CoM Kea LP ROI Des AG Sh “lie”:

seeds RE Da a a Gear Se EA
RAS ete EAS rt A Cae eS
er wD 4 SF Mf er {EERE Se > J; 4 : ON? te
ALY Re a Bets ae Cea BR ck. Ch OSS SMT ears
Porth EL SG CER ee }
f (Fe ay @ BOSS an AY Te ye f i) a4) ¥ HO i ‘ Lo

S- ‘3 > \3 —_—— eyo KAVA Soy f es Li fies RAS x bd A
Pe PS ae RP NS oh AN ys we hee aegis: Ub cdo Ik oe:
RAY | re Al’ RI KA eee. Pa” ROP ssa SK NSA
ne A hoe DO EO ct | See RO IR Oe
Strat EAN UR ROL ee et
Diet te aE 2 ee OR Ba Se ye HO
ee COE NSM pr

ED
e

:
BES,

1? A ps te a
soy oh CORY ne
(aM py y rT he
FNC

et FS ; CDs NY
OSES ST SORTS RS air Rr
PYRE She VO AP NY 9 YS A Te DLA ES SEO PT
EE on y wa See QOS 1 AYU eRe St BAY (See
RSL A TA TD RO tae eae Se eet TS
RO RR ae Se ONE LB EE LA ie
TOMA IY Pe 2 PRR ad SS a C RS AS
a ery | TA Em ch ree Kies
SOM Cae vs Co a KA YRS oh, ay US oD oN
(Oe NAA eh aio ead Koh. ZR NR cS
Y cok EMS ZO ROSY E WE od YS Al
Bee: Je re) ) x APY @ KAY sb ek
Bee ed SRSA Day iota SOs

A a Ose eo Ys SO
RY eae RIN ee
BR SA Sa TRAY OAR AOC WY
ES a CC BF SS oe
yc OF RG RN OE ea NPN Re ED
10 BN Rae 0 YEN Med 5

PBR LRH

a )

a iN we ea SA SB) Ce pt, Wi
oe ae oe oa i Re eoy oe Bee i
bod on Se a SAS SES Sie Cy ee
SS es Sf det ie ae a Argiey aoe oes wy
oa ees ee pe S
me y te rane R os

Wola gh MPS ; eS

CO o PORES 2
ot: fie alr Hae ae Le ae oy ATS on ces
a eae Bike re

» vores A Nee On
ce cee eae

ee 216 se
AY a ae Loa ee e Cs hes tS, enka gS ‘i
Eas Br , Et ed a ALON) Leal a et

‘ Sa a
ve SNe eS De uy
a ft Ae ane ns SENS ae

ee ages ATL WE)

Pat Cae oe ce ae Sy Lae,
AS Ae ma \

: re von UPS Pi
: a ne ee Poe ee

Um

S HR UE TA rey
ee ee oe ey ue pas
BS S oe Oe ES a Sh ee TRY

ao oe me

7s.) sa

~

SMITHSONIAN INSTITUTION
UNITED STATES NATIONAL MUSEUM
Builetin 76

ASTEROIDEA OF THE NORTH PACIFIC
AND ADJACENT WATERS

BY

WALTER KENRICK FISHER
Assistant Professor of Zoology, Stanford University, California

PART 1. PHANEROZONIA AND SPINULOSA

WASHINGTON
GOVERNMENT PRINTING OFFICE
1911

BULLETIN OF THE UNITED STATES NATIONAL MUSEUM
Issvep June 30, 1911.

ADVERTISEMENT.

The scientific publications of the United States National Museum consist of two
series, the Proceedings and the Bulletins.

The Proceedings, the first volume of which was issued in 1878, are intended pri-
marily as a medium for the publication of original, and usually brief, papers based
on the collections of the National Museum, presenting newly-acquired facts in
zoology, geology, and anthropology, including descriptions of new forms of animals,
and revisions of limited groups. One or two volumes are issued annually and dis-
tributed to libraries and scientific organizations. A limited number of copies of
each paper, in pamphlet form, is distributed to specialists and others interested in
the different subjects as soon as printed. The date of publication is printed on
each paper, and these dates are also recorded in the table of contents of the volumes.

The Bulletins, the first of which was issued in 1875, consist of a series of separate
publications comprising chiefly monographs of large zoological groups and other
general systematic treatises (occasionally in several volumes), faunal works, reports
of expeditions, and catalogues of type-specimens, special collections, ete. The
majority of the volumes are octavos, but a quarto size has been adopted in a few
instances in which large plates were regarded as indispensable.

Since 1902 a series of octavo volumes containing papers relating to the botanical
collections of the Museum, and known as the Contributions from the National Her-
barium, has been published as bulletins.

The present work forms No. 76 of the Bulletin series.

Ricuarp RATHBUN,
Assistant Secretary, Smithsonian Institution,
In charge of the United States National Museum.

Wasuineron, D. C., May 4, 1911.

il

TABLE OF CONTENTS.

Page.

TUES GUC AO Tate ata wc Taree a Lari eP eT Re NE hy PR oA ey Le 1
Classification of the Asteroid cases nrr cin 2 ee acre oho): es eR a ie seating SS Ge Se 4
History of systematic work on North Pacific Asteroidea................2.---eeeeeeceeeeeeeeee 5
Distribution and relationships of North Pacific Asteroidea..............................-..... 10
Sve vemia tic) GiscUBsION Othe Tali Als sale Cyst Ses ie arte Ses anja Nee ne oe areae See Se eee < 16
Ke vatonibe Orders Ok -AStCLOLd Galt = cs aennoe Meese eee She a eal 16
Keyitoviheiknowndamiliediof Phanerozonia=ss--) sete cine Sane yen Je eronck occ. 17
Characters used in the classification of Phanerozonia..............-.-.-.-----+-serereeeure-eee 18
Warn veh orcellanasteridsee == ster Santa eae oe Ae Sa ee OY ee dg 22
Key to the known genera of Porcellanasteride..__.........-...-----se-+e----------- 23

Sub iam ysPorcellan astern sess. etiam See a ee ea ee a Be ie ee 23
GeTmUR PET ENILCOS Cena seen cates ae eens oe oe ee ee yet a een ee ee ee 23
Pupiamily Ctenodigcinsessjsss > cele tees sa ee a ne eC a eae ae 31
GenusiGlencdiscus tease cn nc sae ted ee SEER EINE) Ae Ts Uy RN pee Op ae 31

amily PA stLOpec iW dest joo ts Fase aka es ee See) es Bene Se) ee a 37
Key,toitbe known: genera of Astropectinidse52 54.0445. shee ae aeeeaee seen 39

CGonunp epi Chaster erat ee is 2 oo Sos A aaa) ASS. eS ne Se io ARE 42
GenUvAS iO MEClEn Nene, ca os ors eee te Ura Poe Se neni = Mea ee eae 55
Gomustisilastentt anaes ae a2 sc ae eel oe oe aoe Ree ON tor chy es ER BE toes 71

emt Mintssacanthaena anos. 2 SOY See. Sale OME PS Sey RE ee ee Pe 78
GenuntDupsacaster: Saas. Sines = case so 25 ees i es I SocHerBeGeaetate tee 85

Genus yy) tasters ee ears shh Sen ey) Cae oe Big oe ee eee Cys” 101

arn ypu nd ase seein ey St eet he eal OER A UL ou pee oe 105
(CRETE EYOTA tea eR srl Na pee NS SIN le eet, en oS yt vi be vm tal ok PP 105

Family Benthopectinide .........-...........- Bee state eter eee oO ee ee ee 120
Key to the’ known! genera of Benthopectimidse...:.-2.-2--. 5.0... 2 cesses 120

Gonusri ectinasten tots. cta stmt: Slates Sen oot We hepa Wh net Mees 5. Sling 122
Gonusiiuidvasterss ie eee Ake tet Me eR eee ee ae See Dens ee! eS meh. 127

Gens Near chasters into ene Pee atte ee ON Fan, 9 Ean 2 ORR ye sea gee 132
GonuskMyOTiOtus i ee se eeee te eanecna a gets cee mon Garton aaNeiins tec oe wee aoe 140

GOnUs TR entho DECten ane mete neem ea eer ree eee rani Se ene eae ae ae 142

Family Odontasteridse ssi (Sos ate eo acerca eae ee Be a Phe ee 153
Key to the known! genera of Odontasteridse- 2 7.222523 2222222225202 2 oes nee 153

(Gears Odartlirster masts era yc etre eee eee be Oe Le Wen Bt ee 154

(Ramil yGonissteridtess <- 7). 34.- eter eaten cee eRe eee ate ee ee hs 158
Notes on certain genera of the Goniasterid .............-.......----.--------------- 163

Key tothe known genera of Goniasteridse ~~ (2 220202 e222 oe oe oe enn cee eeecace- 169
Sublamily#Mimaatermses).ce oss eases ss cee lates ce ok oe coe Uetdn Soe seme amen cae wee. a 175
Genus Genhyr caster Sasi atc) vce see eas eee wd dee ee eee eT si Fe ene 175
Subiamily, Peoudarchasterinse= <2 =e 2s fio oS ae wed so Saas ee aa ee eae ees «Soe Ro 179
GenusePseudarchasters = 5 pee eas - = eee aera ere aeie atte CE eta renee 179
Sublamily;Goniasterinte ne s- eeece ea. voeseeeee SP erecta fae sso mani Sa trate ee 196
GQenrisNMediasten sa srcte ree ee oe ce tee ae Nee cee a Ee on eee 196

Genus Ceramiastenm cece e eee Sats oe os ea Seen neers | Sean eSoewe catuwese 204

Gonna Cladaster ase ae aces crian see se co cnies saw sean amine aliceweuc sos eeeu sees 221

vi TABLE OF CONTENTS.

Page
Family Goniasteride—Continued.

Subfamily Hippasteriine..........-.--.----- 22 e eee eee ee eee eee eee cence eee cece eee 223
Genus Hippasteria..-......0- 222. e eee e cee teen eee cree tees ener eter ecenecescere: 223
Genus Cryptopeltaster ........-.-.-- 2-0-2 2e eee eee ete eee cence cece ence e eee 237

Family Linckiidee..........- 2.02. ccce esc eee ee cece ee ne nn ence ne cca enscstenscmsescare 240
Key to the known genera of Linckiide.......--..-------+---+---+--+- +++ +252 eee eee 241

GR ENNTIB LATUCR GL eae mcm eterna ta at eae a ew af faa 242
Mamiliy Agteropidee's cos ele oc ccm lola ia wieraisiis oie mi ian tala om wloscin oni = min ole nfase ayn mia micitaln tote erate 247
Key to the known genera of Asteropide..........-------+--------+---++--22+--++---- 247
Gemua 2 ermasterntds io <2 oe nae cer ae ein San ee ee ee aaa ar ee err 248
Order Spimulosa sass sce sone cee oo ee a ela are aia alae win e Say age ele oe = eae ieee 251
Key to the known families of Spinulosa............-.------<-222--2++- 20020222 ee eee nee 251
inn hae ch (aoe ae ar Ree Aan ee hap E ad onerc nd suse I aeneeernoEsatSdataceoescr 253
Key to the subfamilies and known genera of Asterinide ........-...---------+------+--- 253
ASOTIUIB A SEETATUG ike we rec wim lee alelnle sre tia asa te eae alm een acer e ape et a ora ama 254

PEP eATDRT pr CERIN STON CH ete aap at wt ntact ats fale et oleae ate) lee eee oe eee 258
Key to the known) genera of Mchinasteridse. <2 = omc ies ot cjn n-ne ee meee 259
GONUE PE ONAMO DSSS ane as <2 wma sin ote wes aio See ose oeriesee ene asim eee eee 260
Genuatienricianes = serials a atainn/ala'aaatsoe ees ae Be eee eels ate ssn cee ee ee Sone 266
RST VSO MAURIE a = Se lean ta alatel tle eraiatelctormiansiernet arate seera ee aisle ina intra AS et ee 305
Key to the known genera ‘of Solasteridee: . -- ~~ ..-...02 2 32s sn asnet 5 ep 25s eee 306
Gremuis) Solaster << a) cpa aiasa a a oat wien ajale moins wha 5 ate Ss clalaaaisinialt ots = ajo efas ies) sa ee eee 306
Genus Helerozonias! ==). c= ssc ec ee ee seers Face stiera Sel eee ete eee 330
Genus Pro piisten eo otfo5 oc a. eiatictecie nee cine =o sige rae TON arse Se nae oA ee 334
Family Kore thrasteridee so /sos2/. 0 sacle screenees eines esi, ce se ce eels occ ee ne ee eee eee 340
Key to the known genera of Korethrasteride.... - Me Sn eeie Sein chee Sere eee eee 341
Genus eembolaster os sc. oe a8 wot react eh see se ee i ee ee ee eee 341
Mani pab berasterid ge 8 a= /atatctaajaicictete rs aratete wis efaiets, ce we etn mee ee chee tee eee eee Ree 343
Key tothe knowngenera of Pterasteride -. 22:2... - (22sec. 2s shone e de ee ee oe ee 344
Genus Pteraster ..... Beet eee sta alate olen eile eee nore eee es ae eee eet saree 344
Key tothe known species'of Pteraster:.j222<.<: 2. 6522 ee sen le sae ee eee 368
Genus Dinlo pieraster 25.2" Aocte Poe ibe os emcees ca het bc See cee Se ee 370
Mem Ey mendasten ee aos oho Sek oa nee SE ke hes doe es 373

BID MOSTA DU yen tate ea eee tes ee ee Re ROR Pe here Site, See. a a 385
Woxplansiton) Ob pla tease fr. xoaclas ech ones 6 <b soe te ascsh eee Oe er 391

TMU RP ee eee eine ree acne RAs ee ee oe ae hen 407

ASTEROIDEA OF THE NORTH PACIFIC AND
ADJACENT WATERS.

By Water Kenrick FIisHer,
Assistant Professor of Zoology, Stanford University, California.

PART 1. PHANEROZONIA AND SPINULOSA.

INTRODUCTION.

The region covered by the present report embraces the western coast of North
America from the thirty-second parallel of latitude to Point Barrow on the Arctic
Ocean, all of Bering Sea, the coast of Asia from East Cape to Sakhalin, and the
Kuril Islands. It thus includes all the waters north of a line drawn from the south-
ern end of Sakhalin to the southern boundary of the United States. Information
concerning the fauna of the Sea of Okhotsk is very meager; and within the region
outlined nothing is known concerning the starfishes of the great depths of the mid
North Pacific.

In the preparation of this report six thousand nine hundred and twenty-seven
specimens have been listed, and many more examined. These specimens belong to
the following collections:

1. The very extensive collection of the U. S. National Museum, the bulk of
which was obtained by the U.S. Fisheries steamer Albatross between 1888 and 1897,
inclusive. The dredge hauls at which Asteroidea were taken number about two hun-
dred, and the specimens listed number four thousand three hundred and eighty.
Many specimens were taken by Dr. W. H. Dall and others in Alaska.

2. A collection made by the Albatross in Alaska in 1903, and forwarded to
me for report by the U. S. Bureau of Fisheries.

3. A large collection made by the Albatross off California, principally in the
regions off San Diego and Monterey, during the spring of 1904, and sent to me by
the U. S. Bureau of Fisheries.

4. A collection made by the Albatross during the summer of 1906, in the
region of the Aleutian Islands, Kamchatka, and the Kuril Islands.

5. A small collection, principally from off San Pedro, California, and belonging
to the University of California, kindly lent by Prof. W. E. Ritter.

6. The small collection of the Department of Zoology, Stanford University,
principally from Alaska, and Monterey, California.

7. The Museum of Comparative Zoélogy, through Dr. H. L. Clark, has ‘ent
several specimens for examination.

57444°—Bull. 76—11——1 1

2 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

I had completed special reports on collections Nos. 2 and 3 when the collection
from the U.S. National Museum, sent by Dr. Richard Rathbun, Assistant Secretary
of the Smithsonian Institution, was received. This is probably the largest and most
complete single collection ever assembled from a restricted region (the west coast of
North America). Doctor Rathbun at one time contemplated working up this
material, and to that end made a preliminary sorting of specimens in several groups.
He was, however, prevented by routine work from carrying out his plans.

In view of this new accession, it was deemed best not to publish separate papers
on the Alaskan and Californian collections, but to combine all in one report. Under
each species the source of the material is in all cases given in the lists of speci-
mens examined, so that it is easy to ascertain from which of the several collections
the information was derived.

The material proved to be richer than was anticipated, and it has been neces-
sary to issue the work in two parts, of which this is Part 1. Part 2 will contain
an account of the order Forcipulata.

In 1907 I examined, through the courtesy of Mr. Samuel Henshaw and Dr.
H. L. Clark, the extensive collection of starfishes in the Museum of Comparative
Zoélogy, Cambridge, Massachusetts. This collection contains many types and
includes numerous genera not in the national collection. Prof. A. E. Verrill, of
Yale University, New Haven, Connecticut, kindly showed me many of his types,
especially those of his new genera of Goniasteride, Astropectinide, and Gonio-
pectinide.

At Washington I examined all of the national collection, with a view to
determining, if possible, the relationships of genera and families, a subsidiary task
of the present work.

Since many genera had never been assigned a definite type, I made a list of all
the genera of Asteroidea with their type-species in order to render the nomenclature
of the present report as stable as possible. This necessitated drawing upon the
resources of the following libraries: those of the National Museum and Smithsonian
Institution, the Congressional, Army Medical, Philadelphia Academy of Natural
Sciences, Museum of Comparative Zodlogy, and the private library of Dr. C. Hart
Merriam. The nomenclature follows without compromise the rules of the Interna-
tional Code. It has been necessary in a few cases to alter familiar names, because
they have long been incorrectly applied. It is not anticipated that this course
will greatly inconvenience anyone.

In the preparation of this paper several objects have been kept in mind. These
are as follows:

1. To describe and figure all forms known to occur in the region covered by
this report in such a manner that reference to the scattered literature will not be
necessary. It has not been possible, however, to figure a few species, as specimens
were not accessible:

2. To throw as much light as possible on the morphology, especially whenever
this promised to lend aid to classification. A good many forms, especially in the
Astropectinide and Goniasteride, have been examined with this object in view:

3. To revise the classification of genera and families and to work toward as
natural a system as possible:

ASTEROIDEA OF NORTH PACIFIC AND AQJACENT WATERS—FISHER. 3

4. To describe the variations of the different forms represented as thoroughly
as the material would permit. In many instances large series of specimens em-
phasized in a very striking manner the wide and remarkable variations which
characterize most species of Asteroidea:

5. To describe the young stages of the different forms:

6. To consider the affinities of the Asteroidea of the North Pacific to the
Asteroidea of other regions; under each species the geographical distribution is
given, and, whenever possible, its probable relationships are considered in detail:

7. To present keys to all the families of Asteroidea, and to all the genera of
every family represented in the region under consideration. Care has been ex-
ercised to eradicate the numerous errors of current keys, but it is too much to hope
that mistakes do not still persist. A considerable portion of the time spent in
completing this report has been consumed in verifying structural details in genera
not directly considered; but, as genera have greatly multiplied in the last ten years,
it is believed that these keys will prove useful to students of Asteroidea.

In these days it is impossible to complete a piece of systematie work without
the aid of others, and it becomes a pleasant duty to acknowledge this cooperation,
always cheerfully rendered.

Dr. Richard Rathbun, Assistant Secretary of the Smithsonian Institution, in
charge of the U. S. National Museum, has allowed the freest use of all material in
the museum under his charge, and forwarded the extensive collection upon which
this report is primarily based. As previously stated, Doctor Rathbun at one time
contemplated preparing a report on this collection. Throughout the progress of
the work Doctor Rathbun has rendered every possible aid. In 1907 I worked for
a few months at Washington and Cambridge under a grant from the National
Museum.

I am under obligations to Dr. H. L. Clark, Mr. Samuel Henshaw, and Prof.
A. E. Verrill for the privilege of examining collections in their charge. Doctor
Clark subsequently sent several specimens for examination, and Professor Verrill
examined for me specimens of Solaster and forwarded photographs of several of
his new species.

The U.S. Bureau of Fisheries has allowed the use of its collections made
during three cruises of the Albatross under the direction of Doctors Jordan and
Evermann (1903), Doctors Jordan and Gilbert (1904), and Doctor Gilbert (1906).

To Dr. Theodore Gill, Dr. D. S. Jordan, Dr. L. Stejneger, Mr. H. C. Oberholser,
and Mr. Austin Hobart Clark I am indebted for advice and aid in untangling
nomenclatural knots. Mr. Clark has not only looked up literature inaccessible to
me, but also both he and Miss Mary J. Rathbun have greatly aided me by reading
a set of proofs.

Perhaps not least important has been the encouragement I have received from
Doctor Jordan and Doctor Gilbert, for the course of the systematist lies rather
more frequently through briar tangles and over rough places than where the way
is open and fair.

The drawings and negatives of the photographic illustrations were made by
the writer. The prints of the latter were made by Mr. T. W. Smillie, of the U.S.
National Museum.

t BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
CLASSIFICATION OF THE ASTEROIDEA.

In this paper an attempt has been made to combine the best features of the
classifications of Sladen, Perrier, and Verrill, but the results of my own work have
led me to modify my former scheme (1906) in many important details. The
Astropectinide has been enlarged and the Plutonasterid abandoned, a part of
its genera going to the Astropectinide, a part to the Benthopectinide, and Mimas-
ter to the Goniasteride. It was a very unnatural group and the reasons for the
change are discussed under the Astropectinide. The Goniasteride has been
enlarged to include the old family Antheneide and a new subfamily for Chitonaster
has been instituted. Chztaster has been removed from the Linckiide and regarded
as a separate family, Cheetasteride Ludwig, 1897. The Acanthasterine and Val-
vasterine have been raised to family rank.

The weakest point in every scheme of classification of the Asteroidea is between
the Phanerozonia and Spinulosa. Under the latter order Perrier ranges Porania
and its allies in the Poraniidx, while Dermasterias, Asterope, and Petricia are placed
in the Phanerozonia, or rather Valvata. I have merged the Poraniide with the
Asteropide (Gymnasteriide of authors) as a phanerozoniate family. Sladen regards
Asterina and its allies as belonging to the Phanerozonia. Ihave placed them in the
Spinulosa, following Perrier’s system. Sladen classifies the Linckiide in the Crypto-
zonia near Zoroasteride; I have placed the family in the Phanerozonia. The
meaning of this is that no sharp line exists between the Phanerozonia and Spinulosa.
It is, however, not at all difficult to separate the Forcipulata from the Spinulosa.

In grouping the families I have followed as nearly as a lineal arrangement
will permit what seems to be the most natural order, but there are two or three
other combinations which are probably just as ‘‘natural.’’ The Porcellanasteride
are placed first to avoid sandwiching them between the Astropectinide and some
other family, and because they stand apart from the other Phanerozonia.

The following classification is used in this paper:

ECHINODERMA ASTRORADIATA.

Class ASTEROIDEA.
Order PHANEROZONIA.

Family PORCELLANASTERID.

Subfamily PORCH LLANASTERIN 2%.
Subfamily CTHNODISCIN 2*.

Family GONIOPECTINIDAD.
Family ASTROPECTINID®.
Family LUIDIID/.

Family BENTHOPECTINID.
Family ARCHASTERID.®.
Family CHAXTASTERID.
Family ODONTASTERID®.
Family GONIASTERID2.

Subfamily MIMASTERIN At.
Subfamily PSEUDARCHASTERIN 2
Subfamily NECTRIIN_®.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER.

or

Subfamily GONIASTERIN 2.
Subfamily HIPPASTHRIIN 2.
Subfamily LHPTOGONASTERIN_X.
Subfamily CHITONASTEHRIN 4.
Subfamily ANTHEHENBHIN <4.

Family OREASTERID®.
Family LINCKIIDZ.
Family ASTEROPID®.

Order SPINULOSA.

Family GANERID.

Family ASTERINID.

Subfamily ASTHRININ 4.
Subfamily ANSHROPODIN 2.
Subtamily TREMASTEHERIN 2.

Family ECHINASTERID/.
Family ACANTHASTERID®.
Family MITHRODIID®.
Family VALVASTERID.
Family CRYASTERID/.
Family SOLASTERID/.
Family KORETHRASTERID®.

Family MYXASTERID®.

Subfamily PYTHONASTERIN 2
Subfamily MYXASTERIN 42.

Family PTERASTERID.
HISTORY OF SYSTEMATIC WORK ON NORTH PACIFIC ASTEROIDEA.

As compared with the North Atlantic, little work has been done on the Aste-
roidea of the region under consideration, notwithstanding that the west coast of North
America is more prolific in species and individuals than any other portion of the
world. The littoral fauna is especially well developed, and the family Asteriide
contains a surprisingly large number of forms. The following sketch is intended to
serve merely as an historical outline and not as a complete summary of every paper
published. The full titles are given under the Bibliography.

The earliest as well as one of the most important papers was the Prodromus of
Brandt which appeared in 1835 and contained short Latin diagnoses of starfishes
collected by H. Mertens in the vicinity of Sitka, Bering Strait, and Kamchatka.
The new forms named by Brandt are:

Asterias miniata. Asterias pectinata.

Asterias ochracea. Asterias helianthoides.

Asterias janthina. Asterias affinis.

Asterias epichlora. Asterias alboverrucosa.

Asterias camchatica. | Asterias endeca, var. decemradiata.

In the Annals and Magazine of Natural History, volume 6, November, 1840,
J. E. Gray described Asterias katherinze from the mouth of the Columbia River,

6 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Miller and Troschel, in their System der Asteriden, 1842, described Astera-
canthion margaritifer, which is synonymous with Pisaster ochraceus (Brandt).

In Middendorft’s Reise in den diussersten Norden und Osten Siberiens, 1851,
Brandt described, from the Okhotsk region, Asteracanthium ochotense, near A. rubens,
with two varieties (a) macrobrachia, (b) brach ybrachia.

In 1857, Edward Grube, in his Beschreibungen neuer oder weniger bekannter
Seesterne und Seeigel, described and figured Asteropsis imbricata, collected by
Dr. Wilhelm von Bock at Sitka.

In 1857, William Stimpson, in the Boston Journal of Natural History, volume
6. described Asterias gigantea, Tomales Bay, California, Asterias brevispina, San
Francisco Bay (both Pisaster), Linckia leviuscula (Henricia), Puget Sound, and the
new genus and species Mediaster equals, Puget Sound.

This article was followed in 1862 by a paper in the Proceedings of the Boston
Society of Natural History, volume 8, in which the following west coast forms were
diagnosed:

Asterias acervata, Bering Strait.

Asterias exqualis, Monterey.

Asterias capitata, San Diego.

Asterias conferta, Puget Sound.

Asterias cribraria, north of Bering Strait.
Asterias fissispina, Shoal Water Bay, Oregon.
Asterias hexactis, Puget Sound.

Asterias Viitkenii, coast of Oregon.

Asterias paucispina, Puget Sound.

Asterias troschelii, Puget Sound.

In 1865 Edward Grube in an article entitled, Ueber einige Seesterne des Bres-
lauer Museums, described, without locality, Chetaster californicus. This species is
possibly Henricia leviuscula.

In the following year he diagnosed Luidia foliolata (the L. foliata of Sladen),
without locality.?

In July, 1867, A. E. Verrill published an important article On the Geographical
Distribution of Echinoderms of the Pacific Coast of America.?

The following species are listed from Sitka: Asteropsis imbricata Grube,
Patiria miniata (Brandt), Solaster decemradiatus (Brandt) Stimpson, Pycnopodia
helianthoides (Brandt), Asterias epichlora Brandt, Asterias ochracea Brandt. The
following are stated to occur in Puget Sound and along the coast to Cape Mendocino:
Mediaster xqualis Stimpson, Patiria miniata, Cribrella leviuseula (Stimpson),
Pyenopodia helianthoides, Asterias epichlora, A. ochracea, A. conferta Stimpson,
A. fissispina Stimpson, A. litkenia Stimpson, A. paucispina Stimpson, A. troschelir
Stimpson, A. hexaxtis Stimpson. The following species are listed as occurring
between Cape Mendocino and San Diego: Mediaster xqualis, Patiria miniata,
Pyenopodia helianthoides, Asterias gigantea Stimpson, A. brevispina Stimpson, A.
ochracea, A. capitata Stimpson, A. xqualis Stimpson.

For nearly ten years no new species were characterized. Perrier in his Révision
des Stellérides (1875) diagnosed two new forms, Asterias brachiata from the Gulf

® 43 Jahresber. schles. Ges. vaterl. Cultur, Breslau, 1866, p. 59.
6 Trans. Connecticut Acad. Arts and Sci., vol. 1, pt. 2, p. 323.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS——FISHER. i

of Georgia,* and Asterras vancouveri from Esquimalt Harbor, Vancouver Island.
The latter was briefly described in the Annals and Magazine of Natural History.?
Here also was named, with very insufficient description, Asterias douglasi, which had
been confused, in the British Museum, with the true Asterias katherine Gray, from
the mouth of the Columbia. A. douglasi was without locality and probably does
not belong to our region. It is fully described in the Revision.

In an article entitled Etude sur la répartition géographique des Astérides ¢
Perrier named but did not diagnose Luidia californica, a species mentioned but not
described by Liitken.4 The name is a nomen nudum.

A. E. Verrill, in 1880, described Archaster (now Lwidiaster) dawsoni from off
the Queen Charlotte Islands, 111 fathoms, Solaster stimpsoni from Ramsay Island,
British Columbia, and S. dawsoni from 8 to 15 fathoms Virago Sound, British
Columbia, in an article by J. F. Whiteaves entitled: Appendix C. On Some Marine
Invertebrata from the Queen Charlotte Islands.¢

In Ray’s report of the international polar expedition to Pomt Barrow, Alaska,
1885, John Murdoch listed six species of starfishes, of which one was described as
new, Leptasterias arctica, from off Point Franklin, 134 fathoms.

In the following year H. Ludwig published the first important article dealing
with the echinoderms of Bering Sea. He listed six starfishes, of which Ctenodiscus
krausei (=crispatus) and Pteraster aporus (=militaris) were characterized as new/

P. de Loriol, in Notes pour servir & l’étude des Echinodermes, II, described two?
new species collected by Alphonse Forrer at Santa Cruz, California: Asterias forreri
and A. exquisita.

In 1888 J. E. Ives described Pteraster tesselatus from Kadiak Island.” In 1892
the same writer published privately a List of the Echinoderms and Crustacea in the
Cabinet of Frederick Stearns, Detroit, Michigan. On page ii is a list of west coast
Asteroidea, but no new species. I am indebted to Prof. A. E. Verrill for information
concerning this paper.

In the Proceedings of the U. S. National Museum for 1894, page 268, A. E.
Verrill characterized Acantharchaster, new genus, with Archaster dawsoni for type.

P. de Loriol in a fifth paper of his Notes pour servir 4 1’étude des Echinodermes‘
admirably described and figured seven species from Saanich Inlet, Vancouver Island,
three of them new:

Crossaster vancouverensis. | Asterias kahleri.
Asterias saanichensis.

In May, 1901, an important paper by H. L. Clark appeared in the Proceedings

of the Boston Society of Natural History, volume 29, pages 323-329 (Asteroidea),

> <= —— as —

a@ Specimens in the Museum of Comparative Zoélogy, as for example, No. 1191 are so labeled.
Perrier’s specimen came from this Museum.

b Ser. 4, vol. 17, Jan., 1876, p. 187.

¢ Nouv. arch. du mus. d’hist. nat. Paris, ser. 2, vol. 1, 1878, pp. 35 and 91.

@Vid. Medd., 1871, p. 229.

¢ Report of Progress; Geol. Survey of Canada for 1878-79, 1880.

? Echinodermen des Beringsmeeres, Zool. Jahrb., Abth. Syst., vol. 1, 1886, pp. 287-295.

9 Rec. zool. Suisse, vol. 4, No. 3, June 23, 1887.

h Proc. Acad. Nat. Sci. Philadelphia, p. 421.

4 Mémoires soc. phys. et d’hist. nat. Genéve, vol. 32, 1897.

8 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

on Echinoderms from Puget Sound: Observations made on the Echinoderms
collected by the parties from Columbia University, in Puget Sound in 1896 and
1897. Eight species of starfishes are listed, of which three, Pteraster multispinus,
Retaster gracilis, and Cribrella spiculifera, are new; and two new varieties of Cri-
brella leviuscula are characterized: C. leviuscula crassa and (@. leviuscula attenuata.
There is reason to believe that Echinaster tenuispinus was not from Puget Sound.
I am informed that specimens from the West Indies became mixed with the original
Puget Sound collection before it was sent to Doctor Clark.

Up to 1905 the fauna of the deep water off the west coast of North America
was practically unknown, for only one species had been described (Archaster daw-
soni, by Verrill, 1880), although a wealth of material had been accumulating in the
U.S. National Museum since 1888, as a result of the numerous cruises of the U. S.
Fisheries steamer Albatross.

In the Bulletin of the Bureau of Fisheries for 1904, volume 24, June 10, 1905,
W. K. Fisher described one new genus, two new subgenera, and twenty-four new
species based upon collections made by the Albatross in Alaskan waters during the
summer of 1903 and off the coast of southern and central California in 1904, as
follows:

Eremicaster, new subgenus. Hippasteria heathi.
Porcellanaster (Eremicaster) tene- | Hippasteria californica.

brarius. Cryptopeltaster, new genus.
Bathybiaster pectinatus. Cryptopeltaster lepidonotus.
Dipsacaster eximius. Lophaster furcilliger.
Persephonaster penicillatus. Peribolaster biserialis.
Benthopecten acanthonotus. Pteraster jordani.
Dytaster gilberti. Hymenaster quadrispinosus.
Mimaster swifti. Zoroaster ophiurus.
Odontaster crassus. Myxoderma, new subgenus.
Pseudarchaster alascensis. Zoroaster ( Myxoderma) sacculatus.
Pseudarchaster pusillus. Zoroaster (Myxoderma) evermanni.
Tosia leptocerama. Brisinga exilis.
Mediaster tenellus. Freyella fecunda.

Although not directly concerned with the region under consideration, the
report by H. Ludwig on the Asteroidea collected off the Galapagos Islands, Mexico,
Central and South America, by the Albatross in 1891, is of importance because
some of his new species occur as far north as Alaska (as, for instance, Porcellanaster
pacificus, P. waltharii (= Eremicasier tenebrarius Fisher) and Plutonaster abyssicola
(=Psilaster pectinatus Fisher). Others are very closely related to forms described
in the following report. Ludwig’s paper appeared July 17, 1905, as volume 32 of
the Memoirs of the Museum of Comparative Zodlogy.

In the Zoologischer Anzeiger for June 19, 1906, W. K. Fisher described Astro-
pecten californicus and Alexandraster inflatus (=Poraniopsis inflata) from Monterey
Bay; and in New Starfishes from the Pacific Coast of North America,” one new

@ Proc. Washington Acad. Sci., vol. 8, 1906, pp. 111-139.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 9

genus and ten new species, based on the National Museum collection, were char-
acterized :

Leptychaster pacificus. Henricia aspera.

Leptychaster anomalus. Henricia polyacantha.

Astropecten ornatissimus. Crossaster (now Heterozonias) alternatus.
Tnidia ludwigi. Crossaster (now Solaster) borealis.
Tnidia asthenosoma. Rathbunaster californicus (new genus).

H. Ludwig in the Zoologischer Anzeiger for March 5, 1907 (p. 317), raised
Eremicaster Fisher to a full genus and changed the type to Porcellanaster crassus
Sladen. In the same journal for July 23, 1907, W. K. Fisher criticised the char-
acters used by Ludwig to separate Eremicaster from Porcellanaster and called
attention to the fact that this writer had no power to change an established type-
species, which is tenebrorius, not crassus.

In the Smithsonian Miscellaneous Collections (Quarterly Issue), volume 52,
May 27, 1908, W. K. Fisher called attention to a number of generic names which
had long been incorrectly used, and proposed substitute names. The changes
instituted, some of which affect the nomenclature of the present paper, were as
follows:

Anasterias Perrier has precedence over Sporasterias.

Anasterias Ludwig (not Perrier) becomes Lysasterias Fisher.

Gymnasteria Gray becomes Asterope Miiller and Troschel.

Asteropsis of authors (not Miller and Troschel) becomes Petricia Gray.

Crenaster Perrier (not d’Orbigny) becomes Dytaster Sladen.

Ctenaster Perrier (not Agassiz) becomes Letmaster Fisher.

Diplasterias Perrier is a synonym of Pisaster Miiller and Troschel.

Goniodon Perrier (antedated by Goniodon Herrick) becomes Diplodontias
Fisher.

Patiria Gray is a synonym of Asterina Nardo.

Patiria Perrier (not Gray) becomes Parasterina Fisher.

Pararchaster Sladen is a synonym of Benthopecten Verrill.

Pentaceros Schulze (nonbinomial) becomes Oreaster Miiller and Troschel.

The same facts were published in the Zoologischer Anzeiger for August 18, 1908.

A. E. Verrill, in Descriptions of the New Genera and Species of Starfishes from
the North Pacific Coast of America,* diagnosed the following forms:

Solaster galaxides, Victoria, British Columbia.

Solaster constellatus, Puget Sound.

Pteraster octaster, Bering Island.

Pteraster hebes, Departure Bay, British Columbia.
Hippasteria spinosa, Departure Bay, British Columbia.
Tosia arctica, Bering Island.

Asterias (Pisaster) papulosa, Vancouver Island.
Allasterias rathbuni (new genus and species) Maloska.
Asterias (Urasterias) forcipulata, Departure Bay, British Columbia.
Asterias polythela, Arctic coast of Alaska.

Asterias victoriana, Victoria, British Columbia.

@ Amer. Journ. Sci., July, 1909, pp. 59-70.

10 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

In a general article in the American Naturalist for September, 1909, entitled
Remarkable Development of Starfishes on the Northwest American Coast; Hy-
bridism; Multiplicity of Rays; Teratology; Problems in Evolution; Geographical
Distribution, the same writer named a number of new genera and species, some
without description, some with a few lines of description, and three with figures.

New GENERA.

Pterasterides for Pteraster aporus | Bunodaster for B. ritteri Verrill, “Cal-
Ludwig. ifornia.”’

Glyphaster for Leptychaster anom-
alus Fisher.

New SPECIES.

Asterias acanthostoma. Leptychaster millespina.
Asterias dubia. Henricia tumida.
Asterias epichlora, var. alaskensis. | Henricia spatulifera.

Nomina Nupa.

Asterias columbiana. Asterias macropora.
Asterias troschelii, var. rudis. Henricia sanguinolenta, var. pectinata.
Asterias forcipulata.

In the Annals and Magazine of Natural History for February, 1910, W. K.
Fisher published diagnoses of Thrissacanthias, Gephyreaster, Spheriodiscus, and
Heterozonias, new genera, and preliminary diagnoses of Pteraster trigonodon, Pt.
coscinopeplus, Pt. marsippus, Pt. temnochiton, Hymenaster kahleri, and H. perissono-
tus. In the Zoologischer Anzeiger for March 29, 1910, twenty-four new species
and subspecies were characterized, as follows:

Leptychaster propinquus. Ceramaster clark.
Dipsacaster borealis. Cladaster validus.
Dipsacaster letmophilus. Hippasteria leiopelta.
Dipsacaster anoplus. Poraniopsis inflata flexilis.
Benthopecten claviger. Henricia leviuscula multispina.
Benthopecten mutabilis. Henricia leviuscula annectens.
Acantharchaster aciculosus. Henricia asthenactis.
Acantharchaster variabilis. Henricia longispina.
Acantharchaster variabilis pedicel- | Henricia clarki.

laris. Solaster exiguus.
Acantharchaster intermedius. Solaster hypothrissus.
Cheiraster agassizi evoplus. Lophaster furcilliger vexator.

Pseudarchaster dissonus. |
DISTRIBUTION AND RELATIONSHIPS OF NORTH PACIFIC ASTEROIDEA.

A full account of the distribution of species must be reserved for the conclud-
ing installment of this paper. It is intended here to present only the main features
of the relationships and distribution of the Phanerozonia and Spinulosa. Tables
of the bathymetrical range of species will also be given in a subsequent report.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. iat

Before commenting upon the distribution it will be necessary to present a few
tables.

1. Specres common TO Norra Paciric AND NortH ATLANTIC.

Ctenodiscus crispatus. Solaster papposus.
Leptychaster arcticus. Lophaster f. vexator ?
Pseudarchaster parelii. Pteraster militaris.
Henricia sanguinolenta. Pteraster obscurus.
Henricia s. eschrichtir. Pteraster pulvillus.
Solaster endeca. Diplopteraster multipes.

2. Species common To NortH Paciric AND SoutrH Paciric.
Ctenodiscus crispatus. | Ceramaster patagonicus.
3. SPECIES COMMON TO WEST CoAsT oF NorTH AMERICA AND JAPAN.@

(Those marked with an asterisk are found also in the North Atlantic; cireum-
polar species.)

’
Ctenodiscus crispatus.* Henricia sanguinolenta.*
Leptychaster arcticus.* Solaster paaxillatus.”
Leptychaster anomalus.® Solaster borealis.”
Pseudarchaster parelii.* Solaster dawsoni (var.)?
Ceramaster japonicus.” Diplopteraster multipes.*

4. SPECIES FOUND ON THE WEST coast OF NortH AMERICA AND EXTENDING SOUTH ALONG
Mexico AND CENTRAL AND SourtH AMERICA.

Eremicaster tenebrarius, to Galapagos Islands.

Eremicaster pacificus, to Gulf of Panama and Galapagos Islands.
Ctenodiscus crispatus, to southern end of South America.
Psilaster pectinatus, to Gulf of Panama.

Astropecten armatus*, to Ecuador.

Astropecten californicus, to Lower California.

Luidia foliolata, to Mazatlan.

Benthopecten acanthonotus, possibly to Galapagos Tslands,
Mediaster tenellus, possibly to Panama.

Cryptopeltaster lepidonotus, to Tres Marias Islands, Mexico.
Linckia columbiz,° to Colombia and Galapagos Islands.
Asterina miniata, to Gulf of California.

Poraniopsis inflata flevilis, probably to Galapagos Islands.

@ Doctor S. Goto’s list of Japanese deep-water forms has not been published; the species listed are
those I have personally examined. This list should be augmented when the report on the Albatross
collection of 1906 is published.

b Not found outside the North Pacific.
c Intrusions into the United States fauna from the Panamic fauna, and shallow-water species. The

rest, with the exception of Luidia foliolata, Astropecten californicus, and Asterina miniata, are deep-water

forms.

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Lophaster furcilliger, to Galapagos Islands.
Hymenaster quadrispinosus, probably to vicinity of Panama.

5. GENERA PECULIAR TO THE NortH Paciric. @

Thrissacanthias (extends into Japanese waters).
Bunodaster Verrill (probably synonymous with Blakiaster).
Nearchaster.

Myonotus.

Gephyreaster.
Dermasterias.
Cry ptopeltaster.
Heterozonias.

6. SPECIES PECULIAR TO THE NORTH Paciric; THEIR DISTRIBUTION AND PROBABLE AFFINITIES.5

Leptychaster propinquus .

Leptychaster anomalus 4.
Leptychaster pacificus. . .
Astropecten californicus .
Astropecten ornatissimus
Thrissacanthias penicillatus.

Luidia foliolata...............
Luidia ludwigi................
Luidia asthenosoma f
Pectinaster a. evoplus
Luidiaster dawsoni

Nearchaster pedicellaris
Myonotus intermedius ........
Benthopecten claviger.........
Benthopecten mutabilis.
Odontaster crassus............
Gephyreaster swifti............
Pseudarchaster p. alascensis. . .
Pseudarchaster pusillus
Pseudarchaster dissonus
Mediaster zqualis
Ceramaster leptoceramus ......
Ceramaster japonicus é

Okhotsk
Sea,Kam-) Bering
chatka, e

Kurils.

Sea.

Puget

Aleutian Sitka to South of
Kadiak Sound to

Islands to Puget | Monterey

Kadiak. | © Sitka.) sound. ied Bay.

@ Glyphaster Verrill, and Pterasterides Verrill, Id t t.
» Bunodaster ritteri Verrill is not included. ft 1s stated ¢

as south of Monterey Bay.

It is stated to occur in ‘“ California.”

4¢Commander Islands only, not off the mainland.

¢ Ranges to Sea of Japan.
¢To Lower California.

/ From north of Monterey Bay to Los Coronados Islands, Lower California.

9 Monterey Bay only.
4 To Oregon. ce
‘To Japan.

4To Washington.

Related species.

L. anomalus.

L. inermis (Ludwig. )
L. arcticus.

A. fragilis?

A Japanese species.
D. borealis.
D. eximius.

.| D. nesiotes.

D. exilis.

L. brevispina.

L. lorioli.

L. sarsi.

P. agassizi, P. mimicus.
L. horridus.

N. aciculosus.

N. variabilis.

B. spinosus.
O. grayi.

Ps. pulcher.
Ps. pectinifer.

Probably its range should be given

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER.

6. SPECIES PECULIAR TO THE NORTH PACIFIC; THEIR DISTRIBUTION AND PROBABLE AFFINITIES—Con.

Okhotsk
Species. chatka,
K

Sea, Kam-

Aleutian
Islands to
Kadiak.

Berin,
Sea. :

PAMETITE MTN an san anion | Seeks ams

Poraniopsis inflata

Henricia leviuscula.....-.-----|..--------
Henricia l. annectens..........|......----

Henricia 1. multispina.....-.--

Henricia 1. dyscrita

Henricia aspera.......--------
Henricia spiculifera .
Henricia asthenactis.
Henricia longispina
Henricia 1. aleutica...-
Henricia polyacantha .
Henricia clarki..
Solaster stimpsoni
Solaster dawsonif.
Solaster parillatus 9.
Solaster exiguus.. .
Solaster borealis 9.....--.-----|.---------
Solaster hypothrissus
Heterozonias alternatus........|...-.----.
Peribolaster biserialis.......---

Pteraster trigonodon.......--.-.].---------
Pteraster jordani
Pteraster marsippus
Pteraster coscinopeplus.......-|..-.--.---

Pteraster temnochiton

Pteraster t. arcuatus
Hymenaster kehleri

Kadiak
to Sitka.

Puget
Sound to ,
Monterey

Bay.

South of
Monterey
Bay.

Related species.

C. rudis.
H. phrygiana.

A. pectinifera.
P. echinaster.

H. pauperrima.

S. syrtensis(?)
S. parillatus.
S. australis.
S. borealis.

Hymenaster perissonotus

« Commander Islands only, not off the mainland.
b Simushir, Kuril Islands.

¢ To the Tres Marias Islands.

@ To Lower California.

¢ To the Revillagigedo Islands, Mexico.
/ Through Bering Strait.
9g To Japan.

14 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Of the ninety-six species and subspecies listed, seventy-three, or over 76 per
cent, are confined to the North Pacific, although five of these extend to Japan
(Table 3) and six others to Mexican waters but not beyond the Tres Marias and
Revillagigedo islands, leaving 61 per cent confined within the geographical limits of
this report.

Of the remaining twenty-three species, twelve are found in the North Atlantic;
and of these twelve, one (Ctenodiscus crispatus) extends also to Japan and to the
South Pacific (Tables 1 and 2), while four others range into Japanese waters. The
other eleven species are part of another fauna and their ranges extend southward
variously to Panama, the Galapagos Islands, Ecuador, and the extremity of South
America. If Ctenodiscus crispatus, a very wide ranging form, is omitted it will be
seen that eleven of the species are derived from the North Atlantic fauna and
eleven others from the South American.

Of the latter eleven, two species belong to the shallow water Panamic fauna and
have pushed northward into southern California. They are—

Astropecten armatus. | Linckia columbiz.

Considering the seventy-three species which are confined to the North Pacific,
only four (Leptychaster propinquus, Hippasteria spinosa kurilensis, H. leiopelta, and
I. leiopelta armata) are restricted to the Kamchatkan district, and two of these to
the area between Kamchatka and Japan. Turning to the west coast of North
America we find a different state of affairs.

1. TWENTY-ONE SPECIES DO NOT EXTEND NORTH OF SAN FRANcIScO Bay, AS FOLLOWS:

Astropecten californicus.* Hippasteria californica.
Astropecten ornatissimus. Cryptopeltaster lepidonotus.
Dipsacaster eximius. Poraniopsis inflata.
Dytaster gilberti. Henricia l. dyscrita.
Luidia asthenosoma. Henricia polyacantha.
LInidia ludwigi. Henricia clarki.

Myonotus intermedius. Solaster exiguus.
Pectinaster a. evoplus. Pteraster trigonodon.
Odontaster crassus. Pteraster coscinopeplus.
Pseudarchaster pusillus. Pteraster t. arcuatus.

Ceramaster leptoceramus.

2. THE FOLLOWING SOUTHERN CALIFORNIAN FORMS EXTEND THEIR RANGE TO WASHINGTON (*)
AND SOUTHERN ALASKA ({) BUT NOT NORTH OF SITKA:

Thrissacanthias penicillatus.* Henricia 1. annectens.*
Dipsacaster anoplus.* Heterozonias alternatus.*
Luidia foliolata.t Pteraster jordani.*

Asterina miniata.+

®To Bodega Head, just north of San F rancisco Bay.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 15

3. TEN CALIFORNIAN SPECIES ARE OF STILL WIDER RANGE, EXTENDING TO THE VICINITY OF Kapiak

IsLnaNp (*) on TO BERING SEA (f).

Nearchaster aciculosus.*
Mediaster xqualis.*
Ceramaster clarki.t
Hippasteria spinosa.*
Henricia leviuscula.t

Solaster borealis RANGES TO JAPAN.

THEY ARE AS FOLLOWS:

Henricia aspera.t
Henricia asthenactis.t
Solaster borealis.}
Peribolaster biserialis.t
Hymenaster perissonotus.t

4. THREE OTHER CALIFORNIAN SPECIES WHICH RANGE LITTLE IF ANY SOUTH OF MONTEREY Bay
EXTEND TO SirKA (*) OR THE ASIATIC REGION (}), AS FOLLOWS:

Leptychaster anomalus.}
Dermasterias imbricata.*

| Solaster dawsont.t

In the above lists there are thirty-three species which do not reach the coast of
Asia and are distinctive West American forms.

Considering the more northern forms, there are twenty-eight species which do
not reach California, and of these twenty do not extend their range to the Asiatic
coast or outlying islands, making with the above thirty-three species, fifty-three
characteristic west coast forms, not found on the Asiatic side.

5. OF THE TWENTY-EIGHT SPECIES WHICH DO NOT RANGE SOUTH OF OREGON, NINE ARE CONFINED
TO THE REGION OF THE ALEUTIANS AND BERING SEA, TWO OF THEM RANGING AS FAR WEST AS THE

COMMANDER ISLANDS (*), AS FOLLOWS:

Dipsacaster borealis.
Nearchaster pedicellaris.
Ceramaster arcticus.*
Cladaster validus.

Henricia longispina aleutica.

Solaster hypothrissus.
Pteraster marsippus.*
Pteraster temnochiton.
Hymenaster kahlerv.

6. THE REMAINING NINETEEN SPECIES ARE FOUND IN THE REGION BETWEEN KAMCHATKA AND
OREGON, BUT THOSE MARKED WITH AN ASTERISK (*) ARE WIDE RANGING FORMS WHICH REACH JAPAN.
THOSE MARKED ** REACH BERING SEA, AND THOSE INDICATED WITH A | EXTEND TO THE ASIATIC

COAST OR OUTLYING ISLANDS,

Leptychaster pacificus.**
Dipsacaster letmophilus.
Benthopecten claviger.**
Benthopecten mutabilis.
Inidiaster dawsont.t
Nearchaster variabilis.**
Gephyreaster swifti.
Pseudarchaster parelii alascensis.
Pseudarchaster dissonus.**
Ceramaster japonicus.*

Hippasteria heathi.

Henricia spiculifera.

Henricia longispina.

Henricia leviuscula multispina.*
Solaster stimpsont.t

Solaster paxillatus.*

Pteraster multispinus.

Pteraster gracilis.

Pteraster tesselatus.t

16 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Even making liberal allowance for our meager knowledge of the region, it will
be evident that the coast of Asia north of Japan is much poorer 1n peculiar species
than the northwest coast of America, the former possessing four and the latter over
fifty characteristic forms.

” Tt has been stated above that 76 per cent of the Phanerozonia and Spinulosa
are indigenous to the North Pacific. It remains to point out that of the intrusions
12 per cent are derived from circumpolar species, while about 12 per cent come
from the south, even as far as the extremity of South America, and that the shallow
water southern forms do not push so far north as the deeper water species. The
case with the cireumpolar species is somewhat similar, for the forms which range the
farthest south are inhabitants of deep water.

The derivation of the indigenous species, or rather of their ancestors, is too
clouded with uncertainty to admit of exact treatment. It may be suggested, how-
ever, without greatly overstepping the bounds of ascertained fact, that these forms
probably came from the same sources as the species not peculiar to the region,
namely, from the Arctic Ocean, from middle and South America, and to a slighter
extent from the direction of Japan. There remain, however, a number of forms
whose nearest relative outside the region can not be ascertained.

SYSTEMATIC DISCUSSION OF THE FAUNA.
Class ASTEROIDEA Burmeister.

Free echinoderms with radially disposed gonads; with the ambulacral append-
ages, which are each connected with a double or single ampulla, confined to an open
ventral ambulacral furrow and regularly arranged in two or four rows; radial
ambulacral water tubes and radial nerves exterior to ambulacral ossicles and not
covered by any plates; digestive system radiate and extending into the rays; respl-
ration by means of dermal gills or papule; madreporic aperture abactinal; anal
aperture usually present.

KEY TO THE ORDERS OF ASTEROIDEA.

a’. Marginal plates usually large and conspicuous, defining the contour of body; abactinal skeleton in
the form of paxilliform plates,¢ or flat, tessellate plates which are smooth or armed with granules or
spines, and either naked or covered with thin or thick membrane. Pedicellaria never peduncu-
late forcipiform, but spiniform, pectinate, valvate, or excavate. Papule restricted to abactinal
area (except in some Linckiide and Asteropide) circumscribed by the marginal plates. Mouth
plates prominent; ambulacral plates well spaced; tube feet in two series; rays usually 5 normally

(except in some Duidia) ...............---------2------------- Phanerozonia Sladen, p. 17.

a?. Marginal plates not usually conspicuously large; abactinal skeleton not composed of true paxilliform
plates, nor in the form of a tessellated pavement, but usually more or less reticulate or imbricated.
Papule frequently but not invariably also intramarginal and actinal. Some form of abactinal
spinulation always present; tube feet with well-developed sucking disks.

b'. Actinostomial ring with adambulacral plates prominent; pedicellaria very rare, never peduncu-
late forcipiform nor excavate; ambulacral ossicles not crowded; abactinal skeleton composed
of thin, close-set overlapping plates, or forming a more or less open reticulate network, either
regular or irregular; plates often cruciform with or without connecting independent essicles.

Spinulosa Perrier, p. 251.

a The abactinal skeleton when paxilliform may be joined by definite and regular intermediate ossicles,
but is never irregularly reticulate; compare Solasteride, which bear pseudopaxille.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. iy

b°. Pedicellarie pedunculate, either forcipiform or forficiform (crossed or straight); actinostome with
ambulacral plates prominent except in Brisingide. Skeletal plates bearing spines, often long
and isolated, on or about which are usually grouped the pedicellariw; or the pedicellariwe may
be isolated; marginal plates inconspicuous or aborted; ambulacral plates often very crowded
with tube feet in four series. Abactinal skeleton formed of skeletal arches (transverse on rays)
independent or bound together by intermediate plates, forming a network with rectangular or
very irregular meshes. These skeletal arches correspond to every other or to every third adambu-
lacral, and are composed of pieces corresponding in the ventral, lateral, and dorsal regions of the
body. Mouth plates usually inconspicuous. .......................--- Forcipulata Perrier.

Order PHANEROZONIA Sladen, 1889.

=Paxitiosa-++ Vatvara (VaLyuara) Perrier, 1884, 1894, Noromyora Ludwig, 1910.
With the characters given above.
KEY TO THE KNOWN FAMILIES OF PHANEROZONIA.

a‘. Tube feet pointed without a definite flat sucking disk, but sometimes with a small pointed knob at tip.

b'. Ampulle single, one to each tube foot. No intestine, no intestinal cecum, no anal pore. Mar-
ginal plates thin and lamelliform. Cribriform organs between the marginal plates.

Porcellanasteridx, p. 22.
c'. No actinal fascioles; cribriform organs localized in interradial arcs; no superambulacral plates;
paxillz or simple embryonic abactinal plates with a few spicular spinelets.
Porcellanasterinx, p. 23.
ce’. Actinal fascioles; cribiform organs of a simple nature and between all the marginals; super-
ambulacral plates present; paxillee...............- Dataepsus fates oes Ctenodiscing, p. 31.
6°. Ampulle double; simple cribriform organs, or more or less specialized marginal fascioles usually
present; abactinal plates always paxilliform; superambulacral plates always present.

c'. Superomarginal plates replaced by paxille; broad inferomarginals; no anus, no intestinal
cwcum,molntestinesicompound| papules .s.5-2-2s--n<-2-c- +2 50-200 ose ce Iwidiide, p. 105
ec’. Superomarginals never aborted, though sometimes small; not replaced by paxille similar to
those of dorsum; intestine present; coecum only exceptionally absent (Blakiaster); anal pore

usually present, frequently of minute size, and sometimes absent; papule simple.
d'. Marginal, adambulacral, and actinal plates bordered by a single row of spinelets united bya
_web-like membrane, forming specialized fascioles identical with cribriform organs; mar-
ginals and actinals usually covered with thin membrane; actinals in double or single
interradial series, between which are the specialized fascioles leading from marginal fas-
cioles to adambulacrals; adambulacral armature simple usually in a single marginal series;
paxillz with marginal spinelets webbed ........................--.---- Goniopectinidx.
d*, Marginal and actinal plates never bordered by a single row of webbed spinelets, nor are spine-
lets of paxille webbed; superomarginals variable; inferomarginals always massive;
marginals never overlaid by a smooth membrane; actinal plates never in double interradial
series, with each double series separated by specialized fascioles; actinal plates with paxilli-
form groups of spines or spinelets; adambulacral armature always in several series, inter-
gradinpinto actinallspineletsr-. sas-se feces cc ceL nesses cence ee eee Astropectinidx, p. 37.

a’. Tube feet with well-developed sucking disks.

b'. Abactinal plates paxilliform or tabulate with groups of coordinated granules or spinelets; or in the
form of spinopaxille; or simple and flattened with few spinelets (in the latter case the marginals
are alternate and very spiny).

c'. Marginal plates appreciably alternate and with sharp conspicuous spines; a pair of muscle
bands extending along dorsal wall of each ray, from region of third to eighth ambulacral
ossicle to tip; papule usually confined to base of ray and adjacent portion of disk, often to a
localized napulariorgan cesses. oscce noc soe nah eeee eres Benthopectinidx, p. 120.

c’. Marginal plates opposite (at least at base of ray) and not conspicuously spiny; no dorsal muscle
bands; papulz not confined to base of ray and adjacent portion of disk and never localized in
a special papular organ.

57444°—Bull. 76—11——2

18 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

@. One or two conspicuous recurved hyaline teeth to each pair of mouth plates; an odd inter-
ranial ammreinnls occ occte sacle nine sani cre letermreictele ote oimse alee stemuneref afta Odontasteridx, p. 153.

@. No conspicuous recurved hyaline teeth; odd interradial marginal present in one genus.

é'. Spinelets of tabulate paxilliform plates very slender and glassy; marginal plates small,
not much larger than adjacent actinal plates; odd marginal present; actinal interradial
areas small; tabulate abactinal plates with internal radiating ossicles; calcareous inter-
brachial septa; ampulle single........---.-----------+-22- eee e ee eee Cheetasteridzx .4

&. Spinelets not slendey and glassy; no odd interradial marginal; abactinal plates tabulate or
paxilliform; ampullz double.

/). Abactinal plates tabulate, paxilliform, arranged in oblique transverse rows on either side
of a conspicuous medioradial series and with special internal imbricating ridges; actina!
intermediate plates aborted or very few; interbrachial septa calcareous; gonads extend-
ing far along rays; general facies astropectinoid ..........--.-.-.------ Archasteridx..d

/?. Abactinal plates tabulate, often paxilliform, not arranged in definite oblique series; no
internal imbricating ridges; actinal interradial areas large, with numerous plates; inter-
brachial septa usually membranous; gonads interradial......... Goniasteridx, p. 158.

?. Abactinal plates are neither tabulate nor paxilliform, but are flat, convex, spinous, tubercular,
granulous, or smooth; are sometimes overlaid by a thin or thick, smooth or granulous, skin;
marginal plates with or without robust spines or tubercles.

c!. Disk large; actinal interradial areas extensive, but no papul on actinal surface.

d'. Marginal plates large, conspicuous; abactinal skeleton composed of polygonal, circular, or
stellate plates united into a firm pavement; with or without secondary abactinal inter-
mediate plates; plates may be smooth, granulous, spiny, tubercular, or covered by a smooth
or granulous skin; papule single, or a few to an area, not numerous and in circumscribed
areas; interbrachial septa usually membranous...............------ Goniasteridx, p. 158.

@. Marginal plates large but as a rule inconspicuous, being more or less hidden by granulous
skin or encroachments of papular areas; abactinal skeleton stellato-reticulate; plates always
granulous; papule numerous, and in definite areas; abactinal plates usually with large
conical tubercles or spines. Disk usually high, or body thick and cushion-like; interbrachial
Bopta/lisvally (calcareous = ae. ojos tomate cise seinem e nereeienieiaete en eiae eae eet eters Oreasteridx.

d@. Marginal plates small, more or less imbricated and whole body covered by thick, smooth,
tough skin; abactinal skeleton tessellate or reticulate, the plates loosely imbricated; prom-
inent spines exceptional, small spines present in some genera........- Asteropidx, p. 247.

ce. Disk small, with very small actinal interradial areas as a rule (if at all well developed, then
papuls on actinal surface); marginals small; skeleton tessellate; tegumentary developments
granulate (skeleton overlaid by a smooth skin, without prominent spines in Leiaster)
Linckiidee, p. 240.

CHARACTERS USED IN THE CLASSIFICATION OF PHANEROZONIA.

In dividing an order into families, one endeavors to use only those characters
which are of more than generic value yet which are not wide enough in scope to be
of use in subdividing the class, that is, are not of ordinalimportance. In determining
the order Phanerozonia the ambital skeleton and distribution of the respiratory

® The classification of Chetaster presents unusual difficulties owing to the intermediate nature of
the genus. It does not appear to belong in the Linckiide, nor in any other family. Ludwig created
the Cheetasteridz asa provisional group. The presence of an odd interradial marginal, and in one spe-
cies of rudimentary recurved teeth suggests Odontaster, but unlike that genus and its allies, the inter-
brachial septa of Chetaster are calcified, and the ampulle single. In some respects the genus shows a
remote alliance with the Ganeriide.

> The old and well-known genus Archaster is an isolated and remarkable type, for, though superfi-
cially resembling Astropecten, it is really near the Goniasteridx. It has heavily calcified interbrachial
septa, no superambulacral ossicles, tabulate abactinal plates, those of the regular oblique transverse
series imbricating by curious internal keels, produced toward the midradial line. The very regular
and enlarged radial series of plates and the readily recognizable primary apical plates recall the Goni-
asteride. The ampulle are strongly double, and the tube feet have heavy sucking disks.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 19

papule are used, although in the transitional or intermediate family Linckiide
these characters are variable, which is tantamount to saying that the order is not
sharply defined. This is indeed the case.

A study of the majority of genera in the Phanerozonia has led me to rearrange
somewhat Verrill’s revision of Sladen’s classification, especially as regards the
Astropectinide. The characters used in determining the families are emphasized
in the key. These relate to the water vascular system, to the modifications of
the abactinal skeleton, to the structure of the alimentary canal, to the modifica-
tions of the marginal plates, and structure of actinal intermediate plates. The
pedicellariz are not used, except in connection with several primary characters.
They are suggestive for determining the relationships of species and genera, but
are unreliable for any higher groups in this order. The superambulacral plates
are important when taken in connection with other characters; their absence is
often an important character in determining kinship. The gonads are chiefly of
use in determining relationship of species. Their arrangement is apparently
constant within a genus, but can not be used for families.

Not all the primary characters are of use for every family, and such characters
as the form of the tube feet and abactinal plates are common to several families.
They might better be called superfamily or subordinal features. Thus the order
may be divided into two groups, those with pointed tube feet (no definite sucking
disk being present) and those in which the tube feet have a definite sucking disk.
The first group nearly coincides with Perrier’s Paxillosa, and includes the families
Porcellanasteride, Goniopectinide, Astropectinide, Luidiide. The first family
is sharply marked off by the possession of only a single ampulla to each tube foot;
the other three have two. The first family also has a peculiar modification of the
marginal plates, in the form of cribriform organs,? and differs from all the others
except the Luidiide in lacking an intestine and intestinal cecum. The Luidiide
are separated from the Goniopectinide and Astropectinide by the structure of
the alimentary canal just referred to, and by the abortion of the superomarginal
plates. The Goniopectinide constitute an aberrant type near the Astropectinide,
from which they differ in having simple marginal cribriform organs as well as
webbed fascioles between the actinal intermediate plates. The family is not
large, and holds an intermediate position between the Porcellanasteride and Astro-
pectinide. But the segregation of the group from the Astropectinide is desirable,
as it leaves the latter much more homogeneous. The Astropectinide include all
genera with pointed tube feet, double ampullx, intestine, superambulacral plates,
and true paxille (or the readily derived parapaxille). The presence or absence
of an anus does not figure, as this character is worthless, for it varies within a
genus. Thus the family includes Plutonaster and its allies, classified by Sladen,
Perrier, Ludwig, Koehler, Gregory, MacBride, and their followers in the Archasteride,
and by Verrill in a distinct family (which can not be diagnosed), the Plutonasteride.

Before considering the second group, that characterized by having tube feet
with suckers, a word concerning the pointed tube feet is necessary. Several
authors have stated that the pointed tube feet are directly related to the habit
of living on a muddy bottom, and have in a way discredited the character for classi-

4 Present also in the Goniopectinide, although always of a simple type.

20 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

ficatory purposes. This may be true, but paxille, which leave arcade-like spaces
for the circulation of water about the papule, are just as directly related to the
same habit, as also are the various sorts of fasciolar channels between the marginal
and actinal intermediate plates, and the still more specialized cribriform organs
of the Porcellanasteridw. Unless climbing, a starfish does not move by attaching
‘ts suekers to the bottom and drawing itself along, as is sometimes stated, but
it rests on the tips of the tube feet,* which are manipulated like a centipede’s legs.
Consequently a starfish with suckers can walk or crawl on muddy bottom, and
those with pointed tube feet can crawl on hard bottom, as they do. For example,
Leptychaster, Persephonaster (Psilasteropsis), Astropecten, Ctenophoraster, Patagiaster,
Luidia, Ctenodiscus, and some other genera with pointed tube feet frequently
occur on hard sandy bottom, on gravel, on sand, coralline, and shells, or on mixed
coral and coral sand and shells. Similarly, members of the following families
with suckers on the tube feet live on soft mud, and, for aught we know, have for
aves. If the tube feet changed readily it is curious we do not find some of these
families having genera with pointed tube feet. Benthopectinide, Chetasteride,
Odontasteride, Archasteridx Goniasterid (all subfamilies), Oreasteride, Asteropide,
Linckiidw, and practically every family of the Spinulosa and Forcipulata (Crypto-
zonia). There are more species, with suckers, living on globigerina ooze than there
are with pointed tube feet (for comparison see Sladen, 1889, p. 713). I think it is
evident that the pointed tube feet constitute a conservative and deep-seated
character.

Since the pointed tube feet are of superfamily importance, it follows that
Sladen’s Archasteride, still used by most writers, is an unnatural group, as it
contains both types of structure. The family can not be diagnosed in a rational
manner. When the heterogeneous elements composing it are partitioned in other
groups where they belong, nothing is left but the genus Archaster, which is an
aberrant type of uncertain relationships. As Verrill has already indicated, the
family should be retained for this genus only.

The reasons for enlarging the Astropectinide are discussed under that family.

The families having well-developed sucking disks to the tube feet are as follows:
Benthopectinide (Pararchasterine of the Archasteride Sladen), Archasteride
(restricted), Odontasteride (part of Pentagonasteride Sladen, and Archasteride
Perrier), Chetasteride (part of Linckiide, most authors except Ludwig), Goni-
asteride (including Antheneide), Oreasteride, Asteropide, and Linckiide. The
Benthopectinide are readily separated on account of their more or less alternate
very spiny marginals, dorsal muscle bands, long slender rays, restricted papular
areas, rudimentary paxillx, and the characteristic form of the pedicellarie when
present. They possess a combination of characters not duplicated even in part in
any other family. Ludwig has recently raised this group to ordinal rank (Noto-
myota). The Odontasteride, Chetasteride, and Archasteride are small families.
The first and perhaps also the last are related to the Goniasteride. The Cheetas-
teridw are extremely puzzling. I have placed them in the key next to the Odon-
tasteride, but the group is difficult to classify. The glassy spinelets are duplicated
elsewhere only in the Spinulosa, but the structure of the abactinal tabulate plates,

* See Jennings: Behavior of Asterias forreri.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 21

the marginals, and the presence of superambulacral plates, favor the Phanerozonia.
The single ampull are noteworthy. They are double in the families mentioned
above. As throwing some light on the relationship of Chetaster with the Odontas-
teride, the odd interradial marginal is important, and the presence, in the recently
described Cheetaster vestitus Koehler, of incipient recurved specialized teeth to each
pair of mouth plates. These spines have an enlarged opaque base and a hyaline tip.*

With the Goniasteride a transition between the paxilliform, or tabulate abactinal
plate, and the flat granulate or smooth type takes place. It would be desirable,
if it were possible, to divide the Goniasteride into two families, one with tabulate
more or less paxilliform abactinals, and the other with flat pavement-like plates,
either with or without secondary intermediate connecting ossicles of various sizes.
The practical difficulties in the way of such a course are discussed under the Goni-
asteride. The diagnosis of the Goniasteride as it now stands is rather too general,
and perhaps vague. The reason for this is the persistence from past ages of many
intermediate intergrading forms. Naturally if there is anything to evolution, these
intergrading forms should be expected. (If they all persisted, it is evident that
there would be no definable orders, families, genera, or species.) The Oreasteride
are not very sharply separated from the Goniasteride, but after the removal of
Amphiaster the group is more homogeneous, and is readily recognized by the stellato-
reticulate abactinal skeleton, large papular areas, with numerous small papular
pores, and the finely granular skin which hides the underlying plates. The Goni-
asteride, on the other hand, have single papule, or a very limited number to an area,
and abactinal plates tessellate, or in a close network through the presence of small
intermediate plates. The transition between the typical tessellate form and the
more open structure such as occurs in Hippasteria and Anthenea is too complete
to keep these genera, which are not very closely related, in a separate family. They
are nearer Goniaster than are Pseudarchaster, Mimaster, or Gephyreaster. The
Asteropide are also difficult to diagnose satisfactorily, although the component
genera differ from the Goniasterid in having small imbricated marginals, and a
variable, loose abactinal skeleton overlaid by thick skin. The Linckiide (omitting
Chetaster) have come to be more or less traditional. The phanerozoniate charac-
ters are rather unstable, especially the papule. Some of its genera are near the
Goniasteride.

As already stated, Ludwig has raised the Benthopectinide, which he divides
into two-families, to ordinal rank. The character upon which greatest emphasis is
laid is the presence of a pair of muscle bands extending along the cwlomic surface
of the dorsal wall of the ray, from near the base to the tip. These are not present in
any other group. The alternate marginals are correlated with the muscles, as greater
flexibility is given to the ray. Ludwig believes the animals can swim by flex-
ion of the rays, and this idea is incorporated in the name of my new genus,
Nearchaster. The pectinate pediceilariz of several genera are also peculiar to the
group.

In this report I have kept the ordinal groups somewhat larger than an accept-
ance of the Notomyota would permit, although it is possible the Forcipulata are

4 Keehler, An account of the Shallow-water Asteroidea, Echinoderma of the Indian Museum, pt. 4,
June, 1910, pl. 19, fig. 11.

92 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

-

coérdinate with the Notomyota. However, the orders of Asteroidea are even less
satisfactory than the families, and they will ever furnish an excellent bone of con-
tention. If the Notomyota is accepted, the Phanerozonia will become a superorder.
This will necessitate the creation of a new order for the Porcellanasteride, and a
rearrangement of Perrier’s Paxillosa and Valvata. Perhaps some investigators will
find the following scheme acceptable:

Phanerozonia. Marginal plates prominent; papule confined to abactinal area
circumscribed by the marginal plates.

1. Order Cribellosa, new. Tube feet pointed, without sucking disk and with

single ampulle; cribriform organs; no intestine.
Family Porcellanasteride.

2. Order Paxillosa Perrier, emended. Tube feet pointed, with double am-
pulle; abactinal plates paxilliform; superambulacral plates; intestine present or
absent.

Family Goniopectinide.
Family Astropectinide.
Family Luidiude.

3. Order Notomyota Ludwig. Tube feet with sucking disks; a pair or dorsal
muscle bands present in rays; marginal plates alternate and spiny; pectinate
pedicellarie; abactinal skeleton, reduced paxille, or flat plates.

Family Benthopectinide.

4. Order Valvata Perrier. Tube feet with sucking disks, and without other
characters of Notomyota; abactinal plates varying from paxilliform to flat tessel-

late; pedicellarie bivalved, excavate, foraminate.
°

Family Archasteride.
Family Chetasteride.
Family Odontasteride.
Family Goniasteride.
Family Oreasteridx.
Family Linckiide.
Family Asteropide.

Family PORCELLANASTERID Sladen, 1889.

Phanerozonia with relatively thin and lamelliform (not block shaped) marginal
plates, which are naked or covered with rather thick membrane. Abactinal area
with simple perforated embryonic plates, some of which bear a spinelet, or with
paxilliform plates. A central conical prominence on abactinal surface. Actinal
interradial areas more or less extensive, paved with squamiform intermediate plates,
covered with delicate membrane, and sometimes bearing isolated spinelets or fim-
briated channels. Cribriform organs present. Adambulacral plates large, with a
simple marginal armature, uniserially disposed. Superambulacral plates present in

one genus; no anus; no intestine; no intestinal ceca. Tube feet pointed; ampullze
single.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 23

a KEY TO THE KNOWN GENERA OF PORCELLANASTERID.

a'. Cribriform organs highly developed, localized; actinal interradial areas without fimbriated trans-
verse channels; no superambulacral plates.
b!. Both inferomarginals and superomarginals present.

c', Actinal intermediate plates usually but not always naked; adambulacral plates with a single
series of spines on furrow margin only; terminal plates large, armed with conspicuous spines;
cribriform organs one to eleven.

d'. With a conical apical elongation; abactinal membrane with simple spiniferous spicules;
actinal intermediate plates not imbricated; ambulacral furrows wide and exposed; rays
more or less turned back; cribriform organs one to three.

e!. Component structure of cribriform organs lamelliform.
ife~ One er briform organ eect ocean te Porcellanaster 4 W. Thomson.
f2> Mhree cnibriform orpangs sae nas a aisle = wim ain om alain Eremicaster Fisher, p. 23.
e?. Component structure of cribriform organs papilliform; one cribriform organ.
Sidonaster Keebler.

@. No conical apical elongation of abactinal integument, but a slight peak may be present;
abactinal membrane with pseudopaxille; actinal intermediate plates imbricated and
arranged in columns; ambulacral furrows narrow and concealed; rays not revertible;
cribriform organs five to eleven (three in one case only); component structure papilliform.

e!, Superomarginal plates with robust spines forming a single series along the median line;
marginal plates united along median line of the rays.
jf?. Spines of adambulacral armature long and needle shaped, radiating apart; cribriform
organs three to seven in each interradius; no ventrolateral plates in ray; actinal areas
mot. pranvlated =... jo0-% acc fesse deesies sn dacaee ech bee bse ees sees Styracaster Sladen.
Jf?. Adambulacral spinelets spatulate, united by membrane into a scoop-shaped group;
cribriform organs eleven; a single series of ventrolateral plates to tip of ray; actinal
areas pranulated sc. .cccsscs esac sas Ae aon ae ease iecos cea vew scissile Chunaster Ludwig.
e?. No spines on superomarginal plates; spines of adambulacral armature short, compressed,
forming independent series or fans.......-------- - Hyphalaster Sladen.
c?. Actinal intermediate areas covered with spiniform granules, “Adambulacral plates with papilli-
form spinelets on the outer portion of plates; terminal plate very small and inconspicuous,
MNATMed er DritOrMm OrP ANS fOULLECN en ee = = acon aa eileen ecole a Thoracaster > Sladen.
b?. Inferomarginal plates aborted; one cribriform organ..................-...-Albatrossaster Ludwig.
a’. Actinal fasicioles or fimbriated channels present; cribriform organs of a simple nature and between
all the marginals; small superambulacral plates (in Ctenodiscus).
b!. An unpaired superomarginal and inferomarginal in each interbrachial arc and a corresponding
unpaired double series of actinal intermediate plates..................- Pectinidiscus Ludwig.
b?. No odd interradial marginal; no unpaired double series of actinal intermediate plates corresponding
RH EROLO ne cree sa ctee cyan eet cee an cer eae ce eer cele oe Ctenodiscus Miiller and Troschel, p. 31.

Subfamily PORCHLLANASTERIN<Z®! Sladen, 1883.

Porcellanasteride with cribriform organs highly developed, localized. Actinal
interradial areas with squamiform plates covered with a simple membrane, and not
traversed by fimbriated channels; no superambulacral plates.

Genus EREMICASTER (Fisher.)

Eremicaster (subgenus of Porcellanaster) Fisuer, Bull. Bur. Fisheries for 1904, vol. 24, June 10,
1905, p. 293. Type, P. tenebrarius Fisher.

Eremicaster Lupwie, Zool. Anz., vol. 31, 1907, p. 317. Type, P. crassus Sladen.—FisueEr, Zool.
Anz., vol. 32, 1907, p. 14.

@ Includes Caulaster Perrier; see Ludwig, Zool. Anz., vol. 31, 1907, p. 318, footnote.
» Includes Pseudaster Perrier; see Ludwig, Zool. Anz., vol. 31, 1907, p. 313, footnote.

24 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Diagnosis.—Like Porcellanaster Wyville Thomson, but with three (instead
of one) cribriform organs in each interradius.

Marginal plates thin; superomarginals not united along the median line of
ray and usually bearing a spine, these forming a series on each side of ray. Abac-
tinal area covered with membrane containing very small, simple, perforated plates;
those on disk bearing usually a single spinelet each; in center of disk a sort of pro-
longation of the integument forms a peduncle, containing an extension of the body
cavity, but no intestine. Actinal interradial areas paved with thin plates, some-
times armed with spinelets. Adambulacral armature, one to three sharp spinelets.
Usually ‘segmental pits and papille”’ are present. No superambulacral ossicles;
tube feet conical, pointed. Ampulle single.

Remarks.—When this group was described as a subgenus in 1905, I gave three
characters, in order of importance, by which it could be distinguished from Por-
cellanaster; namely, three cribriform organs, presence of segmental pits and papille,
and one or two adambulacral spinelets. Since then I have examined much more
material, and have found that the last character is of no importance generically,
while the segmental pits and papille are numerically variable. The absolutely
constant character at all ages is the three cribriform organs. The segmental pits
and papille are valuable when taken in connection with the three oributonny organs,
as they show relationship. The second (aboral) spinelet of the adambulacral
plates of Porcellanaster is homologous with the segmental papilla of Eremicaster;
but the outer part of the ray in Hremicaster lacks the segmental papilla, and is
therefore similar to Porcellanaster, except in those specimens in which three spinelets
are present in addition to the segmental papilla or its homologue.

Ludwig has considered the segmental papille and pits as of primary importance.*
He placed his P. pacificus which has three cribriform organs and no pits or papillze
in the restricted genus Porcellanaster. I have studied pacificus and consider it
much more nearly related to tenebrarius than to P. cxruleus. Owing to the slight
instability of the pits and papillae in tenebrarius (includes waltharti Ludwig) it
seems better to rely upon the cribriform organs as mentioned above. Therefore
I have placed pacificus in Eremicaster, although Ludwig relegates it to Porcellanaster.

In raising my subgenus Eremicaster to a genus Ludwie changed the published
type tenebrarius to crassus Sladen. This is contrary to a very elementary rule
in nomenclature. Eremicaster will stand or fall with tenebrarius as type. Asa
matter of fact the two species are so close that it makes no material difference

which is type, but the habit of ignoring well-established rules of nomenclature
is a bad one.
EREMICASTER TENEBRARIUS (Fisher).

Pl. 1, figs. 1-4; pl. 2, fig. 4; pl. 53, figs. 4, 4a.

Porcellanaster (Eremicaster) tenebrarius Fisner, Bull. Bur. Fisheries for 1904, vol. 24, June 10,
1905, p. 293.—Lupwia, Zool. Anz., vol. 31, 1907, p. 317.

Porcellanaster waltharii Lupwia, Mem. Mus. Comp: Zo6l., vol. 32, July 17, 1905, p. 92, pl. 5,
figs. 26, 27. 3;

4 Zool. Anz., vol. 31, 1907, p. 3l7.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 25

Eremicaster tenebrarius Lupwie, Zool. Anz., vol. 31, 1907, p. 318.—Fisner, Zool. Anz., vol. 32,
1907, p. 14.
Eremicaster waltharti Lupwie, Zool. Anz., vol. 31, 1907, p. 318.

Diagnosis.—Rays five. R=38 mm.; r=11 mm.; R=3.45 r. Another speci-
men: R=42 mm.,r=9mm., R=4.66r. Breadth of ray at interradial line, 14 to 15
mm.; at outer edge of lateral cribriform organ, 9 mm.; height of ‘‘epiproctal cone,”
3 to5 mm. in different specimens. Related to F. crassus (Sladen). Rays elongated,
stout fairly slender, abruptly tapering at base, but only very gradually on outer
three-fourths; ambulacral furrows very wide. Superomarginals typically with
one or two spinules on upper edge; inferomarginals unarmed, as many as nineteen
in number; terminal plate with three to six spinelets; adambulacral plates with
one or two spinelets (usually two in large examples), and one segmental papilla
which is transformed into a spinelet distally on ray; mouth plates with two seg-
mental papille, with one unpaired spine at inner angle and with one smaller spinelet
at either side. Actinal interradial areas smooth; intermediate plates extending
to fourth or fifth inferomarginal. Spinelets on abactinal membrane extending
as far on ray as lateral cribriform organ; spinelets membrane-covered, terete, and
spaced; abactinal area on ray very narrow.

Description —The abactinal area is slightly sunken below level of the upper
edge of superomarginal plates, and is narrow on rays; but it varies in width according
as the marginal plates are arched snward. On outer two-thirds of ray it is often
reduced to a mere line and is never wider than a neighboring superomarginal.
Abactinal integument is thin but resistant, and on disk and basal portion of ray
(as far as outer cribriform organ) is covered with simple, slender, spaced spinelets,
which are sheathed in membrane so that they appear terete and truncate. They
are slightly longer toward edge of disk. Beyond the outer cribriform organ no
‘spinelets are present in the abactinal integument. On the central prolongation
of the integument, or ‘‘epiproctal’”’ cone, which is from 3 to 5 mm., the spinelets
are minute, and when seen in connection with their simple, perforated, ‘‘embry-
onic’’ plates greatly resemble ‘‘tables,” such as are present in the skin of many
holothurians, the spinelet representing the spire. The abactinal plates are scattered
on the ray, are closer on the disk, and crowded on the central prolongation or
peduncle. They are, in form, small circular perforated plates, with a smooth or
irregular contour, and do not differ materially from the perforated plates of many
-holothurians. Scattered among the spinelets (but not elsewhere) are numerous
papul, especially in the interradial areas. They appear to be absent from the
epiproctal cone.

The marginal plates always arch inward more or less so that when viewed
from above the contour of body is usually defined by inferomarginals. Supero-
marginals (eighteen or nineteen in large specimens) are nearly quadrate and are
larger than corresponding inferomarginals. They decrease regularly in size distally.
Each plate bears, on its upper edge, one or two robust, tapering, sharp spinules,
these forming a row to terminal plate. These spinelets are sometimes nearly
as long as height of plate. A single spine stands at the top of the suture between
the median and either lateral or cribriform organ. (For variations in presence
of superomarginal spines, see Variations on p. 27.)

26 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Terminal plate, seen from side, is about as long as last two superomarginals,
and is strongly concave on side toward abactinal membrane. There are five, some-
times six, prominent spinules, subtubercular and conical; of these, one (rarely two)
is on the extreme tip, one at either side of the end of ambulacral furrow just below
the terminal spinule, and one on either side, abactinal in position. In Alaskan
specimens there are only three spinules on terminal plate, those mentioned last
being absent.

The inferomarginal plates are longer than high, and correspond to supero-
marginals in number; or there may be one more; they are not always exactly
opposite plates of upper series. They do not extend onto actinal surface but are
confined to side wall of ray. Inferomarginals lack spines or any special armature
except the three cribriform organs shared with the superomarginals.

There are always three cribriform organs to each interbrachial arc. Of these
the median is slightly the largest, having twenty to twenty-four lamelle. These
lamelle are made up of a linear, perpendicular series of flat spinelets or oblong
platelets in a thin web-like investment. More or less of the spinelets are free
around the edges, especially the actinal. Each organ has a slight depression down
the center.

Adambulacral plates are narrow and rather long, conspicuously excavated
on furrow margin. Adoral end is most prominent and is surmounted by one
or two sharp tapering spinelets, one slightly compressed and curved, being directed
into the furrow; the other, usually a trifle smaller, is either reflected back over
the plate or is turned aborad. The latter is frequently absent. Midway between
the extremities of plate is a rather small valve-like, scoop-shaped scale, resembling
one-half of a valvate pedicellaria—the ‘‘segmental papilla.’’ It is articulated
to plate, on edge of furrow by a straight base, and the free edge is broad and rounded.
It closes over a small slit in the plate, which is absent beyond basal fourth of ray.
The valve beyond the proximal third to two-thirds becomes transformed into a
small lanceolate spinelet, so that there are either two or three spinelets to each
plate beyond this point. Alaskan specimens, which are smaller than those from
California, usually have only one regular spinelet, and the segmental papille extend
rather farther along ray.

Mouth plates very prominent, the united pair forming a broad keel, with a wide
lanceolate median suture, the companion plates touching at the inner and outer
ends. The inner extremity of the combined pair is rounded and bears a single
short, conical spinule (frequently absent from being rubbed off); and on margin,
nearer peristome, on either side, is a curved compressed, smaller spinelet. Between
latter and first adambulacral plate are two semicircular papille.

Actinal interradial areas small and triangular, without spinelets, and covered
by a thin skin through which the plates are not discernible until specimen is partially
or wholly dried. The plates are thin subcircular or elliptical, and not regularly
arranged. In small specimens they are distinctly spaced. They extend as far
as the edge of the fifth inferomarginal. Beyond this point there is a single row of
what appear to be isolated intermediate plates, extending to tip of ray. These are
the lower ends of the ambulacral ossicles, are triangular, the apex being adjacent
to a suture between two adambulacral plates. At the tip of ray they are fairly

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. oT

conspicuous and three underlie the terminal plate. All these ‘‘plates”’ are invisible
until specimen is dry. Between the flaring outer ends of companion mouth plates
is a triangular unpaired plate, the exposed portion of the primary oral plate or
odontophore.

Madreporic body large, adjacent to a median cribriform organ. It is about as
wide as the latter. Striations deep and radiating.

Color in life: general tint whitish; bluish cast on disk due to viscera showing
through integument.

Anatomical notes.—Superambulacral plates absent. Peristome contains small
irregular perforated rods and plates especially on the lips. There are also a few
simple straight and curved rods. Tube feet large, conical, without calcareous
particles in walls. Ampulle single, not double. The stomach is very simple,
possessing very short radial pockets. The dorsal wall is closely bound to the
abactinal body wall. There is no intestine and no anus, nor any sign of an intes-
tinal cecum. A prolongation of the ceelom fills the so-called ‘‘epiproctal”’ cone or
dorsal peduncle, but the opening into the peduncle is closed by a layer of muscles.
Gonads small, one on each side of the interradial septum.

Variations.—The specimens from California, six in number, present very few
variations of importance. There is some difference in the width of the abactinal
area on rays as explained above. Superomarginals range from fifteen to eighteen
and there may be one to three superomarginal spines. There is sometimes but one
adambulacral spinelet although there are usually two. The segmental papilla can
be recognized as such on as many as fifteen adambulacral plates, and varies all the
way to only five. It gradually is transformed into a lanceolate spinelet so that it
is not easy to limit exactly the number.

Numerous smaller specimens from Albatross station 2859 vary more widely.
R averages about 22 mm. and r8 mm. (R=2.75 r). But one specimen has R = 22
mm., r=6 mm., R=3.6 r, which is fully as long a ray as in the type. Superomar-
ginal plates twelve to seventeen. Out of forty-five specimens from this station
thirty-three lack entirely the superomarginal spinules, five have a spinelet here and
there along the superomarginal series, and seven have a regular series of supero-
marginal spinules. These are lacking usually from the first two plates, and may be
absent from an odd plate here and there. The marginal plates are a trifle smaller
and weaker than in Californian examples. There is usually but one true furrow
spinelet, which is curved and compressed. Rarely two are present, the second
being placed directly behind the first. Segmental papille can be recognized as such
for about one-half to three-fourths the length of furrow although the rudimentary
pit is present only on the first four or five plates. The terminal plate has three
instead of five spinelets.

The most important difference between these specimens and the Californian
examples is the absence of superomarginal spinules in many of the former. Their
presence in almost typical form in seven specimens shows however that supero-
marginal spinules are not wholly trustworthy as a specific character. The difference
in the number of adambulacral spinelets is variable even in Californian examples.
The difference in size may account for the more frequent presence of two spinelets
in largest examples. The so-called ‘‘segmental papille’’ of the adambulacral

28 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

“-

plates are decidedly variable in the number to a ray and in degree of development.
Size may also account for the presence of only three terminal plate spinules in
Alaskan examples.

Although these Alaskan specimens present slight differences the totality of
characters seems to warrant placing them under tenebrarius. The numerous
specimens from the single station show greater variation among themselves than
do the typical Alaskan and Californian examples, when compared.

Young.—The smallest specimen, from station 2859, has R=10mm., r=5 mm.,
and eight superomarginal plates, one adambulacral spinelet, segmental papille,
as such, only on two or three plates, after which they stand over furrow as broadly
lanceolate flat spinelets; no superomarginal spinules. The cribriform organs are
smaller than in the adult, with about ten lamelle which tend to split up into indi-
vidual spinelets, and the space between the cribriform organs is wider than in adult,
about as in Sladen’s figure of Z. gracilis.¢ Abactinal spinelets are confined to the
interradial areas, and in fact are scarce on the radial areas of the disk on many
considerably larger specimens.

Type.—Cat. No. 22326, U.S.N.M.

Type-locality.— Albatross station 4397, about 200 miles off San Diego, California,
2,196 to 2,228 fathoms, on gray mud.

Distribution.—Southern Alaska to the Galapagos Islands, 1,569 to 2,259
fathoms.

Specimens examined.—Fifty-one specimens from the following stations:

2859, off Prince of Wales Island, southern Alaska, 1,569 fathoms, gray ooze,
bottom temperature, 34°.9 F., forty-five specimens.

4394, off San Diego, California, 2,259 fathoms, soft gray mud, three specimens.

4397, about 200 miles off San Diego, California, 2,196 to 2,228 fathoms, gray
mud, three specimens. Last two stations, Albatross cruise of 1904.

Remarks.—This species is closely related to Eremicaster crassus (Sladen),
which was taken by the Challenger Expedition in 2,335 fathoms in the South Pacific,
midway between Sydney and Valparaiso. From this form tenebrarius differs in
having slenderer and (by comparison with figures) longer rays, and about twice as
many marginal plates. Hremicaster gracilis (Sladen) from off the western coast
of South America, between Valparaiso and Juan Fernandez, is probably only a
very young specimen of crassus because the differences cited are such as would be
expected in a young example. I have already pointed out how tenebrarius varies
at one station in respect to the superomarginal spinules. It would have not been
difficult to have created three species from this collection of forty-five specimens.
Mention has already been made that in the young of tenebrarius abactinal spinelets
are developed first on the interradial areas of disk. This would be equivalent to
Sladen’s phrase ‘‘abactinal membrane with spinelets confined to limited areas”
and is borne out by his figure 1, on plate 20. The fact is, the species of this genus
are more variable than commonly supposed.

After a careful comparison of specimens of E. tenebrarius with Ludwig’s de-
scription and figures of E. waltharii from off Panama and the Galapagos Islands,

@ Challenger Asteroidea, pl. 22, fig. 3.

e

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 29

T am unable to find a single constant difference. The photographic figures of
waltharii are just like tenebrarius. There are far greater differences between the
extremes of Alaskan specimens than between typical tenebrarius and waltharii.
Consequently I believe the latter to be a synonym of the former.

EREMICASTER PACIFICUS (Ludwig).
Pl. 2, figs. 1-3.

Porcellanaster pacificus Lupwie, Mem. Mus. Comp. Zodl., vol. 32, 1905, p. 89, pl. 6, figs. 28, 29;
Zool. Anz., vol. 31, 1907, pp. 317, 318.
Bremicaster pacificus FisHer, Zool. Anz., vol. 32, 1907, p. 14.

Diagnosis.—Rays five. R=21 mm.; r=10 mm.; R=2.1 r. Breadth of ray
at interradial line, 11 mm.; at outer end of lateral cribriform organ, 6.5 mm.; length
of abactinal peduncle, 5 to 7 mm. Disk rather large and rays of median length;
interbrachial arcs well rounded. Cribriform organs three; no segmental pits or
papille; adambulacral plates with two sharp spinelets; marginal plates eight or
nine, the superomarginals with one, rarely two, upright spinules; abactinal surface
with spaced spinelets extending to outer cribriform organ; actinal intermediate
areas with spaced spinelets.

Description.—Abactinal membrane capable of inflation. Disk and basal part
of ray, as far as outer edge of lateral cribriform organ covered with spaced deli-
cate, terete spinelets, each invested in a thin membrane. In one specimen the spine-
lets extend a short distance beyond cribriform organ. They are about 0.5 mm.
high and spring from simple subcircular perforated plates embedded in the abactinal
membrane. As a rule these plates have an irregular outline and are about one-half
to two-thirds as broad as height of spinelet. The plates are more or less spaced
and irregularly between them emerge the papule, singly. They have the appear-
ance of being about as numerous as the spmmelets. The central prolongation of the
integument or peduncle contains crowded plates with rudimentary spinelets as in
tenebrarius, and no papule. In the integument of ray beyond the limit of spinelets
are simple subcircular perforated plates usually uneven or more or less toothed on
margin; occasionally smooth. The plates are somewhat larger than those of disk,
the largest being 0.6 mm. in diameter, a small one about 0.3 mm.; whereas on the
spiniferous area at base of ray the largest are 0.35 to 0.4 mm., and are thicker in the
center. The abactinal area varies in width on ray, even in same specimen, according
as marginal plates are arched inward. Occasionally two or three isolated spinelets
occur on outer part of ray.

Marginal plates in form about as in &. tenebrarius. Superomarginals eight or
nine, each with an upright glassy tipped spinule on upper edge of plate, these about
as high as plate when perfect and decreasing in size toward end of ray. Occasion-
ally a second shorter spinule is present. Inferomarginals the same in number as
superomarginals, confined to side wall of ray and longer than high.

Terminal plate fairly large, deeply concave toward abactinal area. There is
one terminal spinule, and below at either side of end of furrow, one; behind termi-
nal spinule on abactinal edge of plate a row of two to five spinelets on either side.
In a small specimen from station 2859 these are lacking.

°
30 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Cribriform organs large, always three. The median has about twenty-four
lamelle, the lateral about twenty. Their structure does not appear to be different
from those of tenebrarius, although there is more tendency for the plate-like spinelets
to remain independent.

Adambulacral plates shaped as in tenebrarius and with two acicular skin-
covered spinelets on the margin. Of these one stands on the prominent adoral
end of plate, the other a trifle shorter at about the middle of concave margin or a
little nearer adoral end. The longest spinelet is about as long as plate; the other
is the homologue of the segmental papilla of Z. crassus and E. tenebrarwus.

Mouth plates very prominent actinally, a little wider at outer end than in tene-
brarius. Armature consists of an odd acicular spinule at inner end of combined
pair and along the free margin a series of two to five similar spinelets increasing
slightly in size toward first adambulacral.

Actinal interradial areas about as in tenebrarius, but covered with spaced
skin-covered spinelets very similar in character and size to those of abactinal sur-
face. These spinelets vary in length and are more numerous near the marginal
plates than in the inner angle. Intermediate plates extend to fourth inferomarginal.
At outer end of combined mouth plates the triangular end of the odontophore is
visible. The triangtJar lower end of ambulacral ossicles is visible, as in tenebrarius,
between the inferomarginal and adambulacral plates, forming a series extending
to tip of ray.

Madreporic body large, about as wide as median cribriform organ to which it
is adjacent.

Anatomical notes.—No superambulacral plates. Gonads interradial. Stomach
simple with two short pouches protruding into base of arm. No intestine; no
anus. There are no deposits in peristome, in the specimen examined.

Type.—Will be deposited in U. S. National Museum.

Type-locality— Exact locality not stated; description based on specimens
from seven stations between lat. 14° 46’ N. and lat. 0° 36’ S., and long. 98° 40’ W.;
region of the Gulf of Panama.

Distribution.—Bering Sea (south of Pribilof Islands) to Gulf of Panama and
vicinity of Galapagos Islands, 859 to 1,879 fathoms, soft mud and ooze and fine sand.

Specimens exramined.—Six from the following stations:

2859, off Prince of Wales Island, Alaska, 1,569 fathoms, gray ooze, two speci-
mens.

3075, off Washington, 859 fathoms, green mud, one specimen.

3603, Bering Sea, between Pribilof and Umnak Islands, 1,771 fathoms, brown
ooze, three specimens.

Remarks.—The specimens agree well with Ludwig’s description and figures
of his Porcellanaster pacificus, which thus, like tenebrarius, has a wide distribution.
There are a few minor points of difference, as in the spinulation of the terminal plate,
but this is subject to variation. The nearly related vicinus is insecurely separated
from pacificus. Among the differences mentioned by Ludwig the only one of any
importance is the absence of actinal interradial spinelets.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. ol

As mentioned under the generic description it seems better to place this
species under Eremicaster than with Porcellanaster, the segmental papille being
subject to individual variation, while so far as known there is no variation in the
number of cribriform organs.

Subfamily CTHNODISCIN 4 Sladen, 1889.

“Porcellanasteride with a simplified form of cribriform organ on the margins
of each pair of marginal plates. Actinal interradial areas traversed by fimbriated
channels, in continuation of the fasciolar or cribriform channels between the marginal
plates” (Sladen). Superambulacral plates present, though not well developed.

Genus CTENODISCUS Miller and Troschel.

Ctenodiscus Miitter and Troscuet, System der Asteriden, 1842, p. 76. Type, Asterias polaris
Sabine=A. crispata Retzius.
Anodiscus Perrier, Ann. sci. nat., sér. 5, vol. 12, 1869, p. 298.

Diagnosis.—General form stellato-pentagonal to stellate; abactinal surface
with true paxille; marginal plates thicker than in typical Porcellanasteride; mar-
ginal, actinal intermediate, and adambulacral plates obscured by a continuous
skin; marginals bearing each a short spine at upper end of plate, these forming two
longitudinal series along ray; cribriform organs in the form of fasciolar grooves
between marginal plates, these grooves continued as simple channels (roofed by
webbed spinelets) across actinal intermediate area and between adambulacral
plates; adambulacral plates with three to five furrow spinelets. No anus. Super-
ambulacral plates present.

CTENODISCUS CRISPATUS (Retzius).

Pl. 3, figs. 1-4; pl. 4, figs. 1-6.

Asterias crispata Retztus, Dissert. sistens species cognitas Asteriarum, 1805, p. 17.

Ctenodiscus crispatus Dipen and Koren, K. Vetensk. Akad. Handl., 1844 (1846), p. 253.

Ctenodiscus australis (Lovén MS.) Lirken, Vid. Medd., 1871, p. 238.

Ctenodiscus krausei Lupwia, Echinodermen des Beringsmeeres, Zool. Jahrb., Syst., vol. 1, 1886,
. 293.

Gecuea procurator SLADEN, Ohallenger Asteroidea, 1889, p. 173, pl. 30, figs. 7-12. For numer-
ous citations to 1899 see Ludwig, Fauna Arctica (Rémer and Schaudinn, Jena) 1900, p. 450;
also, Mem. Mus. Comp. Zodl. vol. 32, 1905, p. 104, pl. 6, figs. 32, 33, for record in equatorial
east Pacific.

Diagnosis.—Rays five, exceptionally four or six. R=1.66rto3.16r. General
form stellato-pentagonal, to stellate, extremely variable. Abactinal surface more or
less tumid and usually with an elevated cone in center of disk; paxille variable,
usually low, with few to many spinelets which are short, clavate, and skin-covered.
Marginal, actinal intermediate, and adambulacral plates obscured by a thin soft
skin. Continuous narrow deep grooves extend between marginal plates, across
intermediate area, and between consecutive adambulacral plates. These are
overhung on either side by a fold of skin embedded in which are numerous
flattened spinelets, as in the lamelle of typical cribriform organs. Between special
raised ridges of marginal plates these furrows are deeper and V-shaped with five to

82 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

seven superimposed lamelle on either side. Marginal plates eleven to twenty, in
each series, from median interradial line to extremity of ray. A single short conical
spine at upper end of each superomarginal, and another similar one on each infero-
marginal on the actinolateral margin of ray. Adambulacral plates with an oblique
series of three to five sharp short skin-covered spines, and on aboral outer corner a
similar, usually shorter spine, covered with the general investment of actinal surface.
Mouth plates prominent; along free margin, about six spines like those of adambu-
lacral and at inner end of plate a single more prominent spine; two or three short
conical tubercles usually stand in a series on either side of median suture, these
sometimes as long as furrow spines. Superambulacral plates present.

Description.—No adequate description of this species is readily available. With
the ample material at my disposal it seems well to give a description and at the
same time to point out some of the most prominent of the variations. Instead of
placing these last, a separate subhead is given to each category of characters and the
variations considered at once.

This is a remarkably variable species, especially in the length of the rays, and
their width, and in the general facies of the animal. Practically all of the most
diverse variations may occur in examples from a single station, and they are thus not
due to locality. As slender and broad armed forms occur among the smaller speci-
mens this difference is not due to age.

Proportions.—A striking series of four nearly equal sized specimens from station
4235 illustrates admirably the difference in form, measurements being given in the
accompanying table.

Measurements of Ctenodiscus crispatus.

< |
N umber | Wiath | Number | Width of
Station. R. Tavera omat- of ray fs ae ee Remarks.
= = = ——— : =
M85 So scccss 38 | 12.0 3.16:1 20 14 5 4.0 Actinal area very narrow; paxillee medium
sized.
ORR oA cece ee 39 | 16.0 2.45:1 18 20 3-5 3.5 Paxille medium sized.
Shia ee ee 36 | 14.0 | 2.57:1 | 16-18 18 344 4.0 Do.
MBG te ae 40 17.0 2.35:1 | 17-18 21 4-5 4.0 Paxille smaller.
42231 35 | 19.0 HS:05 | 13 23 3-5 3.0 Paxillee medium sized; average proportion.
2238 33 | 18.5 | 1.84:1 | 1h he eee 45 2.5 Paxille larger (P1. 4, fig. 5).
MORE ee tes cand 29 | 17.5 | 1.66:1 12 | 20 3-4 2.75 | Paxille small.

2 From interradial line to tip of ray.
© Measured from one interradial line, across abactinal surface to adjacent interradial line.
¢Spines on furrow edge of adambulacral plate; one or more spines on surface of plate not counted.
Specimen 4235' has slender, narrow rays (from above resembling a Psilaster
somewhat) and numerous marginal plates with fairly wide fasciolar furrows. The
intergradation is perfect, through 4235‘ to 4286, a very short-rayed form almost
arcuately pentagonal, with few superomarginal plates. If, in the above table, the
width of ray at base is compared with R, and the porportion R:r taken into account,
the great difference in proportion is at once evident.
Abactinal surface; paxille.—In some specimens the epiproctal cone is inverted,
in others very inconspicuous. In specimens with rays nearly the same length the
width of the paxillar area varies considerably, especially at the end of ray, thereby

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 33

giving some specimens a much more robust appearance. The fact that in some
examples the abactinal wall is nearly plane, except for the epiproctal cone, and much
inflated in others is due, of course, to the condition of the animal at the moment of
death, but is important as magnifying one or two trivial characters, such as the angle
of marginal plates and compactness of paxille. Two extremes, both from station
4223, will serve to illustrate variability of paxille. Eachspecimen has R=34 mm.;
in A, paxille in neighborhood of madieporic body have seven to twelve spinelets, occa-
sionally as few as four on very small ones; of these never more than one is situated
in center of tabulum and rather more than half the paxille have no central spinelet
at all; in B the paxille ordinarily have from twelve to twenty-two spinelets, of which
three to five occupy center of tabulum, and all or very nearly all have central spine-
lets, those paxille on outer part of ray having so few as one central spinelet.* When
these two specimens are placed side by side the difference is very striking. Many
of the specimens, both small and large, lack a central spinelet to paxille altogether,
thus resembling exactly some Atlantic examples. There is as much if not more
difference in the extremes of these specimens than is shown by Sladen’s two figures
illustrating paxille of Ctenodiscus australis and C. procurator,® while in the extremes
of body-form the difference is greater than between figures 1 and 7, illustrating the
same two species. I thought at first that the difference in size of madreporic body
might furnish a character of some constancy, to separate krausei from crispatus, the
latter having the larger body. This character also is very variable in Pacific speci-
mens, some examples having fairly large, others small madreporic bodies. In four
rather poorly preserved Atlantic specimens the madreporic body is more constant
and is one and one-half times greater in diameter than in Pacific examples of the
same size.

The shaft of the paxilla varies in length. In specimens from very deep water
it is longer than in shallow water specimens.

The bases of paxille, or the abactinal ‘‘ plates,” are circular and rather closely
placed, usually not quite touching. They are largest about one-half r from center
of disk, decreasing in size toward tip of ray and center of disk. Along mid-radial
line where there are no papule the plates are smaller and usually broadly elliptical.
Papule are not regularly arranged, four to six usually occurring about a plate.
They are single, and are lacking on a circular area in center of disk (including the
central cone) and on five narrow radial and five narrow interradial areas extending
from the center like the spokes of a wheel. On either side of the radial areas,
papule extend to tip of ray.

Marginal plates.—Superomarginal plates exactly opposite inferomarginals. The
former are thin and confined to side wall of ray. The actinolateral border of ray
with its series of spines is slightly nearer upper than lower border of inferomarginals.
The part of inferomarginal below the actinolateral spine has a rather broad special-
ized ridge, which is broader than intervening fasciolar furrow. This furrow is
roofed over by a single row of spinelets immersed in a continuous web, eight or
nine of these spinelets occurring between lower edge of plate and actinolateral

,

2 Atlantic specimens show the same range of variation.
5 Challenger Asteroidea, pl. 30, figs. 4 and 9.
57444°—Bull. 76—11——3

34 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
spine. When dried the surface of plate shows often numerous minute bosses.
Directly above the actinolateral spine the specialized ridge narrows abruptly and
joins without a conspicuous break, that of superomarginal. The fasciolar channel
is several times broader than the specialized ridge and is a V-shaped trough, on
whose sides are five to seven parallel superimposed but distinctly spaced series of
delicate spinelets in a continuous web. Hach web extends from the upper end of
a superomarginal plate to the actinolateral spine. The whole apparatus of five
to seven webs or lamellie on either side of the V-shaped trough, extending over its
cavity, forms a very delicate cribriform organ or filter. The series of spinelets roof-
ing the furrow are much flattened and the tip ends in several minute points. Owing
to the greater width of furrow, these spinelets are much longer than those of lower
end of inferomarginals. The other spinelets are narrow and more delicate and
decrease in size in each successive tier toward bottom of furrow. In alcoholic
specimens the membrane investing the spinelets is usually so thick that they are
only seen with difficulty. The spinelets act only as a support for the membrane
and do not themselves function as strainers, as in Astropecten. 'The superomarginal
spine which stands at the top of the plate is either terete, tapering, and pointed,
or broadly lanceolate and acute like a spear tip. On outer part of ray there are
sometimes two spines on a few plates.

The variation in number of marginal plates has already been indicated in the
table. The margin of ray and disk is thicker in some examples than in others, the
appearance being heightened by recurved rays, when the actinolateral angle is
rounder and less abrupt. The exposed surface of the specialized ridge of supero-
marginal plates is thin or narrow, but varies more or less, being slightly broader in
four Atlantic specimens; but in another Atlantic example from off Newfoundland
they are as narrow as in Alaskan specimens. The height of superomarginals is
variable, specimens from 1,033 fathoms having much lower ones. This is readily
appreciated by noting the distance between the two series of marginal spines in
the interbrachial arc.

Actinal surface—The adambulacral armature is essentially alike in both
Atlantic and Pacific specimens, some examples showing a preponderance of three
or four furrow spines, others of four and five. Besides the spine (not of furrow
series) which usually stands on the outer aboral corner, there is usually one to
several very much smaller and more delicate spinelets along adoral and outer edge
of plate. These can not be distinguished readily unless specimen is dry.

Actinal interradial areas, like marginal and adambulacral plates, are overlaid
by membrane through which the plates are scarcely visible until dried. Plates
are arranged in series running from marginals to adambulacrals. Deep channels
also follow same course. These are overhung by a series of spinelets embedded in
membrane, being a continuation of the marginal fascioles. The interradial channel
splits and runs on either side of mouth plates. The first on ray always runs between
second and third adambulacral, the second between the third and fourth, and so
on. The photographic figure will show the arrangement of plates.

Anatomical notes.—No intestine, no intestinal ccecum, no anus. The conical
eminence in center of disk contains a prolongation of the ceelom and is divided by a
vertical septum. Stomach large, single, firmly moored to abactinal wall. Hepatic

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 35

ceeca large, extending nearly to end of ray, and with spacious cavity. Gonads
interradial, one on either side of the membranous interradial septum. Ampulle
single; tube feet large, conical, pointed; no deposits in walls; one Polian vesicle
to each interradius.

Superambulacral plates present; absent from first and sometimes second
ambulacral ossicle and also from last six or eight at end of ray. These plates are
not very conspicuous, being dorsoventrally flattened and overlaid by membrane.
Although Ludwig @ states that they are absent from his specimens (Panama region),
I think he must have overlooked them. They are very easily seen if a portion of
the ray is treated with caustic potash solution. They are present in a specimen
examined, from station 3307, from the great depth of 1,033 fathoms. I also dis-
sected a specimen from station 2452, off Newfoundland, 89 fathoms, and the
superambulacral ossicles are present. i

In the center of the conical abactinal prolongation one can easily distinguish
in many specimens a small “pore” evidently connecting with the body cavity.
This is also present in many Hremicaster tenebrarius, and is what Sladen took to be
an anus in Porcellanaster. I think it must be an artificial opening caused by a
stretching of the abactinal membrane at the summit of the cone, and possibly
subsequent wearing, as the rudimentary paxille are usually more or less worn
down here.

The walls of the stomach contain numerous small straight or irregular rods
-and grains from 0.01 to 0.175 mm. in length. They are sometimes provided with
irregularities on sides or are irregularly triradiate. They are found also in the
walls of the hepatic ceca, but are not so numerous. On the lips of the peristome
they are transformed into broader irregular flattened rods with a few perforations,
but in the peristome itself are comparatively few scattered rods like those of stomach
walls, and only near the lip are they perforated. In the wall of stomach near
mouth the rods are usually simple and very regular, and tend to arrange themselves
in close meridional series.

Japanese specimens.—I have eight specimens from station 4818, Sea of Japan,
225 fathoms. One of the largest of these is figured. All have very small low
paxille with comparatively few spinelets. The superomarginals are also slightly
narrower than in shallow-water Alaskan examples, but about as in specimens from
1,033 fathoms, Bering Sea. (The latter have large paxille, with usually high
pedicels, which lack entirely the central spinelets.) The Japanese specimens have
four, but occasionally also three and five furrow spines.

Another specimen from station 5039, south coast of Hokushu, 326 fathoms,
is very different, having narrower and longer rays (R=3 r), large paxille, with
comparatively long pedicels, the summit of which lacks central spinelets, and small
actinolateral spinules. There are four furrow spines, as in the Japan Sea examples.
This is also figured.

Superambulacral plates are present.

Distribution.—Bering Sea, along the north coast of America to Melville Island;
through Barrow Strait to Greenland; south along the east coast of North America
to Cape Cod; west and north of Spitzbergen to latitude 80° 3’ N.; south to Faroe

@Mem. Mus. Comp. Zodl., vol. 32, 1905, p. 105.

36 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Islands, and on the Norwegian coast from Kristiansund to Finmark; Spitzbergen,
Barents Sea, Murman coast, Matochkin Strait, Nova Zembla, and northward;
Kara Sea as far as longitude 79° E. From here to East Cape the species has not
yet been recorded.* From Bering Sea the species ranges south into the Sea of Japan,,
and on the American side to California, and is recorded from the mouth of the

Gulf of California (Ludwig) and Gulf of Panama (Ludwig), and under the name
procurator from off the Chonos Archipelago, Chile, south to entrance of Smyth
Channel (Sladen); off the east coast of southern South America as australis (Sladen).

Specimens examined.—About eight hundred and eighty-three.

Specimens of Ctenodiscus crispatus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms

TMB s2 Near Shumagin Islands, Alaska.........-.- 110 green mud.. 85 U.S. Nat. Mus.
BOAO wo eteten a == WO Seen sweeten = riecamerecesciecs 69 12 Do.
2852.....- SS OO Pa ew ne ana sean sepecienns ace alan 58 45 Do.
2855...... of Sitkalidak Island, near Kadiak Island, 69 14 Do.

Alaska.
2860...... South end Queen Charlotte Islands......... 876 5 Do.
3075...... Off Sea Lion Rock, Washington............ 859 3 Do.
3076...... Offi Washingtour. << <<< aces wows cena se 176 3 Do.
fliseaea Near Prince of Wales Island, southeastern 322 58 Do.
Alaska.

3128...... Off Monterey Bay, California .............. 627 blue mud.... 1 Do.
$216...... South of Alaska Peninsula.................- 61 black sand, mud. 66 Do.
42 black gravel ae 1 Do.
1,033 BTCEN OOZEs emer e cae rale ees 33 Do.
65 green mud, fine sand...._.._ 41 Do.
59 dark green mud............- 16 Do.
Oa) aces do... 66 Do.
59 Preen: Mudk 35.2252 eee eon 4 Do.
76 brown mud... 104 Do.
| 74 green mud... 38 Do.
: 54 blackisand):2<..s.--¢s5-00 5 2 Do.
oe paring Sea (north of Unalaska)........ eee at 987 green mud, black lava sand. 156 Do.
3610...... Bering B00 65 oo s5 oan cco icant co cakcecncecod 75 preenmud essen. case (6) Do.
OSM Ost weiter seccen arcee cna ses ebes|aaeeees ees loca mcisevc cece caweee smeneeeenee 2 Do.

4194...... Gulf of Georgia, British Columbia. ......... 111-170 | soft green mud........._...- 1 Albatross, 1903.
4197..... i Sof O ate chon Sana eee tee aod S100; Wil es eee tks 3S Ah Sde ge ae 1 Do.
4223...... Boca de Quadra, southeastern Alaska....... 48-57 | soft green mud.............. 21 Do.
4228. ..... Near Naha Bay, Behm Canal, southeastern | 41-134 | gravel and sponges.......... 2 Do.

Alaska.
7 ba 2 UO BR SACS ROE Ae nee SORE Re 198-256 | soft graysmud: <4 se ee 1 Do.
Beer ane nla nO ndcnemmnaes tus iuce cst. hesctes ase 113-82 | green mud, fragments of 2 Do.
slate.
4235...... Near Yes Bay, Behm Canal, Alaska........ | 130-181 | gray mud, black specks....- 8 Do.
TAB see: | Kasaan Bay, Prince of Wales Island, south: | 42-47 preenimud |. oc scciscscc ease | 1 Do.
ern Alaska. |

4274......| Alitak Bay, Kadiak Island................ BEIM Nemes ces on's aesciean cae see 1 Do.
B28). Chiignik Baye. .cn.8255 tenenc soem ee | 42-43 BreenMua sees cc oe eee ee see 1 Do.
4286... wunODannks dee Seesaw ater etn Se ban ae | 57-63 | green mud,rocks............ 1 Do.
4287...... al vu yak Bay, Kadiak Island......_. pb ee mae GG-G7a i Weraycmud see aa eee 2 Do.
4202... -| BISUROURRAM GS cack ccsuecoe teeceteucee 102-94 | blue mud.................:. 2 Do.

4768... Bowers Bank, Bering Sea.................. 764 greenish brown mud........ 28 Albatross, 1906.
4775... Sel vay eee Suddcarbeenensrccnnee coer 584 Sreenimud. J=see eee ll Do.

@ The above is condensed from Ludwig, Fauna Riciics: vol. 5 iti i
aay ole gE. vol. 1, p. 451, where authorities are cited. ,

¢ Exact locality unknown.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. oT

Remarks.—Attention is again called to the marvelous variation exhibited by
this species and to its very extended range. It is extremely doubtful if Ctenodiscus
australis Liitken, 1871, from off the east coast of southern South America is a
distinct species. The differences noted by Sladen are among the most variable
characters. Compare the various figures of undoubted crispatus published here-
with, and then the figures published by Sladen. It would not be difficult to make
at least two species in Bering Sea and a third in the Sea of Japan with greater
differences than seem to exist between ‘‘australis”’ and ‘‘procurator.”’ Perrier “ has
already expressed the same doubt as to the difference between australis and European
crispatus.

This starfish, judging by its wide distribution, seems well adapted to life on soft
mud. The creatures are usually gorged with mud, from which they evidently
derive their food materials.

Family ASTROPECTINIDE Gray, 1840, emended.

Astropectinide Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 180 (includes also Luidiide,
Henricia, and Solaster).—PrErrier, Révision des Stellérides, 1875, p. 329 (includes also
Chextaster, Luidiidee, Archaster, Ctenodiscus).—V1GuIER, Squelette des Stellérides, 1879, p. 225
(includes also Ctenodiscus, Luidiide).—Prrrimr, Mém. sur les Etoiles de Mer, etc., 1884,
p- 266.—SiapEN, Challenger Asteroidea, 1889, p. 174 (includes Luidiidee and Craspidaster and
omits Plutonaster and allies).—VERRILL, Trans. Conn. Acad., vol. 10, 1899, pp. 201, 218 (omits
Plutonaster and allies).

Phanerozonia with large marginal plates, true paxille, and parapaxille; with
pointed tube feet (a flat or true sucking disk being always absent); with double
ampulle; no cribriform organs, but frequently well developed marginal fascioles
which are never webbed; with an intestine and usually an intestinal ccecum; anus
absent, small, or well developed; superambulacral plates always present.

Remarks.—This family is distinguished from the Porcellanasteride by the
absence of cribriform organs, and in having double instead of single ampulle;
from the Goniopectinide in lacking webbed fascioles, and webbed paxillar spine-
lets; from the Luidiide in having an intestine, intestinal ececum (as a rule), and
superomarginal plates; from all other families in lacking a sucking disk to the
tube feet.

The family Astropectinide as here limited includes the Astropectinide of
Sladen less his Luidiine plus his Plutonasterine. It comprises Verrill’s (1899)
Astropectinide plus his Plutonasterine. I regret that it is necessary to emend
Verrill’s classification, which was a great improvement upon that of previous authors.
A study of the genera involved, however, demonstrates that his Plutonasteride
can not be maintained, for the group has just the weakness of the old Archasteride
of Sladen and recent authors. The family can not be diagnosed with a requisite
degree of precision to exclude the majority of Astropectinide and yet include the
genera which he would desire to form the nucleus of the group. I have shown
elsewhere that the Pontasterinz of Verrill really belongs with the Benthopectinide.
The Mimasterine have well-developed suckers on the tube feet and belong near
the Pseudarchasterine.

aExp. sci. Cap Horn, 1891, pp. 143, 144.

38 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Verrill included in his Plutonasteride three subfamilies, of which two have
just been mentioned. The third, the Plutonasterine, 1 am. unable to separate by
any trenchant characters from the Astropectinide of Verrill (= Astropectinin of
Sladen). This was first brought strongly to my attention when working on the
Albatross Hawaiian collections. The genus Patagiaster, according to Verrill’s clas-
sification would go in the Plutonasterine, yet there is not a single character which
separates it sharply from undoubted Astropectinide; the same is true of Perseph-
onaster, Thrissacanthias, Dipsacaster, Tethyaster, Plutonaster, Lonchotaster, and
the recently described Ripaster. Of course, even to-day the Astropectinide are
supposed to differ from the Archasteride in lacking an anal aperture. An anal
aperture exists in several undoubted Astropectinidee (Leptychaster, Psilaster, Per-
sephonaster, Astropecten, Tritonaster). No more unstable character can be con-
jured up to separate Astropecten and its near relatives from Plutonaster, Dytaster,
or Dipsacaster. In the first place the absence or presence of an anus is difficult to
determine in some species without sectioning. In the second place, there is good
evidence that the character is variable within a genus (Astropecten, Psilaster, Lep-
tychaster). At any rate, it is proved beyond peradventure that typical Astropec-
tinidses may have an anal pore, sometimes of conspicuous size. Why not then let
the “aproctuchous” myth take its proper place in history ?

Verrill (1899, p. 199) has shown that the Archasteride is an untenable group as
defined by Sladen. He has restricted the family to the genus Archaster Miller
and Troschel. With this course I agree. As noted above, we differ on the dispo-
sition of Pontaster and the Plutonasterine. These are placed in the Plutonasteride
by Verrill, the former being made the type of a subfamily. But in Verrill’s diag-
nosis of the family (1899, p. 210) it is impossible to find any characteristic mentioned
which is not shared by Astropectinidee. In other words, the group is poorly defined
and heterogeneous. After eliminating Mimaster and Pontaster, the remainder (that
is, the Plutonasterinse) is much more homogeneous, but the genera involved (Plu-
tonaster, Dytaster, Lonchotaster) are connected with Astropecten by a very satisfac-
tory series of intergrades. I have been unable even to make two subfamilies in
the Astropectinide. If anyone wishes to try it, numerous possibly available char-
acters are furnished in the accompanying table. Here all the genera are listed,
and one may see how futile it is to attempt to separate Plutonaster and its allies,
either in a family by themselves or as a part of that taxonomical catch-all, the
Archasteride.

All genera here included agree in having pointed tube feet with double ampulle,
typical paxille (either the astropectinoid type or parapaxille), two series of mar-
ginal plates, an intestine, almost always an intestinal cecum (one exception,
Blakiaster), and always superambulacral plates. Their negative characters have
already been mentioned after the diagnosis of the family. Minor characters of
less than family importance are detailed in the accompanying table. So far as
possible the order given to the genera is such that the Plutonasterine of Verrill
would begin after T'ethyaster, and probably would include Persephonaster, as this
genus has been classified in the Archasteride, as also has Psilaster @ by Ludwig,
Blakiaster by Perrier (1884), and Tethyaster by Ludwig (1897) and Sladen (1889),

4 Plutonaster abyssicola Ludwig 1905 = Psilaster pectinatus (Fisher).

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 39

although the last is exceedingly close to Sideriaster, which Verrill (1899) considers
an undoubted astropectinid.

Table of genera of Astropectinide detailing their characters, but not including family characteristics. (The
three marked with an asterisk have not been examined by the writer.)

oe es == S oa
Ps i2./<. 15 |8.183 lee |2 (bjs |sBs
ne a| << )/2 |Bale8 ila2 13 [ges & | SE
ce |38| axle |e Se [S | Ss \8"4cs |S
man g |o-9/F2| G5 | 58/82 (Be isé | Be | 25/83 |5<
fea : BS) 95| 28 | 28) Ss |Ss./8y | 2 lesslse ise .
Som a >| 82 )]a ms le°S8lae S |Fatlasd
8 & os 2 on Ao = > |= 31B5 . a3 SwSaSslsese
a | £ 282) 92| ds |2 | 22 e2eeséla° ledsleasiaees
I am I< & < < na |O = 5 | < a
x
TiCTRY CR OMT oii in a ann enema ass Xx x KX acess seScml lasses Xx ee eeeemaae eee eis (a)
Bathyiaster <2: nein et ee ae oi ene OM teense Mi icskiee SO Nbswocet esese| Somes laccee SE Eo oeeaa
xx
Datlasieri sere os ee ee Se a Scr |e ie | ae Sid Ne SE coy oe ccs: x
x
x
x
x
x

Tee spines comparatively few, in an angular or curved comb intermediate between the two types.
allow.

¢ Lobed, imbricating, convex plates simulating tabulate paxille (perhaps pseudopaxille).

aNot always.

¢ Also protopaxille.

J A special type, derived from the astropectinid.

KEY TO THE KNOWN GENERA OF ASTROPECTINIDE.

a. No specialized spines on either series of marginal plates; marginals armed with a uniform covering
of small papilliform or squamiform spinelets.

b!. Superomarginals smaller than inferomarginals.
c!. Furrow margin of adambulacral plates angular, short, with a series of few long delicate spine-
lets. Madreporic body small, often partly hidden, but without special paxille on its sur-
face; odd interradial actinal intermediate plates ...............- Leptychaster 4 Smith, p. 42.
c?, Furrow margin of adambulacral plates gently curved, long (plate conspicuously longer than
wide), bearing about 8 short spinelets. Madreporic body large, covered with special
paxille; no odd interradial actinal intermediate plates. ............... Lonchotaster 6 Sladen.
b?. Superomarginals not conspicuously smaller than inferomarginals; both series high, confined to side
of ray; incipient marginal spinules in one superomarginal series, and two inferomarginal longi-
GUI BOLICH aan cree eine ae eae cee ces enn ceases Bathybiaster ¢ Danielssen and Koren.

@Includes Parastropecten Ludwig and Glyphaster Verrill.
6 Refers to Lonchotaster forcipifer; see sections a?, b?, c?, d?.
¢ Bathybiaster robustus frequently has the marginal spinules very small.

40 BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

a, More or less conspicuous spines or spinules on inferomarginal plates, and also often on superomargi-
nals as well. : ;

b). Both series of marginals nearly equally developed, forming a more or less vertical lateral face to
ray: at least, the superomarginals are not small and confined to abactinal surface of ray; the
inferomarginals not forming lateral border to ray. f ;

c, Inferomarginal plates touching adambulacrals throughout ray; not separated by a series of actinal

intermediate plates, for a part, or the whole length Obra viesseeeeceee Astropecten Gray, p. 55.

©, Inferomarginals separated from adambulacrals for a part or a whole of the ray by one or more
series of actinal intermediate plates. ‘

d', Superomarginal plates thick and block-like; at least not very thin and confined to side wall

of ray.
¢). Actinal interradial areas rather small and intermediate plates few (40 per cent or less of
minor radius measured on dorsum; madreporic body never large and concealed by paxille;
inferomarginals never defining ambitus).
f?. Marginal plates not tumid, but with special elevations, there being narrow vertical rather
shallow fascioles between the plates; actinal spinelets more or less sacculate.

g}. Adambulacral armature forming a very angulated series on the furrow margin, the
median spine considerably longer than the rest, and with a flap of integument at tip;
all actinal spinelets very sacculate, and marginal spinules inconspicuous; abactinal
paxillz with stellate, imbricating bases.....-.-. Bathybiaster 2 Danielssen and Koren.

g?. Adambulacral armature forming a curved series, the spinelets more numerous as a
rule and central spinelet not enlarged nor bearing a conspicuous terminal flap; actinal
spinelets frequently sacculate, and marginal spines or spinules conspicuous; abactinal
paxillee with subcircular or subpolygonal bases; not stellate. .Psilaster Sladen, p. 71.

/?. Marginal plates distinctly tumid, the fascioles between them obsolete; actinal spinelets
not sacculate.

g. Gonads confined to interradial area of disk; not extending into arm; superomarginal
spines present or absent; no enlarged actinal adambulacral spines.

2. Anal pore present; intestinal coecum consisting of two bi- or trilobate sacs; a very
narrow median radial area free from papule; adambulacral plates with gently
curved furrow margin bearing numerous spinelets in a regular comb, and on actinal
surface of plate several parallel rows of similar spinelets; abactinal paxille polygo-
nal, independent; one or two, more or less, prominent superomarginal spines
sometimes present; no odd interradial actinal intermediate plates.

Persephonaster 6 Alcock.

#2. Anal pore probably present; no intestinal ccecum; a conspicuous median radial area
free from papule; adambulacral plates with an angular furrow margin bearing five
spinelets, and a few similar ones on actinal surface of plates as in Leptychaster;
abactinal plates lobed, imbricating, with low convex tabula; no superomarginal
spines; odd interradial actinal intermediate plates...........- Blakiaster¢ Perrier.

g. Gonads extending a short distance along dorsal wall of ray, on either side; prominent

bristling superomarginal as well as inferomarginal spines; anus present and easily
detected; one or two enlarged actinal adambulacral spines.
Thrissacanthias Fisher, p. 78.

@ Includes Phoxaster Sladen, and probably /lyaster Danielssen and Koren. Thelatterhasalong apical
cone or funnel, and is evidently a very young individual.

b Includes Psilasteropsis Fisher.

© Near, if not identical with this genus, is Bunodaster Verrill, as yet not fully described. The type
is B. ritteri, from “California.” (See Verrill, American Naturalist, vol. 43, 1909, p. 554, fig. 4.) Pro-
fessor Verrill has kindly sent me enlarged photographs of the abactinal and actinal surfaces of the type.
The abactinal plates are in the form of low convex parapaxille, largest over central portion of disk.
The general appearance of the creature is strikingly like Blakiaster, but I have no details of its anatomy.
The actinal surface is identical with that of Blakiaster, even to the details of the actinal intermediate
plates and the presence of the odd interradial actinals.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 41

e. Actinal interradial areas fairly extensive with numerous intermediate plates arranged in
definite series extending from adambulacrals to inferomarginals; (actinal intermediate
plates occupying 50 per cent or more of minor radius, measured on dorsum, but marginals
not greatly unlike in size, nor inferomarginals extending laterally beyond the
superomarginals). .

f*. Madreporic body hidden by special paxillee on its surface; inferomarginal plates and
sometimes superomarginals with a single stout erect spine or spinule; abactinal paxille
only slightly lobed, or polygonal, not markedly stellate. Adambulacral armature not
astropectinoid; furrow spinelets in a regular comb.............-.-- Plutonaster Sladen.

« f?. Madreporie body naked and large; adambulacral armature astropectinoid; abactinal
plates stellate; marginal spinulation variable but not as above.

g'. Inferomarginals with large flattened spatulate spinelets, chisel-shaped, roundish or
square cut at tip, closely packed, with here and there a transverse row of five, but
usually without order; actinal intermediate plates closely packed with stout flattened
spatulate spinelets. Paxillee with long pedicels and compact crowns of short
iInbending spineleisy..s- se kscsh seen. seesaw Stone ce eon Moiraster @ Sladen.

g’. Inferomarginals with a single transverse row of pointed spines in addition to much
smaller papilliform or squamiform spinelets; actinal intermediate plates with clusters
of papilliform or granuliform spinelets, separated by deep fascioles; paxillee with
ouly moderately long pedicels, sometimes short, the spinelets forming a distinct
central group surrounded by a peripheral series; adambulacral plates separated by
fasciolar grooves.

h‘. Gonads extending far along ray, on either side; a distinct medioradial series of paxille,
slightly larger than two or three adjacent series (especially on outer part of ray);
adambulacral fascioles shallow; deposits in tube feet; first adambulacral plate
considerably compressed; central actinal adambulacral spinelets not increasing in
length toward end of ray; madreporic body moderately large.... Tethyaster Sladen.

h®. [Gonads unknown]; no definite medioradial series of paxille slightly larger than
those adjacent (?); adambulacral fascioles deep, densely lined with small delicate
spinelets; stoutest actinal adambulacral spine increasing in size toward end of
Tayi madreporic body, verylarge — 2 .ces- S55: ce ein eee eee Sideriaster’ Verrill.

@. Superomarginal plates thin and confined to side wall of ray; no fascioles between them.

e'. Marginals with a single, rigid, sharp, perpendicular spine; madreporic body large with
Paxil lees Ong rtsRUTIACe eens apse teen oe tee erie acs emesis Dytaster¢ Sladen, p. 101.

. Marginal plates excessively thin and confined to side wall of ray, both series with a vertical
series of five or more appressed flattened spines; madreporic body small, naked.

Ripaster Koehler.
b?. The two series of marginal plates unequal in size, often forming an angulated rather than vertical
margin to ray; superomarginals much smaller than inferomarginals or at least confined wholly
to abactinal surface on rays; inferomarginals often extending laterally beyond superomarginals,

thus defining the border of rays (an exception in Lonchotaster).
c'. Madreporic body not large and hidden by special paxille on its surface; anal aperature very

sinall, or absent.
d'. No well-developed series of plates between inferomarginal and adambulacral series throughout
ray (rudimentary at base of ray in Tritonaster).

e!. Superomarginal plates not conspicuously smaller beyond middle of ray. Marginal fascioles
wellideveloped esas ys tes Bs Jee. Fae ce teem bcuee okie Astropecten Gray, p. 55.

e*. Superomarginals very much reduced in size on outer part of ray; inferomarginals very tumid
with an oblique series of long seta-like spines; no marginal fascioles. .. Tritonaster Fisher.

aSladen, Challenger Asteroidea, 1889, p. 192. Type, Archaster magnificus Bell, Ann. and Mag. Nat.
Hist., ser. 5, vol. 8, 1881, p. 440; see Keehler, Trans. Roy. Soc. Edinburgh, vol. 46, 1908, pt. 3, p. 630, pl.
12, figs. 107-110.

b Verrill, Trans. Conn. Acad., vol. 10, 1899, p. 219. Type S. grandis Verrill. The genus is very
closely related to Tethyaster.

e Includes Crenaster Perrier, not d’Orbigny,

2 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

@. A single well-developed series of actinal intermediate plates between inferomarginals and
adambulacrals. Inferomarginals broad, short, and band like, their spines many and
appressed to ray. Rays long, disk pmall lremmreetiane Serer ae eater cla Ctenophoraster Fisher.

c?, Madreporic body large, hidden by paxillse on its surface; anal aperature always present, large
in Dipsacaster. 5 ,
d', Margin of ray defined by inferomarginals; superomarginals confined to abactinal surface on
ray; deep marginal fasciolar grooves; actinal intermediate areas large; no pedicellariz;
y; g f
paxille typical. i
e'. Gonads disposed in a series along either side of abactinal integument of ray, extending

beyond middle; papule distributed all over paxillar area......- Dipsacaster Alcock wp. 85.
2. Gonads confined to interradial regions not extending into rays; papule absent from central
portion of disk and a MIGrad ial DANGiee ae ces Sacer es erisate concer e te Patagiaster Fisher.

@. Margin of ray defined by both series of marginal plates which are small and armed with a
single small spine in addition to papilliform spinelets; marginal fascioles shallow; actinal
interradial areas small, and intermediate plates not extending far along ray; actinal granuli-
form pedicellarise and low abactinal paxille; closely related to Dytaster.

Lonchotaster @ Sladen.
Genus LEPTYCHASTER? Smith.

Leptychaster Smrru, Ann. and Mag. Nat. Hist., ser. 4, vol. 17, 1876, p. 110. Type, L. kergue-
lenensis Smith.
Leptoptychaster Surrn, Philos. Trans., Roy. Soc., Zool. Kerguelen Island, vol. 168, 1879, p. 278.
(Emended from Leptychaster.) :
Parastropecten Lupwic, Mem. Mus. Comp. Zodl., vol. 32, 1905, p. 76. Type, P. inermis
Ceo ate. Amer. Nat., vol. 43, Sept., 1909, p. 553. Type, Leptychaster anomalus
Fisher. :
Diagnosis.—Rays normally five in number; abactinal surface flat or slightly
arched by inflation of dorsal integument; actinal surface beveled on sides; disk
variable in size, medium to small; rays typically of medium length, tapered; marginal
plates entirely without specialized spines, the superomarginals typically very small,
the inferomarginals short and band-like; deep fasciolar grooves between specialized
transverse ridges of marginal plates, these grooves lined with close-set delicate
spinelets; abactinal area covered with true paxille with subcircular to irregularly
stellate bases on papular areas; crown of paxilla with numerous very small spinelets;
papuls absent from a circular area in center of disk and from a conspicuous median
radial area (see Dipsacaster); actinal interradial areas small to medium sized, but
much larger than in Asfropecten; actinal intermediate plates low-paxilliform,
imbricating; adambulacral armature, several longitudinal series of three to five
spinelets; mouth plates with long furrow margin, and inner spines enlarged; ambu-
lacral feet conical without a sucking disk; no pedicellerie; superambulacral plates
present, though small; gonads interradial not extending along ray; anus absent or
sometimes present.
KEY TO THE SPECIES OF LEPTYCHASTER HEREIN DESCRIBED.
a‘. Superomarginals small, forming a narrow margin to abactinal area; not conspicuously wider in inter-
radial angle than at middle of ray.

bt. Inferomarginals placed very obliquely, their specialized ridges when cleared of spinelets only
one-half width of intervening fascialar grooves at base of ray. Superomarginals paxilliform.
arcticus, p. 43.

2 Refers to L. tartareus in particular.

» All the evidence is against a typographical error, so the original spelling has been maintained in
accordance with modern usage.

ASTEROIDEA OF NORTH PACIFIC-AND ADJACENT WATERS—FISHER. 43

b?. Inferomarginals more transversely placed, and less numerous; specialized ridge of each plate about

as wide as adjacent furrow. Superomarginals, broader and not paxilliform; quadrate in shape.

propinquus, p. 54.

a’. Superomarginals larger, forming a conspicuous margin to abactinal area and wider and shorter in
interradial angle than at middle of ray.

b'. Stellate with sharp rays; abactinal plates decidedly stellate or lobed on papular areas; R.=2.3 to
3.1 r; inferomarginals (with spinelets) more than twice as wide as long; marginal plates twenty-
iwotopbirtyy-o ven) ad ultapecimens ic, aiseyae eos e se tone deme es pee oe Seas pacificus, p. 45.

6°. Stellato-pentagonal, or stellate with short blunt rays; abactinal plates only obscurely lobed or stel-
late on papular areas; R.=1.6 to 2 r; inferomarginals on ray twice or less than twice as wide as
long; marginal plates ten to eighteen in adults -.........-.-...-----.--.---- anomalus, p. 48.

LEPTYCHASTER ARCTICUS (Sars).
PL. 8, fig. 1; pl. 9, fig. 4.

Astropecten arcticus M. Sars, Reise i Lofoten og Finmarken, Nyt Mag. f. Naturvidensk., vol. 6,
1851, p. 161; Fauna Litt. Norvegie, 1856, Heft. 2, p. 61, pl. 9, figs. 16-18; Oversigt af
Norges Echinodermer, 1861, p. 32.

Astropecten liitkeni Barrett, Ann. and Mag. Nat. Hist., vol. 20, 1857, p. 45, pl. 4, figs. 3a, 6, c.

Astropecten arcticus and liitkeni Dusarpin and Hur, Hist. nat. zoophytes Echinodermes, 1862,

- 428.

baeie arcticus VERRILL, Amer. Journ. Sci. and Arts, ser. 3, vol. 16, 1878, p. 214.—Storm, K.

Vidensk. Selskabs Skrifter, Throndhjem, vol. 8, 1878, p. 252—Prrrrer, Nouv. archives
du mus. d’hist. nat., ser. 2, vol. 1, 1878, pp. 32, 88.

Leptoptychaster arcticus SLADEN, Challenger Asteroidea, 1889, p. 189.—Lupwice, Fauna Arctica,
vol. 1, 1900, p. 452.

Diagnosis..—Rays five. R=32mm.; r=10 mm.; R=3.2 r. Breadth of ray
at base, 12 mm. (measured from interradial line). Disk medium-sized, rays taper-
ing gradually to a blunt extremity; interbrachial areas rounded but rather abrupt;
abactinal surface slightly inflated ; sides of ray rounded; superomarginal plates small,
resembling enlarged paxille, not markedly wider at base of ray than at middle
or on outer third, forming a narrow margin to abactinal area; inferomarginals short,
but fairly wide, placed very obliquely and forming the rounded edge to ray, most of
the series being visible from above; no specialized spines on either series; adam-
bulacral plates with a furrow series of three or four spinelets, and on actinal surface
two or three longitudinal series (often irregular) of three to five spinelets each, with
sometimes two or three small spinelets out of line; abactinal paxille very compactly
placed, longest on interradial area of disk and on either side of median radial area
along ray; actinal interradial areas fairly large, the intermediate plates extending
far along the ray.

Description.—The paxillx are closely placed and the difference in size between
those of the midradial region of ray and center of disk, and those along side of paxillar
area (particularly in the interradial region) is very marked—much more so than in
Leptychaster pacificus. A large paxilla of the interradial region of disk presents
a slightly convex crown of small terete, compactly placed spinelets, varying in
number according to the size of the paxilla, a fairly large one having twenty to
twenty-five peripheral and about the same number of central spinelets, a trifle
shorter than column of paxilla. Papuls absent from center of disk and along median

«From specimen taken at station 4792, in the vicinity of the Commander Islands.

44 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

radial area: five or six (sometimes four) about each plate irregularly, on inter-
radial regions and along border of paxillar area of ray. Abactinal plates slightly
and irregularly lobed on papular areas, much less obviously so than in pacificus.

Superomarginal plates, forty-one in number from interradial line to extremity of
ray, are much smaller than in either of the two other species described below, and
have the form of large paxille, being not conspicuously wider in interbrachial angle
than midway along ray. They are irregular or subcircular in outline, rather straight
sided adjacent to inferior series, have the appearance of being obliquely oriented,
and do not correspond exactly to inferomarginals. Opposite the first ten supero-
marginals are seventeen abactinal paxille (or irregular transverse series). The
superomarginals are confined to the abactinal surface and their spinelets are heavier
and slightly more numerous than on the other paxille. Terminal plate wider than

with a rounded end; notched adjacent to paxille.

Inferomarginals much shorter than wide, obliquely oriented (about 45°) to
transverse axis of ray, and occupying entire side wall of ray. Each plate is therefore
strongly arched, this arch giving the rounded margin. Each plate is covered with
spinelets, stouter than those on superomarginals, somewhat squamiform on exposed
surface, and very slender in the fasciolar grooves. These fasciolar grooves are deep,
and about twice as wide at base of ray (taking transverse axis as length of plate)
as the adjacent specialized ridge of the plate. The latter is a thin, almost laminar
ridge, very much thinner than in pacificus, where the specialized ridge is as wide or
wider than the grooves. Farther along ray in arcticus the ridges become thicker,
nearly or quite as wide as grooves and the inferomarginals are more massive. A
longitudinal shallow furrow separates the two series of plates.

Adambulacral plates with a curved furrow series of three or four rather long,
slender, blunt, terete spinelets. Lateralmost spinelets slightly the shortest. On
actinal surface of plate the spinelets decrease gradually in size outward, there being
two or three longitudinal series of three or four spinelets, occasionally more. Some-
times a very few odd spinelets stand out of line. The first plate is wider than the
rest, with more numerous spinelets.

Mouth plates narrow, the combined pair forming a salient angle into actino-
stome. Furrow margin long, with about ten spinules like those of adambulacral
plates, decreasing in size outward, and continued along adambylacral margin in
several smaller spinelets; innermost two spinules forming ‘‘teeth” at angle. An
irregular series of spinelets smaller than furrow spinules stand along edge of median
suture; and at outer end of plate are a few intermediate spinelets between the two
series, which throughout the length of plate stand rather close together.

Actinal interradial areas fairly large, the plates being arranged in series between
inferomarginals and adambulacrals. A single series of plates extends about half
length of ray (from interradial angle); a second series, one-fourth; a third series,

one-eighth or less. Between first superomarginal and first adambulacral are about
six plates in an interradial direction.

long,

Madreporic body nearer margin than center of disk; surrounded by six or seven

large paxille which overhang the edges; ridges coarse, transverse rather than
centrifugal.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 45

Variations.—There are three specimens from station 3602 which are difficult
to classify satisfactorily. Two resemble arcticus but have a larger disk, shorter and
broader rays, and larger superomarginals. The third much resembles anomalus,
and in the sum of its characters stands about midway between the two specimens,
above referred to, and anomalus, which was taken at the same station. The aber-
rant specimens have inferomarginals resembling those of areticus. R=37 mm.
(largest specimen); r=15 mm.; R=2.5r. Breadth of ray at base, measured from
interradial line, 18 mm. It is not improbable that this species hybridizes with
anomalus whenever the ranges overlap, and that the very aberrant specimens may
be explained by such a theory.

— Type-locality.—Oxfjord, Finmark, 100 to 150 fathoms.

Distribution.—The distribution of this species is evidently cireumpolar. In the
Atlantic hemisphere it is found along the east coast of North America from lat.
38° to 45° N., and on the coast of Europe from south of Ireland, the Faroe Channel,
off Norway from Trondhjem to Finmark, and eastward to Barents Sea and the
Murman coast.? In the north Pacific region the species ranges over Bering Sea and
south on the Asiatic side to Yezo, Japan.

Specimens examined.—One typical from station 4792, vicinity of Commander
Islands, 72 fathoms, pebbles; two aberrant forms, 3602, vicinity Pribilof Islands,
Bering Sea, 81 fathoms, green mud, sand; one from 5047, off Hokushu, Japan, 107
fathoms, dark gray sand, broken shells, pebbles.

Remarks.—The specimen from near the Commander Islands agrees in most
particulars with an example from the coast of Maine (station 21, Cashes Ledge).
The Atlantic specimen has the raised ridges of inferomarginals at base of ray,
slightly wider.

LEPTYCHASTER PACIFICUS Fisher.
Pl. 8, fig. 2; pl. 9, fig. 2; pl. 50, figs. 1, la.
Leptychaster pacificus Fisher, Proc. Wash. Acad. Sci., vol. 8, Aug. 14, 1906, p. 112.

Diagnosis.—General form similar to that of Z. arcticus (Sars). General form
flattened; rays evenly tapered, bluntly pointed; interbrachial angle slightly
rounded, but abrupt; abactinal surface subplane; margin of rays defined by
inferomarginal plates, rounded; superomarginal plates well developed, relatively
larger than in L. arcticus, forming a fairly conspicuous margin to abactinal paxillar
area; actinal surface slightly convex; paxille compact, the larger with about
twenty-five peripheral and thirty central spinelets; plates of papular areas lobed;
papule in fives and sixes about each; adambulacral plates with four or five furrow
spinules and on actinal surface two or three longitudinal series of about four similar
spinules. Rays five. R=43 mm.; r=14 mm.; R=3r. Breadth of ray at base,
16 mm.

Description.—Abactinal paxillar area fairly compact, the paxille decreasing
in size toward center of disk, midradial line, and end of ray; smallest paxille in
center of disk, the largest on margin of area at base of ray. Paxille similar in

@ Condensed from Ludwig, Fauna Arctica, vol. 1, p. 452.

46 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

character to those of L. arcticus, but slightly larger, and spinelets a trifle longer,
Base of pedicel flaring into a roundish plate with four or five short, rather irregular.
lobes by which the plates touch or imbricate slightly, and between which the
papule emerge. Larger paxille with about twenty-five peripheral and thirty
central slender delicate terete blunt spinelets; spinelets occupying center of tabulum
form a codrdinate flat-topped group, usually stand upright, and are crowded;
peripheral spinelets usually radiate and are not equal in length. ol

Abactinal plates, in a prepared specimen, from inner or coelomic side: Plates
rather small and close-set irregularly stellate with short irregular lobes by which the
plates overlap. Papule absent from a narrow radial line, interrupted on the
interradial line, and absent from a circular area (with radius about 0.4 r) im center
of disk. On the radial line the plates are more crowded, and are without regularity
as to the number of lobes and in arrangement. The plates are also variable in
size and more uneven in contour than those on the papular areas. The plates
on the central nonpapular area of disk are small, roundish, and crowded, with a
number of larger scalloped or irregularly incised plates scattered here and there.
On the papular areas the plates are arranged in transverse rows parallel with inter-
radial line, but the regularity is frequently interrupted. Adcentrally to the mad-
reporic canal is a large tumid plate, rounded on the outer side and angular toward
the canal. On either side are two smaller more elongate plates, while a small plate
completes the circle on the outer side.

Marginal plates short, band-like, but both series more conspicuous than in
L. arcticus; superomarginal plates, thirty in number, from interradial line to ex-
tremity of ray much wider than long on proximal half of ray, the width rapidly
decreasing on outer portion until plates are nearly quadrate. Plates form an
arched bevel to margin of abactinal area, are separated by deep fasciolar grooves,
and are covered with short delicate terete spinelets, which form a close nap all over
exposed surface.

Inferomarginals corresponding to superomarginals, beyond which they extend
laterally, forming the rounded margin to ray. They are separated from supero-
marginals by a rather wide groove, which is not so deep as the transverse fasciolar
furrows, these being deeper between inferomarginals than between superomarginals.
Plates short, band-like, obliquely oriented to radial line (superomarginals being
transversely oriented), forming a well-arched bevel to actinal surface. First plate
about twice as wide as corresponding superomarginal (sometimes somewhat more).
All plates densely covered with small spinelets similar to those of superomarginals,
but a trifle larger, those of transverss median region slightly squamiform and
directed outward.

Actinal interradial areas about the same size or a trifle smaller than in L. are-
ticus; one series of intermediate plates extending about three-fourths length of ray
or to eighteenth inferomarginal; a second series extending to seventh or eighth
plate, and a third series confined to angle bounded by adjacent first two plates.
Intermediate plates with a low tabulum crowned by a codrdinate group of fifteen or

twenty papilliform spinelets, those in center being slightly thicker and more clavate
than the peripheral ones.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 47

Adambulacral plates about as wide as long, with a rounded furrow margin,
but first two or three plates wider than long and with more angular margin. Arma-
ture consists of (1) a furrow series of four (more rarely five) slender, rather long,
blunt cylindrical spinules, the two central being slightly the longest or the four
subequal; (2) on actinal surface are two or three longitudinal series of about four
similar spinules, which decrease in size toward outer edge of plate; third series
when present more irregular, its spinelets distinctly tapered, slenderer, shorter, and
sharper. Furrow spinelets usually bent back from furrow, and armature has a
decidedly crowded appearance.

Mouth plates narrow, the free margin of each being longer than that adjacent
to first adambulacral, and the combined plates forming a salient angle into actinos-
tome. Margin of plate with a series of about 15 slender tapering spinules, decreasing
in length from inner to outer end of plate. About eight to ten of these are more
regular and occupy the free actinostomial margin, the rest being adjacent to first
adambulacral plate, between which and the mouth plate there is a fairly wide suture
furrow, and sometimes an incomplete, irregular, intermediate series is present.

Madreporic body situated about its own diameter from inner edge of supero-
marginal plates, fairly large, surrounded and partially obscured by large paxille;
striations deep, coarse, eedlen centrifugal.

Superambulacral lees present, but small; one opposite each ambulacral plate
except the first plate and a few at the very tip of ray.

Variations.—The variations exhibited by specimens of nearly equal size are
slight. The relative dimensions, with the number of superomarginal plates, are
listed below for several specimens:

Measurements of Leptychaster pacificus.

| | Breadth of | Number of
Station. R:r. ray at supero-

| 8 aa base. marginals.

Tees pate |
2862 24 Ee 9.0 2,6 31 10.0 22
2862 40 14.0 2.92:1 16.0 | 26
3223 32 14.0 2.3 :1 16.0 24
4194 39 12.5 3.12:1 15.0 | 30
4198 23 7.5 3:1 9,0 | 27
2862 44 | 14.5 3:1 16.5 | 35

The paxillar area is slightly narrower in some examples than in others, and
the relative amount of inflation of the abactinal integument also changes the
external facies somewhat.

Young.—There are three small specimens which seem to be the young of this
species; two from station 3223 have broader, more robust rays than a nearly equal-
sized individual from 4194. The characters of the latter are as follows: R=9 mm.;
R=4 mm.; R=2.25 r. Breadth of ray at base, 5 mm.; superomarginals, fifteen;
paxille with three to five central and eight to ten peripheral spinelets; supero-
marginals inconspicuous; abactinal integument inflated; furrow spinelets four
usually, the two central longest; actinal intermediate plates extending less than
one-third length of ray measured from interradial angle; terminal plates conspicuous.
One of the examples from station 3223 has the following characters: R=10 mm.;

48 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

r=5mm.; R=2r; breadth of ray at base 6 mm. ; superomarginals, fifteen; larger
paxille about the same as in foregoing specimens, but spinelets shorter and area
broader; furrow spinelets usually three ; actinal intermediate plates extending more
than one-third length of ray (nearly one-half); superomarginal plates more robust
than in foregoing example; terminal plate conspicuous.

Type.—Cat. No. 21925, J.S.N.M.

Type-locality.—Albatross station 2862, near north end of Vancouver Island
(inside) in 238 fathoms, on gray sand and pebbles.

Distribution.—Southern part of Bering Sea to Vancouver Island and from 56
fathoms in northern limit of range to 238 in the southern. Dwells on sand, pebbles,

and soft mud.
Specimens examined.—Sixteen from the following stations:

Specimens of Leptychaster pacificus examined.

| Number
Station Locality. Depth. Nature of bottom. of speci- Collection.
aoe) , mens.
| Fathoms.
SEBG rege na sai Off Afognak Island, Alaska.........-- 68 Shellse a eceeseoreee cceema 1 U.S. Nat. Mus.
BROR ne eatseee Vancouver Island, British Columbia 238 gray sand, pebbles.........- 8 Do.
(near north end, inside).
ERR cae eee Bering Sea, vicinity of Unalaska...... 56 blackspebblest ===: 2-5-=----5 3 Do.
BAI Ger een ee GO sere cc nce can nee roce ee wasrats 61 black sand, gravel, shells... 1 | Do.
4100 Straits of Georgia, off Nanaimo, Brit- | 111-170 | soft green mud..........-..- 2 Albatross, 1903.
ish Columbla.

MIOB Soo cases clans Oso se aeneinoe se aeeee ance ceteanae 157-230 |....- CO bere ajeneeciee noe eee 1 | Do.

LEPTYCHASTER ANOMALUS Fisher.

Pl. 7, fig. 4; pl. 9, fig. 1; pl. 50, figs. 2, 2a.

Leptychaster anomalus Fisner, Proc. Wash. Acad. Sci., vol. 8, Aug. 14, 1906, p. 115.
Glyphaster anomalus VERRWL, Amer. Nat., vol. 43, Sept., 1909, p. 554.

Diagnosis.—Rays five. R=27 mm.; r=1.6 r. Breadth of ray at base, 19
mm. Usual form stellato-pentagonal; disk broad, rays short, broad, and blunt;
interbrachial arcs shallow and wide; general form depressed; abactinal surface
subplane, capable of slight inflation; actinal surface slightly arched due to rays
bending upward; marginal plates conspicuous, few in number, devoid of any
enlarged spines or specialized armature, but covered with short spinelets; actinal
intermediate areas broad; adambulacral plates with three or four furrow spinules,
and on actinal surface with three longitudinal series of smaller spinelets, decreasing
in length toward outer edge of plate; small superambulacral plates present; a very
small anal pore present.

Description.—Abactinal paxillar area compact; paxille arranged in not very
regular oblique transverse rows at sides of ray; without order in median radial area
and center of disk. Paxille largest at base of ray and in interradial areas, decreasing
conspicuously in size toward center of disk and tip of ray; larger at sides of paxillar
area than in midradial region. Column of paxilla about as high as breadth of base,
flaring at summit, the largest crowned with a coordinate floriform group of about
forty or forty-five short, terete, often clavate, round-tipped spinelets; of these

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 49

about one-half form a peripheral series and are a trifle slenderer and longer. On
the smaller paxille the spinelets decrease markedly in size, but only slightly in
number.

Abactinal plates in a prepared specimen, from inner or ceelomic side. Plates
small, closer together along midradial line and in center of disk where there are no
papule than at sides of area where papule are present. Plates of latter region are
circular in general shape, but nearly always more or less irregular; with four, five, or
six short, abrupt, lobes irregular in length, thickness and in distribution on the periph-
ery of plate. These plates of papular area are arranged in perceptible, although
irregular, oblique transverse rows (parallel with interradial line). Usually six
papule occur around each plate (often five, and rarely four) emerging between the
irregular lobes by which plates usually touch. Plates along median area of ray are
even less regular than others, and frequently there are no lobes at all. They are
slightly smaller, and papule are absent from a strip about four plates wide. Toward
center of disk, plates decrease rapidly in size, become more crowded, and lobes if
present are very inconspicuous and of reeilee occurrence. The large primary
interradial plate placed just adcentrally to the madreporic canal is convex internally
(bears a large paxilla externally), and is slightly concave on the outer edge, next to
madreporic canal; three or four other less regular and smaller plates complete the
circle about the madreporic canal.

Superomarginal plates, fifteen in number from median interradial line to
extremity of ray form an arched bevel to border of abactinal surface; plates shorter
than wide, but increase in length on outer half of ray. Plates of both series separ-
ated by transverse narrow deep fasciolar grooves and a narrow deep groove from
inferomarginal series. Superomarginal plates covered with short, terete, blunt,
granuliform spinelets, similar to but larger than paxillar spinelets, becoming well
defined, slender spinelets in the fasciolar grooves. The superomarginal spinelets
are close-set and small, forming an even nap on the exposed surface of plates.

Inferomarginal plates much wider than long, encroaching more onto actinal
area than do superomarginals onto abactinal, and corresponding in position to
superomarginals. Spinelets, densely covering surface of plates, larger than those of
superomarginals, and increasing in size toward outer end of plate which projects
slightly beyond adjacent end of superomarginal, thus defining the ambitus. Infero-
marginal spinelets somewhat granuliform in middle of plate, often attaining a
squamiform appearance near outer end; spinelets in fasciolar furrows, slender.
Spinelets on extreme upper or outer edge of plate (above the squamiform spinelets)
are similar to those of adjacent portion of superomarginals, the true ambitus or
edge of ray being a little below the margin of plate. It is on this rounded edge of
ray that the largest spinelets occur. No enlarged spines or tubercles on either
marginal series. Terminal plate small, granulose, deeply notched below.

Actinal interradial areas large; intermediate plates low-paxilliform, arranged
in chevrons, the series adjacent to adambulacrals extending about three-fourths
length of ray or to eighth inferomarginal. Plates decrease in size toward margin,
are strongly imbricated internally, and the paxillar crowns which are composed of
about twenty-five to thirty clavate obtuse, not very crowded, spinelets (slender

57444°—Bull. 76—11——4

50 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
when dry) surmount a very low convex elevation or tabulum. Well-defined fas-
ciolar channels separate these tabula.

Adambulacral plates about as wide as long, with a slightly rounded, angular
furrow margin, the angularity being more pronounced in vicinity of mouth plates.
Armature consists of (1) a furrow series of four (sometimes three) terete or slightly
flattened bluntly pointed tapering spinules about as long as plate and graduated in
length orad, the longest spine being on aboral end of plate; or the spinules | may be
disposed like rays of fan and graduated in length toward either end of series. @)
On actinal surface are about three longitudinal series of smaller spinelets, decreasing
in length toward outer edge of plate where the spinelets are like those of actinal
intermediate plates. Four spinelets commonly occur in the inner actinal series
and about three to five in each of the outer; or the two latter series may be wanting,
the spinelets, instead, forming an irregular group, especially on outer part of ray
where there are frequently upward to sixteen or twenty actinal spinelets.

Mouth plates narrow, rather prominent actinally, the free margins of the
combined plates forming a salient angle into actinostome; free margin of each
plate slightly angular near inner end and longer than the margin adjacent to first
adambulacral. Armature consists of a furrow series of about six or seven tapering
spinules decreasing in length from the inner enlarged tooth, outward, and thence
continued along margin adjacent to first adambulacral in about nine much smaller
spinelets similar to those of actinal intermediate plates. A superficial series of
similar spinelets follows margin of median suture, increasing in size toward inner
angle of plate, and an incomplete more or less irregular series often, but not always,
occurs between marginal and superficial series. There is more or less variation in
the details of dental armature. The exposed, outer, slightly convex surface of the
combined plates has the appearance of being covered with a bristling armature of
short papilliform spinelets, very similar to those on adjacent actinal intermediate
plates.

Madreporic body rather large, about midway between center and extreme edge
of disk. Striations coarse, centrifugal, very irregular; madreporic body sometimes
nearly hidden by five or six large paxille.

Superambulacral plates present, though small. Absent from the first ambu-
lacral plates, and from the distal six or seven, which are much reduced and crowded
against the adjacent inferomarginals. Gonads forming a tuft of tubules on either
side of the interradial septum, five or six tubules (two or three times dichotomously
divided) to each tuft. The gonads do not consist of a series of tufts extending
along the ray as in Dipsacaster. A Polian vesicle in each interradius. Interradial
septa uncalcified.

Variations.—The specimens assembled under this form present a very con-
siderable amount of variation, and when the extremes are placed side by side it is
hard to believe that there are not two valid species. But there is such a bewildering
number of more or less perfect intermediate stages that one is forced to range them
all under one head.

. The most important variations occurring in this species are in respect to
dimensions and the size of the marginal plates. Some examples have a more stellate

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER,. 51

form, while others verge onto an arcuate pentagonal shape. Again, one extreme
has fairly massive superomarginals, which are large up to the tip of the very blunt
ray, while in the other extreme, the upper marginals are more numerous and rela-
tively smaller, and the tip of the ray is not so blunt. The latter form is not so
numerous as the first, which is ‘‘typical.’”’ The effect of this difference of shape
on the relative dimensions is shown in the following table:

Table showing effect of shape on relative dimensions in Leptychaster anomalus.

] | ]
Breadth | Number of
Station. R. iv Rir. of ray at supero-
| base. marginals.
|
3310 | 7 1.6:1 | 19.0 15
4281 21 | 13 1. 62:1 16.0 10
3334 25) ae eS Tel el el. 13
4281 29 17 nee 19.0 13
3486 28 14 2:1 16.0 18

The number of spinelets on the outer half of the adambulacral plates (i. e., on
actinal face) varies slightly, and they are a trifle longer and slenderer in some indi-
viduals than in others. The tip of the ray in this species is rounded and blunt,
while it is sharp in pacificus. This, with the fewer and more transversely placed
inferomarginal plates of anomalus, will serve to distinguish the longer rayed indi-
viduals from pacificus. Variation within narrow limits occurs in the size of the
paxillz, in general the examples with more massive marginals having the larger
paxille. The greatest difference in general facies is caused, however, by the varia-
tion in size of marginals referred to above. A slight difference in width causes a
varying amount of encroachment upon the abactinal paxillar area. In general the
shorter armed individuals have wider and more massive marginal plates, but this is
not invariably the case. The extreme variants in proportions are found at the
same station.

Young.—The smallest specimen (station 4538) measures R=8 mm., r=6 mm.;
rays broad, stout, blunt, with seven rather massive superomarginals, which are
larger relatively than in adults. Paxille with nine or ten peripheral and two
to five central spinelets; adambulacral plates with four furrow spines. Terminal
plate small. This species has much larger marginal plates than equal sized Pseu-
darchaster pusillus, and the latter species has very large terminal plates and spiny
inferomarginals, so that there is no danger of confusing the two forms in a super-
ficial examination.

Type.—Cat. No. 21926, U.S.N.M.

Type-locality.— Albatross station 3310, vicinity of Unalaska, Alaska, 58 fathoms,
on fine dark sand and mud.

Distribution —Bering Sea (vicinity of Pribilof Islands and west of St. Paul)
to southeast Alaska, and off Monterey Bay, California; on the Asiatic side to the
Sea of Japan. Bathymetrical range, 32 to 688 fathoms in Bering Sea and northern
part of range, to 871 fathoms off Monterey Bay. Found on fine gray or black
sand, green mud, or on pebbles.

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

or
bo

Specimens exami ned.—The following is a complete list of localities from which
one hundred and seventy-eight specimens have been examined:

Specimens of Leptychaster anomalus examined.

| Number
Station. Locality. Depth. Nature of bottom. of specl- Collection.
mens.
Fathoms.
fy eee Vicinity of Shumagin Islands, Alaska... 48 fine gray sand............-.- 1 U.S. Nat. Mus.
110 green mud..........-------- 13 Do.
GO) eiioe ss GO kisses aoaiseces cecl= tne ose 3 Do.
58 black sand.......-...------- 2 Do.
SRDAL coace Off Kadiak Island, Alaska........--..-- BOM aissen 0 pereteseey ot cee nme tela 1 Do.
BRED once Off Sitkalidak Island, vicinity Kadlak 69 green mud...........-.-..-- ; 10 Do.
Island, Alaska.
| Bering Sea, vicinity of Unalaska, Alaska. 56 black pebbles.......-..----- 1 Do.
North of Unimak Island, Alaska... ....- 121 black sand and gravel.....-.- 1 Do.
81 gray sand and gravel.......- 2 Do.
61 black mud..........-- Bear 12 Do.
58 fine dark sand and mud...-. 15 Do.
85 green MU ca ese ses eaae eee 28 Do.
68 fine black sand.........----- 50 Do.
56 mud and sand........-.---- 1 Do.
3486...... West of St. Paul Island, Bering Sea... . 15 green mud, fine sand_....--. 2 Do.
SASS saceuniacene OO gece fone daceee ss ete osecneee on See 106 green mud, gray sand......- 5 Do.
$501...... South of St. George Island, Bering Sea... 6838 green Mud eacce-n- or seonetee 1 Do. -
3548...... Near Unimak, Aleutians.........-..... 91 Dplacksimud!! =e. eeeee ees 1 Do.
$602...... Bering Sea, vicinity of Pribilof Island... 81 green mud, sand........---- 1 Do.
Shumagin Islands, Alaska.........-.--- SSM peeaceee eres eee Seeman ate 1 Do.
UO NSASKA seetes seocin e+ ose ee ee 80 SANG's=-oceee eee eee eens 1 U.S. Nat. Mus., W. H.
Dall, No. 6073.
| Nollocality 0 ocean en saee as seaeens a2 |a--ae aon 9 U.S. Nat. Mus.
4230... | Vicinity of Naha Bay, Behm Canal, | 240-108 3 Albatross, 1903.
southeastern Alaska. |
4233...... Vicinity of Yes Bay, Behm Canal ...... 39-45 | gray mud, rocky..........-. 1 Do.
4285.05: Off Sitka Sound, Alaska............... » 590 green mud, rocky...... 1 Do.
4280...... Chignik Bay, Kadiak Island............ 32 green mud, black sand. 2 Do.
S206 2. scl eeu se OD scaes tee nee a seta snes eee ae 57-63 green mud, rocks.... 2 Do.
4538...... | Off Monterey Bay, California...........| 871-795 | gray sand and rocks... 1 Albatross, 1994.
B1BB 2s c555 | Bowers Bank, Bering Sea....-......... 764 greenish brown mud. 3 Albatross, 1906.
AT bis ac [enntO sane cesta tae een eee 584 | green mud......... 1 Do.
AISA sk | Vicinity of Attu Island, Aleutians ..... 135 coarse pebbles. . . 1 Do.
Cb ee Sea of Japan (38° 08’ N.; 138° 31’ E.).... 225 fine brown mud. 1 Do.
4867.2... BOR Of AB PAN ee weet ee ee oad oe 150 green-Mud's sco. ane aasiese 1 Do.
2 Bottles broken by earthquake, 1906. b Not typical.

Remarks.—The reasons which may be advanced for placing this species under
the genus Leptychaster are the following: Leptychaster arcticus, which ranges into
Bering Sea, has in the same region a close relative, pacificus, which differs in
possessing broader and less numerous marginal plates. The superomarginals of
arcticus are throughout the ray scarcely larger than paxille, but are of conspicuous
size in pacificus, and in the interbrachial region are considerably wider than midway
along ray. After a close comparison of the two forms it has not been possible
to separate them generically. The character of the marginals, abactinal paxille,
adambulacral plates and armature, distribution of gonads, and distribution of
papulx are essentially the same in the two forms. The width of the superomar-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 53

ginals remains the most conspicuous difference. By the same reasoning it is impos-
sible to separate generically anomalus from pacificus. In the former the width of
the superomarginals has been found to vary considerably; hence if this character
is unstable for a species it obviously can not be used to diagnose a genus. The
wholly unarmed marginals remain, however, very characteristic of the genus.
Thus L. arcticus and L. anomalus represent two extremes, with pacificus in between.
L. propinquus, described below, somewhat resembles L. kerguelenensis when viewed
from above, and still more Mimaster cognatus, although the interbrachial ares are
rounder and the inferomarginals wider in proportion to length. As noted under
Gephyreaster swifti, in view of the range of variation within the genus Leptychaster,
it is very doubtful if Mimaster cognatus is a Mimaster. It seems to be a Lepty-
chaster.

Leptychaster anomalus bears a striking resemblance to Parastropecten inermis
Ludwig, which appears to be based on young specimens, however. From this
species anomalus differs in having three or four furrow spinelets instead of six or
seven, in having five or six papule about each plate or paxilla instead of four,
and in having a minute anal opening.

The diagnosis given by Ludwig for Parastropecten is as follows: Die neue
Gattung unterscheidet sich von Astropecten durch den vélligen Mangel von unteren
und oberen Randstacheln, durch kurze Arme und durch verhaltnismiissig grosse
ventrale Interradialfelder, deren Platten (=Ventrolateralplatten) eine paxilliire
Form haben. In other words, the genus is erected on the strength of the unarmed
marginals and large actinal areas—just the features in which Leptychaster differs
from Astropecten, for the size of the superomarginals has been shown to be variable
within a species, and not diagnostic of the genus.

The impossibility of recognizing a separate genus Parastropecten for inermis
and anomalus is the fact that Leptychaster propinquus would have to be ranked
under Leptychaster on account of the small superomarginals, L. anomalus, of course,
under Parastropecten, L. pacificus probably under Parastropecten, on account of
the larger superomarginals, although it is obviously a close relative of L. arcticus.
To divide the genera on the presence or absence of an anus would lead to the
ranking of Parastropecten inermis and Leptychaster arcticus under one genus and
the other forms under another, and would necessitate ignoring the very characters
upon which we base genera.

Consequently, although Parastropecten seemed fully warranted when described,
it seems best now to merge it with Leptychaster.

The genus Glyphaster Verrill based (without diagnosis) on L. anomalus seems
to be synonymous with Parastropecten, which would therefore have precedence if
anomalus constituted a separate genus. Since there are intergrading forms with
arcticus, I have no hesitation in saying that neither Glyphaster nor Parastropecten
can be so diagnosed as to include Leptychaster propinquus and exclude typical
Leptychaster. .

I have examined two small specimens from the Sea of Japan, collected by the
Albatross in 1906. If the tiny specimen from off Monterey is certainly anomalus,

a Mem. Mus. Comp. Zodl., vol. 32, 1905, p. 76.

54 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

the species probably has a continuous distribution by way of Bering Sea. It is

on . e a
of interest to note that the Japanese examples are fairly typical. Both of them
have epiproctal cones, and in one the tiny anal pore is visible (in sunlight) with a
strong glass.

LEPTYCHASTER PROPINQUUS Fisher.
P29) fig. 3:
Leptychaster propinquus FIsHER, Zool. Anz., vol. 35, March 29, 1910, p. 545.

Diagnosis.—Similar in general form to L. anomalus, but with narrower and
more numerous marginal plates, less massive margin to disk, and smaller paxille.
R=29 mm.; r=15 mm.; R=1.93 r. Width of ray at base, 17.5 mm.

Description —This form is obviously such a close relative of anomalus that
the description will be largely a comparison with that species. Paxille arranged
as in anomalus, but smaller and less compactly placed, and with longer spinelets.
The paxillar area has the appearance of being more open and the outlines of the
paxillse are less well defined. While a large paxilla of anomalus has forty to forty-
five spinelets, a similarly placed one in propinguus has but about twenty-five, the
peripheral series containing about fifteen. The spinelets themselves are longer
than in anomalus, especially the centrally situated ones. Distribution of papules
and form of abactinal plates practically the same as in anomalus. Plates show a
slight irregular lobing in some specimens; in others, none at all.

Superomarginal plates are nearly quadrate, except in interbrachial angle,
where the width is slightly greater than length, and they are only one-half or
slightly over one-half as wide as in equal sized specimens of anomalus, thus appear-
ing more typical of the genus Leptychaster. The superomarginals are, moreover,
more uniform in width throughout the ray than in anomalus. Superomarginals,
twenty-one in number; in type of anomalus, which is practically equal sized,
fifteen. Spinelets covering marginals similar to those of anomalus, but, as in the
case of paxille, a trifle longer.

Inferomarginals shorter and narrower than in anomalus, corresponding in
number and position to superomarginals, the covering being a trifle but not much
longer than in anomalus. As compared with arcticus the inferomarginals encroach
more onto actinal surface and are placed less obliquely to transverse axis or plane
of ray. (The form bears no particular resemblance to pacificus, this species having
large superomarginals.)

Adambulacral plates and spinulation not differing materially from anomalus.
Furrow spinelets three or four, generally three; usually four in anomalus. The
spinelets on actinal surface of plate are very similar if not identical in arrangement
with those of anomalus, but are a trifle longer. Actinal interradial areas vary
slightly in size and the paxille are a trifle smaller, and the spinelets slightly longer
and slenderer than in anomalus. The mouth plates and armature do not differ
in any important respect from those of anomalus.

Variations.—The essential features which distinguish this species are well
marked in all of the fifty-seven specimens, and there is very little variation of
consequence. In other words, all have small paxille and numerous and narrower
marginal plates. The smaller specimens naturally have shorter rays, and there is
a slight variation in the width of the superomarginals and inferomarginals.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER,. 55

Young.—The smallest specimen has R=7 mm., and is arcuate-pentagonal
in shape. When compared with very young anomalus, the latter are seen to have
fewer and much more massive marginal plates. The other differences between the
adults are not so obvious but occur in a slight degree.

Type.—Cat. No. 24338, U.S.N.M.

Type-locality.— Albatross station 4792, vicinity of Commander Islands, Bering
Sea, 72 fathoms, pebbles.

Distribution —The plateau upon which the Commander Islands are situated.
Bathymetrical range 54 to 72 fathoms.

Specimens examined.—Fifty-seven from three stations, all in the vicinity of
Commander Islands: 4787, 54 fathoms, green sand, eleven specimens; 4788, 57
fathoms, green sand, forty-five specimens; 4792 (type-locality) one specimen.
Taken by the steamer Albatross, June 14, 1906.

Remarks.—The only specimens among those assigned to anomalus which
approach dangerously near the present species are six, which are unfortunately with-
out locality. These have considerably narrower and more numerous superomar-
ginals than typical examples, and more numerous inferomarginals, but there is no
difficulty in distinguishing them at a glance from propinquus. When placed next to
a series of the latter, their superomarginals are evidently much more conspicuous,
especially in the interradial angle, while the inferomarginals are neither so numer-
ous nor so short. Although the general form of the two is so nearly the same, the
spinulation of the aberrant anomalus is typical, and not like that of propinguus.

I was at first inclined to regard this form as a race of anomalus, but in the absence
of intergrades have classified it as a distinct species. It combines characteristics
of typical Leptychaster and the synonymous genera Parastropecten and Glyphaster.

Genus ASTROPECTEN Gray.

Astropecten Gray (from Linck, 1733) Ann. and Mag. Nat. Hist., vol. 6, Nov., 1840, p. 180.
Type A. aurantiacus, designated for the first time, 1908.—Fisurer, Smiths. Misc. Coll., vol.
52, 1908, p. 93.

Stellaria Narpo, De Asteriis, Oken’s Isis, 1834, p. 716. Type S. aurantiaca. Invalidated by
Stellaria Meller.

Asterias Acassiz (not Linnzeus), Mém. soc. sci. nat. Neuchatel, vol. 1, 1835, p. 168.

Crenaster D’ORBIGNY, Prodrome de paléontologie, 1850, vol. 1, p. 240.

Diagnosis.—Rays normally five in number; abactinal surface flat, not arched;
actinal surface slightly beveled on sides; disk variable in size, usually medium
to small; rays usually long and tapered; marginal plates large, the inferomar-
ginals always broader than superomarginals and sometimes extending laterally
beyond them; inferomarginals armed with spinelets and a variable number of
spines which increase in size toward the edge of ray; superomarginals, in addition
to small granules or spinelets, may also bear tubercles or enlarged spines extending
in one or two complete or interrupted rows along ray; or enlarged spines may
be entirely absent; exposed surface of consecutive marginal plates separated
by deep fasciolar grooves lined by minute capillary spinelets, these grooves acting
as percolaters or filters; abactinal area covered with true paxille; papule single,
usually absent from a narrower or wider midradial line, and from center of disk;

56 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

usually six about each paxilla; actinal interradial areas typically very small,
with few paxilliform intermediate plates which do not extend far along ray, the
inferomarginals and adambulacrals being in contact on the ray proper; inter-
mediate plates never extend beyond middle of ray and rarely beyond proximal
fourth; ambulacral plates with an angular furrow margin, bearing typically three
spines, of which the middle is slightly the longer; two or three rows of spines
on actinal surface of plates; first adambulacral plate compressed, much wider
than the rest; mouth plates narrow, with inner spines enlarged ; tube feet conical,
without a true sucking disk; no true pedicellarie; anus typically absent; gonads
interradial, not extending along ray; superambulacral plates well developed.

KEY TO THE SPECIES OF ASTROPECTEN HEREIN DESCRIBED.

a, A more or less complete series of spines along upper edge of superomarginals, and frequently also
a second parallel longitudinal series, spaced from the above; central granules of paxillee usually
considerably enlarged on disk; enlarged adambulacral spine usually abruptly truncate and
often sulcate.....-..-------- +--+ 2222 ee ee eee eee ee eee eee eee eres Soo ga ee a rmatus, Dp. 56.

a®, Superomarginal plates entirely devoid of enlarged tubercles or spines; central granules or spinelets
of paxillse not enlarged on disk.

b'. Paxille smaller, about four or five transverse series opposite two superomarginals at base of ray;
more or less irregular along radial lines; abactinal plates of papular areas of rays not lobed;
upper edge of ambulacral ossicles not produced into a thin lamina between consecutive ampul-
lee, and bearing a comb-like series of irregular teeth or spinelets.......-..-. californicus, p. 61.

b?, Paxillee larger, about three transverse series opposite two superomarginals at base of ray; paxille
uniform in arrangement across ray; abactinal plates of papular areas very distinctly lobed;
upper edge of ambulacral ossicles curiously compressed and produced into a thin lamina,
bearing a few irregular teeth or spinelets; upper end of same ossicles toothed . .ornatissimus, p. 67.

ASTROPECTEN ARMATUS Gray.
Pl. 5, figs. 1-2; pl. 7, figs. 3, 6; pl. 50, fig. 4; pl. 51, fig. 3.

Astropecten armatus Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 181 (not Miller and Troschel,
Syst. Ast., 1842); Synopsis Starf. Brit. Mus., 1866, p. 3—VeERRLL, Trans. Conn. Acad.,
vol. 1, pt. 2, 1867, p. 332.

Astropecten erinaceus Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 182; Synopsis, 1866, p.
3.—VerriLL, Trans. Conn. Acad., vol. 1, pt. 2, 1867, p. 332—Prrrier, Révision des
Stellérides, 1875, p. 358.—StapeNn, Challenger Asteroidea, 1889, p. 734.—FisHErR, Proc.
Wash. Acad. Sci., vol. 8, 1906, p. 118.

Astropecten Orstedii Liirken, Vid. Medd., 1859, p. 159.—VeErrit, Trans. Conn. Acad., vol. 1, pt.
2, 1867, pp. 274, 328, 330.

Diagnosis..—Size frequently large. Rays five. R=129 mm.; r=30 mm.;
R=4.3r. Breadth of ray at base, 32mm. Species variable in relative dimensions
and in ornamentation. Sides of ray rather high, vertical. Marginal plates massive,
the superomarginals encroaching more or less onto abactinal area, and usually
armed with one or two short stubby spines, placed one above the other and forming
two longitudinal rows of spines along ray except at base, where they may be as
many as four spines to a plate, and at tip, where there is usually but one. Great
variation in this respect; frequently only one series of spines, or rarely only a few
scattered. Paxillew fairly large, the central granules or spinelets larger than

@From a specimen in the Stanford collection, collected by E. C. Starks at San Diego, California,
figured on pl. 5, fig. 1.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 57

the peripheral. About three prominent lateral inferomarginal spines in an oblique
series, with numerous accessory spinules on actinal surface. Adambulacral plates
with three furrow spines, and on actinal surface a longitudinal series of two spines,
the adoral small, somewhat flattened, the other greatly enlarged, subterete at base
but much flattened beyond middle, and chisel-shaped or decidedly scoop-shaped
at tip. Out side of this series, three or four spines, about equal to adoral member
of above series, form an oblique irregular row or a group.

Description —Paxille fairly large, those on disk and along proximal radial
regions with spinelets very robust, short, globose, or with tips flattened and flaring.
In large specimens two or more central spinelets are frequently fused together to
form a very irregular plate or ossicle occupying the whole or a part of the tabulum
of paxilla. Peripheral spinelets or paxilla robust, but smaller than the central
ones; more uniform in size in small than in large specimens where they are of
unequal caliber, often enlarged at tips and ten to fifteen in number. Paxillx
largest on radial regions at base of ray, decreasing in size toward margin of area,
center of disk, as well as toward tip of ray.

When the abactinal integument is removed and examined from the cclomic
side the midradial areas are seen to be paved with irregularly arranged, subcircular,
elliptical, or even faintly scalloped plates without papule between them. The
areas occupied by these plates broaden at base of rays, and the central area of disk
is also paved with similar ossicles, decreasing in size and overlapping more and
more toward the very center. On the rays, the nonpapular areas are from three
to four plates wide. On either side of this midradial area the plates are very
much smaller, and at base of ray are not very regularly substellate or six-lobed,
but soon become elongate as in the following species. The plates do not touch
and are surrounded by six or five papule. On the interradial papular areas of
disk the plates are stellate and usually touch. The connective tissue and muscle
layer between the ossicles and ccelomic epithelium is thicker than in either of the
following species. The arrangement of plates surrounding madreporic canal
resembles that of A. californicus.

Superomarginal plates very massive and high, especially interradially, occu-
pying two-thirds or three-fourths of the vertical side of ray, and encroaching
conspicuously onto abactinal surface, about which they form a slightly raised
border. There is a large degree of variation both in the size and amount of encroach-
ment onto abactinal paxillar area. Variability in the latter respect affects the
width of the paxillar area and considerably alters the general facies of the specimen.
A large specimen? (Albatross, 1904) from San Diego Bay (Beacon 3 Shoal) has a
row of small tubercles, one to a plate, all along the upper edge of series, extending
about two-thirds the length of ray. These tubercles are very irregular in size and
shape, sometimes becoming reniform or even cordate. Not quite half way to lower
end of each plate is another small tubercle forming a similar but less conspicuous
longitudinal series. Another large specimen, that from which the diagnosis is
taken, has both series of tubercles well developed, the upper series extending
nearly to tip of ray, the lower within two plates. Spines of lower series conical

“See pl. 5, fig. 2.

58 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

or bluntly bi- or tri-dentate at tip; those of upper series shorter, more irregular,
often flattened, and apparently composed of two fused tubercles; sometimes two
subequal or very unequal tubercles stand side by side at upper end of plate. On
first plate three tubercles (shaped something like molar teeth) stand in a transverse
row at upper end of plate, and below them two or three short conical spines, the
lowermost in line with the lower longitudinal series. Scarcely two proximal plates
are alike in armature. The opposite extreme is found in a large specimen from
off San Pedro in which there is an inconspicuous tubercle at the upper end of a
few of the proximal plates, the rest being unarmed. Any number of intermediate
forms may be found between these extremes. General surface of plates covered with
spinelets, which are flat, roundish, or polygonal and granuliform on the abactinal
surface of ray, becoming slender and capillary in the fasciolar grooves and toward
lower edge of plate.

Inferomarginals extending laterally slightly beyond superomarginals, or not
at all. Armature consisting of about three or four tapering, slender, sharp, slightly
flattened spines in an oblique series on outer end of plate, and two or three shorter,
sharp, lanceolate spines on the aboral edge of the actinal surface of each plate.
Numerous shorter, lanceolate or blunt spinules usually accompany the regular
series. The actinal spines are broader, flatter, and more blade-like in some speci-
mens than others and the larger lateral spines vary considerably in length in pro-
portion to width of ray (one-fifth to one-half width), while some are slenderer
and less obviously flattened than others. General surface of plates is covered with
flattened, round-tipped, squamiform spinelets.

Adambulacral plates with an angular furrow margin. Furrow spines three,
of the usual astropectinoid type; central spine slightly longer than the two laterals,
compressed, blunt; lateral spines of series flattened, tapering, blunt. On actinal
surface, just behind furrow series, is a longitudinal series of two spines, the adoral
somewhat flattened, equal to or shorter than adjacent furrow spine; the aboral
spine greatly enlarged, subterete at base but much flattened beyond middle and
chisel-shaped or decidedly scoop-shaped at tip. In large specimens this spine
is 4 to 6 mm. long and 2 mm. broad at tip. Outside of this series, three or four
flattened spines, about equal to adoral member of above series in length, form
an oblique, irregular, or a longitudinal row, or a group. Sometimes at base of ray
the aboral member of the outer series is also enlarged, but it never attains the size
of the large spine. The great size of the enlarged spine often pushes the adoral
member out of place, usually toward the furrow.

Mouth plates prominent actinally and rather narrow. Armature variable.
In large specimens there is usually a single series of prominent, blunt, or truncate,
flat spines parallel with suture between the two plates and increasing in size toward
the inner end of plate. On free margin of plate is an irregular series of much
smaller spines, and between the two numerous very short subconical spinelets.
Sometimes a second superficial series, smaller than the regular one, closely parallels
it. The first adambulacral plate is much compressed and bears two closely appressed
ee series of truncate spines, fairly regular in small examples, but irregular
in large.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 59

Actinal interradial areas very small, paved with eight to twelve paxilliform
plates, sometimes even fewer, the spinelets being similar to those on adjacent parts
of the inferomarginals—i. e., slightly tapering, terete or flattened at tip, and skin
covered. In specimens with very spiny superomarginals, the actinal intermediate
paxillz usually bear one or occasionally two stout tapering spines, surrounded by
the smaller spinelets. Many specimens do not have this spine. Actinal inter-
mediate spines are strongly developed in a large specimen from San Diego, from
which the diagnosis is drawn.

Superambulacral plates well developed. Under edge of ambulacral ossicles
normal, and smooth, not produced into a thin lamina between consecutive ampulle,
nor toothed in any way. Gonads axillary. Six Polian vesicles, two in the inter-
radius containing the madreporic canal (one on either side of it),and one in each
of the other four interradii.

Madreporic body large, exposed, 5 to 6 mm. in diameter (in big examples),
the adcentral edge about midway between center of disk and edge of marginal
plates. Striations very irregular often with tiny protuberances along ridges, these
sometimes fusing after the manner of the central granules of the paxille.

Variations.—This species, as has already been said, varies greatly, not only
in respect to many details of ornamentation but also as regards dimensions, the
relative size of rays and disk, and the width of paxillar area. It is not surprising,
therefore, that the general appearance should also differ considerably, according
as the rays are long or short, or the superomarginals encroach considerably upon
the paxillar area, or as the latter are very spiny or not, or as the inferomarginals
extend laterally beyond the superomarginals. If one encountered the isolated
extremes he would certainly be justified in regarding them as examples of distinct
species. It is not at all improbable that several species now regarded as distinct
may really belong to this variable species. The character of the paxille and of
the inferomarginal armature appears to be more stable than that of the marginal
plates and is hence more valuable in an offhand determination of specimens.

A few specimens have been analyzed to show some of the more obvious differ-
ences or variations.

Table showing variations in Astropecten armatus.

| |
‘ | Number | a a | Width of
. Rin rin of supero-| Upper row of superomar- = .| paxillar
Locality. mm. | mm ue mar- Pinal spines present. EO et LO Bresette Lee eel rea, base
ginals. |omarginals.| of ray.
129 30 4.3:1 36 x ie NR Raters 27
150 30 5:1 42 G0 ee | ee tanerere cabehis se Capen ticle [execpt 21
103 20 5.1:1 | 40 1 or 2spines>........... Scattered spines....... x 15
108 | 25 4.3:1| 38 | 3 or4spines>. Bee Xae res 18
2| 18 3.4:1 30 5or 6spinesb...........| Xac Trifle. 7
105 23 4.6:1 | 39 SOn4 CU DSECled Os aet as ao mcch dna giee le kon iae mates x 20
32 9 3.5:1] 25 Lispine'v:.:.cc-esteacece Xac Trifle 5
———E —_— | — — — ———_—__—_—__—__ — ——____
a Absent from terminal fifth of ray. » At base of ray. ¢ Absent from first five or six plates.

The smallest specimen of this species I have seen has R=17 mm. and a few
enlarged superomarginal tubercles. The central spinelet of paxille is granuliform

60 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

and subglobular and the enlarged adambulacral spine is square cut at base of ray.
These features will serve to distinguish small specimens from young of californicus
and ornatissimus, in which the central granules of paxille is not more robust than
the peripheral ones, and where there are never superomarginal tubercles or spines.

Type-locality.—‘ Puerto Portrero, South America, on sandy bottoms, 9 fathoms,
H. Cuming, Esq.’’ (Gray).

Distribution.—Ranges from San Pedro, California (and probably from north
of this point) to Ecuador (Punta St. Elena, Puerto Portrero). Only Californian
specimens have been examined, however. Shore to about 30 fathoms.

Specimens examined.—Thirty-four from San Pedro and San Diego, as follows:

Specimens of Astropecten armatus examined.

Locality. Depth. Nature of bottom. | Number. * Collection.
eae |
Long Beach, California............~...-.---------- 1 U.S. Nat. Mus.
San Diego, California...... Shore. | 1 Stanford University.
San Diego, California, Beacon 3 Shoal. ........-.- : | 1 Albatross, 1904.
San Pedro, California (II-2)¢....................--- 2h 1 University of California.
San Pedro, California (ITI-1)...............-...---- 7 dark-brown mud.....- 7 Do.
San Pedro, California (VI-1)...........-.--.------- li shore'sand ==; =--=----- | 6 Do.
San Pedro, California (XV-1) aeons 4d |(coarse'sand.....=..---- 7 Do.
San Pedro, California (XVI-3)......-.....------ ee arg ihes eR oe eens 4 Do.
San Pedro, California (XIX-2)..................... 102-30 | soft sandy mud.......| 1 Do.
ING localltviss aoe eh eenee ence c sont ae ae Oe ene ae eee | 5 Do.

@ Station and haul, University of California Series.

Remarks.—This species is so variable that it is difficult to make its positive
characters intelligible through the median of description. It bears a great resem-
blance to Astropecten duplicatus Gray, from which it differs in having broader rays
relative to the length, heavier paxillar granules, much stouter and longer infero-
marginal spines, and more conspicuously enlarged subambulacral spines. In
duplicatus the superomarginal plates encroach more onto the paxillar area (width
equals more than one-half paxillar area) and are not nearly so high as in armatus,
the ray being more depressed.

Armatus may be distinguished from the other two Californian astropectens by
the presence of at least a few superomarginal spines or tubercles, by the broader
and more tumid superomarginal plates, by the enlarged central granules of the
paxillwe, by the heavier and more bristling inferomarginal armature, and especially
by the chisel-shaped enlarged adambulacral spine, which is broader at tip than
at base, and usually more or less hollowed at the end, on the upper (or outer) side,
like a gouge. Even young specimens of armatus have a few superomarginal tuber-
cles, although these are usually inconspicuous. In lieu of the superomarginal
spines the enlarged central granules of paxille and the specialized adambulacral
spine may be used to determine doubtful specimens.

The name erinaceus has usually been used for this species. Perrier considers
Litken’s Grstedii and Gray’s armatus to be identical with ervnaceus, and in view
of the great variability of the species, he is probably correct. The name armatus,
however, has page priority over erinaceus. The Astropecten armatus of Miller

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 61

and Troschel, 1842, is a homonym of this species and a synonym, according to
most authors, of A. polyacanthus of the same writers. It consequently does not
affect the validity of Gray’s name, which has two years priority.

ASTROPECTEN CALIFORNICUS Fisher.
Pl. 6, figs. 1, 2; pl. 7, fig. 1; pl. 50, fig. 5; pl. 51, figs. 2, 2a.
Astropecten californicus FisHER, Zool. Anz., vol. 30, June 19, 1906, p. 299.

Diagnosis.-—Rays five. R=100 mm.; r=18 mm.; R=5.5 r. Breadth of
ray at base, 20 mm. (measured between first and second superomarginal plates).

Disk small; rays long and narrow, pointed; interbrachial angles rounded;
abactinal surface plane or arched, bordered by narrow margin formed of supero-
marginal plates which are confined chiefly to side wall of arm; superomarginals
without enlarged spines or tubercles; inferomarginals rather narrow with trans-
verse aboral series of about three spines on edge of ray, continued actinad along
aboral edge of plate in two to four smaller-spaced spinules; adambulacral spines
im about three parallel longitudinal series, the aboral spine of the middle series of
two or three conspicuously flattened, round-tipped and stouter than others (see
p. 62); paxille small to medium sized, about four or five transverse series, corre-
sponding to two superomarginals at base of ray, six or seven at middle and about
eight to ten near tip. Upper edge of ambulacral ossicles smooth.

Description—Abactinal paxille fairly uniform in arrangement, and usually
compact, largest on outer half of radius of disk, decreasing in size toward center
of disk, very quickly toward margin, and gradually along rays toward tips; arranged
in definite transverse rows along sides of paxillar area of rays, elsewhere without
regularity. Crowns of paxille circular or broadly elliptical when spinelets are
radiating, more irregular when spinelets stand erect and compact. Each paxilla
pedicel surmounted (in larger paxille) by fifteen to twenty short, stout, round-tipped
or subtruncate, often clavate spinelets in a peripheral series, and eight to fifteen in
center. When the spinelets are crowded together the paxilla is flat-topped and
the tips. of the spinelets resemble flat-topped granules. In many specimens the
spinelets are slender and stand on the pedicel more like the petals of a flower.
These specimens usually have the paxillee more definitely spaced, with fewer spine-
lets—ten to fifteen in peripheral and five to eight in the central group. Opposite
the suture between second and third superomarginal plates about eighteen to
twenty paxille can be counted across ray to similar point on opposite side.

The bases of the paxillz or abactinal plates are subcircular and slightly over-
lapping, without papule between, along midradial areas and central portion of
disk. Along the ray this radial area of roundish plates is about five plates wide and
is broader than the same region in A. ornatissimus. The plates are irregularly
arranged, without evident spaces between except occasionally on disk. On either
side of the radial areas the plates are very regularly arranged in transverse rows,
are six-sided but much longer than wide; far along ray they become rudely elliptical,
with truncate ends, and on disk broader than on base of ray. Faint lobes are

2

aDiagnosis and description based on specimen from station 4559, Monterey Bay, California; 2
to 8 fathoms, fine gray sand.

62 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
sometimes seen, but the plates are not distinctly lobed as in ornatissimus. In these
lateral regular series two papuls occur between every two plates, each plate being
surrounded by six. On the disk the papule are confined to the interradial angles.
A very large plate stands on the adcentral side of the madreporic body, and on
either side of this a slightly smaller one, while four others, much smaller, complete
the circle. eee ;

Superomarginal plates forty-six or forty-seven from interradial line to extremity
of ray are wider than long, the length increasing, the width decreasing toward end
of ray; and they gradually encroach more and more onto abactinal surface as they
proceed along ray. Plates form an arched bevel to distal half of abactinal edge of
ray, and a rounded lateral margin to ray in proximal portion. Plates slightly tumid
along aboral edge, and covered with low granuliform spinelets squamiform in middle
of upper half, thence decreasing in size and becoming slender toward margins of
plate. No enlarged spinelets or tubercles on any of the superomarginal plates.

Inferomarginals corresponding exactly to superomarginals, but not extending
laterally beyond them except a very little, sometimes, near end of ray. Armature
consists of a series of five or six slender, tapering, sharp, slightly curved spines and
spinelets along aboral edge of plate; of these, three stand on outer end of plate, form-
ing a triple series all along margin of ray, the middle spine being longest of the
three and about equaling in length the width of plate; on actinal surface two or
three spaced, sharp, slightly flattened, much shorter spines continue this series
to inner end of plate. Adorally to the three marginal spines and close to them,
four or five sharp, flattened spinules (shorter than inner spines of above series)
form an oblique row meeting the aboral series just below the inner spine of marginal
row (which is also set transversely oblique). There may be four spines in the mar-
ginal series, and usually on the first four plates the largest spine is broader, flattened
and lanceolate or oblong-lanceolate in shape. It is here only about one-half as
long as width of plate. General surface of plates covered with flattened, round-
tipped, upright, papilliform spinelets, increasing slightly in size toward the auxiliary
lateral spinules.

Adambulacral armature in three or four fairly regular longitudinal series. (1)
Furrow spines three, the central longest and saber-shaped when dry, the laterals
shorter and slender when dry, but in life with more or less evident membranous expan-
sions at base, giving the spine a flattened appearance. (2) First actinal series consist-
ing of two or three slightly shorter spines, of which the aboral is very much the
stouter, being flattened with a rounded or truncate tip. The adoral spine is shorter
and when there are three in the series they are graduated in length. (3) On outer half
of plate are two to five shorter slightly flattened spines either in an irregular group,
or forming a series of two or a series of three, with one out of line in addition, or
sometimes two series of two. In large specimens the armature is less regular.
Sometimes a series of two smaller spines is interpolated between the furrow series
and the first actinal, or again only one such spine. Irregularity is frequent on
plates near the mouth, but usually the larger flattened spine is recognizable on the
actinal surface. First adambulacral much compressed, the second less so; former
with a double transverse series of closely appressed spines, about 30 in number.

Combined mouth plates narrow with a crowded armature consisting of three
series, a marginal, rather high on side of furrow, an intermediate, and most con-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 63

spicuous of all, a superficial, the latter consisting of a nearly straight series of four-
teen or fifteen short flattened, round or square-tipped, or subacute spines, increasing
rapidly in length toward inner angle of plate, where there is an enlarged tooth (or two
to each mouth angle). Marginal spines small, forming an angular series of seven
or eight, between base of tooth and inner end of first adambulacral plate, thence
continued along side of first interadambulacral suture in the form of a few strag-
gling very small spinelets. Intermediate series extending whole length of plate,
parallel and close to superficial series, against which the much smaller spinelets are
often appressed. All spines more acute, and somewhat slenderer in dried specimens.

Actinal interradial areas very small, having six small paxilla to each area,
one opposite each of the first three adambulacrals. Each paxilla bears several
(four to seven) closely grouped, clavate skin-covered, papilliform spinelets. Fewer
than six plates may be present.

Superambulacral plates well developed. The upper edge of the ambulacral
ossicles, though sharp, is not produced upward into a thin lamina provided with irreg-
ular comb-like teeth as in the following species. This fact will serve to distinguish
californicus at a glance. A Polian vesicle in each interradius except that containing
the madreporic canal, which has two; the same interradius has two septa. Gonads
axillary, not extending along rays.

Madreporic body partially concealed by paxille, situated about one-third
distance from margin to center of disk; striations deep, irregularly centrifugal;
ridges with numerous little knobs.

Color in life, ferruginous to light claret brown above, lighter below.

Variations.—The chief variation in this species is due to the relative size of
specimens, the large examples having long, narrow rays, the small ones having
shorter thicker rays. As fully seven-eighths of the specimens from off the southern
part of California are small, and the maj ority of those from Monterey Bay are large
it would seem, if one had only a few examples, that the southern form had shorter
rays. Several good-sized specimens collected by the Albatross in 1889 off the
Santa Barbara Islands have rays fully as long, however, as average specimens from
Monterey Bay, while small examples from the latter locality are of about the same
dimensions as equal-sized examples from southern California. It is possible that
the region about Monterey Bay is more favorable for the species and that it there-
fore develops to larger size than in southern waters.

Specimens from Monterey Bay of nearly the same size vary in respect to the
length and width of ray, some specimens having narrower, or more pointed rays
than others, but as shown in the accompanying table these differences are not great.

Table showing variations in Astropecten californicus.

!
gern eas Number | width | R is to
Station. Locality. mm, | mm: Riv peromar- of ray | breadth
ginals. ee base.| of ray as:
| |
| 4559 | Monterey Bay.........--.-- 10 | 180] 5.5: 46 | 20 | 5:
4459/2). doses eee 94 | 16.0 | 5.87:1 wo | 7 5. 53:1
MaDa NINN Ded oucee Cee Fe 94 17.0 | 5.53: 5 | 47 5. 53:1
7 eee dO: See eee 3 | 13.5 | essa 51 15 6.2:1
dO we syest eee eee 9 | 19.0 | 4.781 43 20 4. 5:1
| 2975 Off Santa Cruz Island....... 79 | 13.5 5.121 45 |

15 5. 33:1

64 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
Variations due to size are shown in the following table:

Table showing variations due to size in Astropecten californicus.

‘ ; Number Width | Risto
ee a fat | R= atte aS
| pors e 3 q 7 =m

4420 Monterey Bay.............. 11 4.5 2.44:1 | 14 5 2.21
$420 edges, ete eee 21 8.0 | 2.62:1 20 7 | 3:1
{ido P11 Sido: 2.22. 5: ee esata e0c ln | tote eer 10 3.2:1
| 4535 S00 moan nee omen -| 52 14.0 | 37:1 | 132 14 | 3. 7:1
| 4. 4:1

| 4453 pdQre.-+- Semeenae ease 70 15.0 4.7:1 46 16

Besides variations in dimensions, some specimens have the superomarginal
plates encroaching onto abactinal surface more than others, especially at base of
ray; and the paxille vary slightly in specimens of the same size. Some examples
have the paxillee more crowded and their spinelets may be shorter and more granuli-
form, as well as thicker in caliber. This causes the paxilla itself to appear more
compact and alters somewhat the general facies of the specimen. In specimens
with slender paxillar spinelets the abactinal area is more open on the rays. The
flattened granules on the superomarginal plates are coarser in some specimens than
others. The outer actinal adambulacral spines vary somewhat in arrangement,
but the enlarged spine is always flattened in a longitudinal plane and has a round
or subtruncate tip.

Young.—The young of course have shorter arms, fewer marginal plates, and
smaller paxille. In a specimen with R=20 mm. (r=6 mm.; R=3.3 r) there are
four or five transverse rows of abactinal paxille opposite every two superomarginals
at base of ray, and about twelve to fifteen paxille in a transverse row opposite
second superomarginal. These are smaller than in a corresponding example of
A. ornatissimus, the smaller and more irregular paxille of the midradial line and
center of disk of californicus being particularly well marked. In ornatissimus there
is no difference in the midradial region. Lateral spines in examples of this size are
one to three, the auxiliary lateral spines being poorly developed or absent. (Com-
pare with corresponding stage of ornatissimus). Actinal inferomarginal spinelets
are inconspicuous, and frequently there is only one series of actinal adambulacral
spines, the enlarged spine being not nearly so well developed as in the corresponding
stage of ornatissimus. The third series when present is represented by one or two
spinelets slightly shorter than those of first or inner actinal series. Most of the
small specimens have a prominent epiproctal cone.

Type.—No. 157, Stanford University invertebrate collection.

T ype-locality.—Monterey Bay, California; 70 fathoms (approximately).

Distribution.—From north of Bodega Head (north of Point Reyes), California,
to Lower California; 10 to 244 fathoms. This species is one of the commonest in the
shallower depths of Monterey Bay as determined by the steamer Albatross in 1904.
It is abundant also off San Pedro, where it has been collected in considerable num-
bers by the staff of the Biological Station of the University of California (1901).

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER.

65

Specimens examined.—The following is a complete list of localities, Californian
except as noted, from which in all 1,458 specimens have been examined:

Specimens of Astropecten californicus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
2804-2... Off San Miguel Island. .....-..- 53 stones, broken shells... ....... 1 U.S. Nat. Mus.
2900...... Off Point Conception........... 13 SONGS eon cs oe eee 5 Do.
OE ie aan Off Santa Cruz Island.......... | 68 fine gray sand................. 2 Do.
2006...... Off Point Conception........... 96 sand Mud <..2s2csecesesscsoem 2 Do.
AO pesiaees eae C0 tee eee eee eens eee | 31 ETSY. SANG Seo ce ee ea 1 Do.
201 oo Cortes Bank: <---> - -252.-22-:- 60 LOCKS; SAU Geers sae eos anee 7 Do.
PE Roe | bead DO seen eee re 26 broken 'shells--5 5-2 22-o2u-<---- 18 Do.
ZOLbsecome lasts (1 ee a ser sen eos 55 TAY SANG coca acdsee ae a Do.
2918...... toeeee oie sensoeeaeeescbeee sce <n 67 fine pray.sand-2 2-23-54... = snes 3 Do.
Oe ada |aaee CO ecesecsacca copesesceses ee oe DO Soe peace eens eae 5 Do.
AG cm ceae | eaten WGeesaedes ac weasat ees -aeecn vedi: teacee Oe eae sees aan 1 Do.
2930......| Los Coronados Islands, Lower 60 ea aatiee seen aeeeeate net = 3 10 Do.
California.
BOSD reac vatece se Ose ete fe oie won cies Sores 34 gray sand, shells.............- 27 Do.
Qoagteee alse eee eae eae 36 pray, Sandicc 22s ee we oe 30 Do.
2084.5 S55 Off San Diego. . Sbe SIeESSe WO aoe ans saens mrac esa 5 Do.
2938...... San Pedro Bay. 47 fine gray sand, stones......... 21 Do.
26 shells} .stones;2-- -.-2-5--<-.-- 2 Do.
20 Play Sand2- 2... --csce-ee anes 1 Do.
2950...... Between Santa Rosa and Santa 21 gray sand, broken shells... .._- 2 Do.
Cruz Islands. |
7: eee | Off Santa Rosa Island...-...... 57 broken shells, rocks..........- | 2 Do.
2962.....- | Off Santa Barbara.. 165 sand, stones, coral............. 2 Do.
2963 | do Dor Foe dO. aoe reesess: 16 Dow
214 Sand'stones\ 2-10.55. --4h =n 16 Do.
5 7 finelpray: SAndl <=. -scendesa.~s | 3 Do.
2967. .-.--|..-.. CO Seaees Sassen eee 30 (COBrse MUGS ws == ssa a OD * Do.
200 Estee Geen cate e eaten ee aan: 26 gray sand, pebbles, stones..... 55 Do.
2970S 22 o|b eae OO.222 joesa cate sece ew seee 29 fine gray sand, mud........... 135 Do.
207A e nena s|nee GO sa seen a neon nesee sees 20 it yan DO nena Cee 57 Do.
20f ae caned|xaade OO soo ee co seeece se wsicboace ne 61 Preen mud aon. = =e aeea sees 4 Do.
207 Sees sefaa-e Ota sees nase e CSae Aiea Ose eae eceen ease 4 Do.
DOTA eet aoa Ones sn aneea pe aace econo Tea pee Osea reer ie aeeserek ost 3 Do.
2075=<.-=- Off Santa Cruz Island.......... 36 gravel, broken shells.......... 2 Do.
2983. ...-- East of Guadalupe Island, 58 gray sand, broken shells... .... 4 Do.
Lower California.
3106...... Off Half Moon Bay........... 7 fine gray sand..........-.--... | 1 Do.
3135...-.. | Off Santa Cruz................. CoD ea ceee MOS ear eters seers 15 Do.
S1452..<< | Monterey Bay oe 5B) {leaeee MOS hea ee es ee | 1 Do.
Blais. =. | South of Point Ano Nuevo..... 56 LOW TG eee eee ane 3 Do.
3148...... Off Point Ano Nuevo.......... Ye) Eee’ CO vette n= see oer acl ae | 1 Do.
S16455-- =. PointiReyes: <--. <2. \s.cc~saenen 61 TOGKY/Sace secede Seco ss 1 Do.
OLT2 esses Northwest of Bodega Head..... 62 plack sand <2. a8=.2--e sess se 2 Do.
LOU a ocee | Between Cape St. Martin and 53 fine gray sand................. 2 Do.
Piedras Blancas. t
68 gray sand, mud............... | 1 Do.
56 green mud, sand.............. 1 Do.
..| North of Monterey Bay......... 108 fine gray sand................. | 6 Do.
PUOUEBTG gS hee ter asain eee | eee eine | acts am cieia aoa cea sne Sees ea 17 Do.
Off Santa Barbara...-.......... BL rn cwelscte ata Savec wees cevsaceewes | 1 Do.
Off SaniD lego =<. cn ec. neaee on MPAs PO cee nec us cyan eee ewe see tncanaee 5 Do.
Wo lonalify Gi. 2 en canes iene Ieee eal von ns act anna oetenancscenmeneeace ! 46 Do.

@ Several bottles broken and the contents scattered by California earthquake, 1906.
57444°—Bull. 76—11——5

66 BULLETIN 76, UNITED STATES NATIONAL MUSEUM. '

Specimens of Astropecten californicus examined—Continued.

Station. Locality.
8207 -.-<-- | Off Cerros Island, Lower Call-
fornia.
Monterey Bay :
4304...... Vicinity of San Diego....-------
S852 Soc cat locsee Osc ceva pans eencaes an sessnne
4343...... Near Los Coronados Islands,

Lower California.

Between Santa Barbara and
San Nicholas Island.

4431...... Off Santa Rosa Island...-....-.-

4439...... Monterey Bay (northeast of
Point Pinos).

ME ens |se eae WO. a geen sosceeesceecenea ==

A45S co. wenle=a == DO sen= seecca ee oscese a=:

MART Gece samen OOnteccceccccae serene c= -seee

4464...... Monterey Bay (southwest of
Moss Landing).

4473....... Monterey Bay (north of Point

Pinos).

4476...... Monterey* Bay (southwest of
Moss Landing).

SET wacceclsasea OD 2eettecoscansh = ce cenweaen =

4492...... Monterey Bay (off Hill’s Land-
ing).

S50 Ree. sls-s =o MO aeceeacetsccerss sees

4535...... Monterey Bay (3 miles off Pyra-
mid Point).

4548...... Monterey Bay (between Italian
Ledge and shore).

4500 Secs cs Monterey Bay (near Italian
Ledge). /

AGB Sennen lorena G0 seen ce aero seunne ee

4552......| Monterey Bay (70 Fathoms
Bank).

A00D cewac laren WO netet eee ac acr one kn aoe

MBBS aware Monterey Bay (Italian Ledge)...

4557...... Monterey Bay (Portuguese
Ledge).

4550...... Monterey Bay (off Mussel Point)

4562...... Monterey Bay (vicinity of Santa
Cruz).

Off San Pedroa

Depth. | Nature of bottom.
Fathoms.
.| About 70|........------ee2e- ee ee eee n nee
25 coarse yellow sand, shells,
gravel.
62-183 | gray sand, rocks.....---------
55-155 | fime gray sand.......----------
244
46-50 | dark green mud, fine sand.....
82-134 | green mud
164-85 | gray sand.......--------------
32-33 | fine gray sand...........------
229-298 | gray mud, rocks......---. ore
31-32 | gray sand, shells........------
30-41 mud, gray sand, rocks........
42-40 | gray sand, shells.......-------
49-50 | green mud.........-..--------
49-51 | dark green mud............-.--
46-40 |..... GOs seweseean ce acie sma alone
51-36 | soft, dark gray mud...........
54-65 | gray sand, mud........-.....-
25-39 | soft green mud.............---
19-10) jo <e OO ia rcesaae casos eoone
26'S isceee COSe senee as aac ee ene eee
12-11 hard sand sss-2s.ss-s ease wre
71-54 hard gray sand.............-..
46-54 coarse sand, shells, rocks. .....
50-54 green mud, rocks...........-.-
46-56 | coarse sand, shells............-
73-66 | green mud, rocks........--.-.-
66-69
50-59
53-54
22-8 fine prayisand: =~ .-.-2s:--...0-
10 STC SANG espe ene Cv ecace cance
24-102 | hard and soft sand, green mud,
brown mud and pebbles.

Number.

ll

wn

493

!

Collection.

U.S. Nat. Mus.

Stanford University.
Albatross, 1904.

Do.
Do.

University of California.

1

@ These specimens, kindly sent by Prof. W. E. Ritter, were taken at twenty-nine different hauls, ranging in depth from 2}

to 102 fathoms, but the majority from depths of less than 30 fathoms (nineteen hauls).

The bottom was sand together with

shells, shale, cobbles, pebbles, or sand alone; less frequently mud or sandy mud (seven hauls). Many of the specimens are

quite small.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 67

Remarks.—¥our species of Astropecten from Mexico and Central America must
be considered in connection with A. californicus. A. fragilis Verrill, Panama
to Peru, is a slender rayed form with a narrow paxillar area on the rays and a
small tubercle at upper end of the proximal superomarginals (apparently char-
acteristic). All the adambulacral spines are slender, the middle furrow spine
exceeding width of plate. The enlarged subambulacral spine is still longer, slightly
tapered, with a slight groove on the upper or outer side near the bluntly pointed
tip. Adorad is a very slender spinelet in line, and behind the two are two or
three capillary spinelets. Inferomarginal spines two, rather broad.

A. regalis Gray (San Blas and south) is a short-rayed species (R=3 r) and has
one sharp enlarged actinal adambulacral spine, scarcely longer than the longest
furrow spine. A. verrilli de Loriol (Mazatlan) is likewise a comparatively short-
rayed form (R=3.4 r) and differs from californicus in haying small tubercles on
the superomarginal plates, forming a single longitudinal series. The inferomarginal
and adambulacral armature is also slightly different. A. rubidus de Loriol (Mexico)
is allied to articulatus (Say), having broad superomarginal plates, a smaller disk
than californicus, with rays broader at tip, paxillee with shorter spinelets, adambu-
lacral plates with much smaller spinelets—three small ones in actinal series. The
minor details of inferomarginal armature are different.

Until an extensive series of Mexican and southern Astropectens is compared
by one man it will not be possible to determine the validity of the rather numerous
nominal species. It is highly probable that several so-called species are varia-
tions of a single form. Neither is it possible at present to determine the limits
of Gray’s species, for his descriptions are nearly useless. The species called armatus
may eventually be divided into several races, as Liitken’s drstedii seems to differ
in some respects from typical armatus (a specimen in the Yale Museum, no. 817,
having been examined).

Although californicus and ornatissimus seem to be perfectly distinct, they may
eventually be found to intergrade with southern species.

In the Museum of Comparative Zodlogy (No. 190) are two medium-sized speci-
mens of californicus with rather shorter than typical rays and decidedly truncate
enlarged subambulacral spine. These are supposed to come from Alaska, but
the locality is not correct. As the Albatross in her numerous Alaskan dredgings
has never secured the species (or even north of central California) I have not
considered the record authentic.

ASTROPECTEN ORNATISSIMUS Fisher.
Pl. 6, figs. 3, 4; pl. 7, fig. 2; pl. 51, figs. 1, la-c.
Astropecten ornatissimus FisHer, Proc. Wash. Acad. Sci., vol. 8, Aug. 14, 1906, p. 119.

Diagnosis.—Resembling A. californicus in general features, but differing in
having larger paxillx, longer, slenderer, and more distinctly spaced paxillx spine-
lets; in having the paxille superficially more uniform in size across the ray, not
more elongated along midradial line, and smaller over a large area at center of disk;
in having longer, slenderer, and more tapered adambulacral spines; in having

@ A specimen in the Yale Museum No. 1302, Zorritos, Peru, has been examined.

68 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

the upper edge of each ambulacral plate produced into a thin lamina between
consecutive pairs of ampulle, the free edge bearing a comb-like series of slender
but irregular teeth or spinelets. R=56 mm.; r=14 mm.; R=4 r. Breadth of
ray at base, 16.5 mm. (type).

~ Description.—The paxille afford the most evident difference between orna-
tissimus and californicus. In californicus there is a considerable area around
center of disk in which the paxille are smaller and more crowded than on remainder
of disk and on rays, and paxille of midradial regions are more irregular, at least
in arrangement, than along margins of ray. In the present form the large paxille
extend nearly to center of disk, there being only a small area of small paxille.

The paxillae of sides of rays are not in such regular rows and are not easily
differentiated from the midradial ones. About three or three and one-half trans-
verse series of paxille correspond to two superomarginal plates at base of ray
(usually five in californicus), about five at middle of ray and six or seven near tip.
Opposite suture between second and third superomarginal plates about twelve
or thirteen paxille can be counted across ray to same point on opposite side
(eighteen to twenty in californicus). Large paxille at base of rays with fifteen
to eighteen peripheral and ten to fifteen central spinelets, which are much longer
than in californicus, terete, with rounded or clavate tips. Tabulum of paxilla
fairly broad so that both central and peripheral spinelets appear spaced, giving
the whole an open floriform appearance. Farther along ray, one to six central
spinelets to a paxilla, and upwards to fifteen or eighteen peripheral. At very
end of ray the paxille are much smaller.

Although superficially there is little difference between the paxille of the mid-
radial line and those along sides of area, the plates or bases of the paxille are very
different. If the dorsal integument is removed, treated with caustic potash and
viewed from the inner or coelomic side the outline of the plates is readily seen.
Those along the midradial line and on center of disk are subcircular or very
obscurely lobed, are without papule between them, and imbricate strongly, espe-
cially on disk. In large specimens the plates are usually entire, and roundish,
but tend to become lobed adjacent to the papular areas. On the ray these same
plates become more elliptical and irregular in outline as well as arrangement.
Two or three series have no papule between them. On either side of the mid-
radial line of entire plates as well as over the interradial triangle of disk the plates
are smaller, strongly lobed, and touch or overlap by these lobes, allowing the
papule to pass between. On the ray each plate is longitudinally elongated, sub-
elliptical with two short but abrupt lobes at either end. The plates of consecu-
tive transverse rows alternate so that the processes join every plate with four
others, two adorally and two distally. Between any two plates there are, as a
rule, two papule in a longitudinal row, a single plate having six papule surrounding
it. On the disk these plates are less regular and frequently have as many as six
lobes. Occasionally the slender lobes appear as independent ossicles. In the
largest specimen examined the regular plates on ray do not touch one another,
but do so regularly in a medium-sized example. On the adcentral side of the
madreporic canal where it pierces the dorsal integument is a large convex plate,
hollowed on the side toward the canal; on either side of this, forming a sort of

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 69

horseshoe around the canal, are two other smaller, slenderer plates, closely joined to
the large one. The side toward marginals is without a specialized raised ossicle.

Superomarginal plates thirty-two to a ray, without enlarged spinelets or tuber-
cles. General surface covered with short spinelets, delicate except along median
transverse line where they are clavate to thimble-shaped, increasing in size toward
upper end of plate (same spinelets are markedly squamiform in californicus).

Armature of inferomarginal plates very similar to that of californicus, there
being usually two or three marginal spines obliquely placed, and, in a line, three
more spaced, smaller, spines along aboral edge of plate. The auxiliary lateral
spines situated just adorad to the regular lateral spines on each plate are longer
than the same spines of californicus.

Adambulacral furrow spines three or four, similar to those of californicus. First
actinal series with two spines, the aboral being much the longer, tapering, slightly
flattened, bluntly pointed, longer and slenderer than the corresponding spine
of californicus. The adoral member is about as long as the furrow spine which
stands opposite. Outer or second actinal series usually consists of three slender
untapered spines somewhat shorter than furrow spines and standing in a fairly
regular row. Near base of furrow two or three very small spinelets sometimes
stand on outer end of plate.

Actinal interradial areas very small, paved with six or eight paxilliform plates
in large specimens. Each paxilla bears a floriform group of about eight to twelve
slender spinelets, of unequal thickness (more or less), those in center being usually
the stoutest. Occasionally there is a single central sharp spine. In specimens
from off Lower California, and to a less extent in those from off San Diego, the
spinelets are, on some of the plates, grouped or coordinated to form a sort of pedi-
cellarian apparatus. Even the spinelets on the actinal surface of the adambu-
lacral plates are frequently arranged in a similar manner. As a whole the inter-
radial areas may be said to be very inconspicuous.

Mouth spines similar to those of californicus, but the marginal series stand
slightly spaced from the intermediate spines, so that inner end of combined plates
is broader and the three series, superficial, intermediate, and marginal, are more
clearly evident. All spines are slenderer and a trifle longer than in californicus.
Marginal spines, about seven between tooth and inner end of first adambulacral
plate, and about six or seven more minute spinelets continue the series two-thirds
distance to outer end of plate.

Madreporic body concealed by paxille, situated as in californicus and crossed
by sinuous striz; tiny, spiniform knobs on ridges of californicus apparently lacking.

Superambulacral plates well developed. The upper edge of each ambulacral
ossicle is produced into a thin lamina or ridge between consecutive pairs of ampulle,
the free edge being sharp and provided with very slender, irregular teeth, five or
six in number. These give the appearance of a comb. The upper end of each
plate is produced into a knob, bearing several very slender spinelets; or the knob
may have a sharpened edge bearing teeth. Between the knobs of opposite plates
pass the transverse muscles of the ambulacral ridge. Very tiny spinelets are
discoverable between consecutive knobs and transverse muscles of the ambulacral
ridge. The presence of a comb of spinelets or teeth on the upper edge of the ambu-

70 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

lacral plates will serve to distinguish this species from californicus, especially when
other characters fail. Gonads arranged in a tuft on either side of the interradial
line. Six Polian vesicles present, one in each interradius except that of madre-
poric canal, in which there are two; the same interradius having a double inter-
radial septum, the others a single. Two Tiedemann glands to each interradius.

Color in alcohol, bleached yellowish to whitish; color in life unknown.

Variations—The variations in this species mostly concern minor details of
ornamentation—characters which are unstable in nearly all species of starfishes.
Thus the spinelets are fewer in small specimens and are slightly more robust in
some than in others. As usual the smaller specimens have shorter, stouter arms.
There are a relatively few specimens of A. californicus (stations 3145, 3147, 3148,
3172) which bear a certain resemblance to ornatissimus, chiefly because the paxille
are larger and more uniform, as seen superficially, than in typical examples. The
paxillar spines, however, are shorter and more compact and the auxiliary infero-
marginals spines (that is, the lateral ones) are short, while the comb of teeth on
the ambulacral plates, described above, is entirely lacking. Likewise the enlarged
adambulacral spinelet is of character of californicus—much flattened, relatively
short with a round tip, and not very conspicuously tapered. It is barely possible
that some of these aberrant examples of californicus may be hybrids. There is
no good evidence that the two species intergrade, although it is not impossible or
even improbable that they may converge somewhere off Lower California. The
forms, though different, are so similar in general structure that one is impressed
with the notion that they may be offshoots of a common species still living in the
south, possibly A. verrilli or a nearly related form. There is a wide difference,
though, in the toothed and smooth upper edge of the ambulacral plates.

Young.—In small specimens of ornatissimus the characters of the adult are
fairly obvious, especially when examples are compared with equal-sized individuals
of californicus. The paxille are large and the superomarginal plates encroach
onto the abactinal area rather more than in the adult. In a specimen with R equal
to 20 mm. (r=6.5 mm.; R=3r) there are about three transverse rows of abactinal
paxille opposite every two superomarginals at base of ray, and about ten paxille
to a transverse row, opposite the second superomarginal. Superomarginal plates,
twenty; lateral spines of inferomarginals, two or three, the auxiliary lateral spines
well developed and actinal inferomarginal spinules prominent. Adambulacral
armature similar to that of adult, the long spine of second series (first actinal) well
developed. Usually there are only two actinal series each with two spines, the
outer shorter than those of furrow series.

Changes in dimensions due to size are shown in the following table:

Table showing variations in Astropecten ornatissimus.

No. of |

‘ | Breadth
Station. R. r. sr. ofray at |superomar-|
| base. ginal plates.
4309 26 8.0 | 3.25:1 9.0 21
4304 36 9.0 | 4:1 11.0 25
4306 43 9.5 | 4.5:1 11.0 28
| 4413 | 56 14.0 4:1 16.5 32
4413 | 68 14.0 4.8:1 16.5 | 40

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. to

Type.—Cat. No. 21927, U.S.N.M.

Type-locality.—Albatross station 4413, between Santa Catalina and Santa
Barbara islands, in 152 to 162 fathoms, on gray sand.

Distribution.—Vicinity of Guadalupe and Cerros islands, Lower California,
north to vicinity of Santa Catalina Island, and San Pedro, California; undoubtedly
ranging much farther south. Bathymetrical range, 47 to 162 fathoms.

Specimens examined.—The following is a list of localities from which one
hundred and fifty-four specimens have been examined:

Specimens of Astropecten ornatissimus examined.

Station. Locality. Depth. Nature of bottom. No. Collection.
Fathoms.

202s as. Cortes Bank, off San Diego, Cali- 145 fine gray sand............... 5 U.S. Nat. Mus.
fornia.

2935)..555% Off Point La Jolla, near San Dees (een Oe se ee ee eee 68 Do.
Diego, California.

2083...... East of Guadalupe Island, Lower 58 gray sand, broken shells .... 3 Do.
California.

3662. 222-. Off Avalon, Santa Catalina Island, 47 fine gray sand............... 1 Do.
California.

Notirecorded eic7 cesses snes <5 se 1 Do.

Off Cerros Island, LowerCalifornia. 12 Do.
Near San Diego, California........ 4 University of California.

San Pedro, California. ............ : 10 Do.

67-116 | gray sand, shells ............ 6 Albatross, 1904.

207-497 | green mud, fine sand........ 1 Do.

67-78 fine sand, shells, rocks....... 27 Do.

71-75 | green mud, fine sand........ 5 Do.

81-134 | green mud, fine sand........ 5 Do.

--| 101-129 | gray sand, mud, rocks....... 1 Do.

| 89-90 | coarse sand, rocks........_.. 2 Do.

152-162 | fine gray sand............... 3 Do.

Barbara islands, California.

Remarks.—This species differs from A. fragilis, A. regalis, A. verrilli, and A.
rubidus in the same respects as does A. californicus.

Genus PSILASTER Sladen.

Psilaster SLADEN, Narr. Challenger Exp., vol. 1, 1885, p.611. Type, Astropecten andromeda Miiller
and Troschel; Challenger Asteroidea, 1889, p. 221.
Bathybiaster Fisner (not Danielssen and Koren), Bull. Bur. Fisheries for 1904, vol. 24, June 10,
1905, p. 295.
Plutonaster Lupwia (not Sladen), Mem. Mus. Comp. Zoél., vol. 32, July 17, 1905, p. 30.
Phidiaster Kenuer, An Account of the Deep-Sea Asteroidea collected by Roy. Ind. Mar. Surv.
Ship Investigator, Calcutta, April, 1909, p. 28. Type, Phidiaster agassizi Keehler.
Diagnosis.—Disk small; rays five, high at base, tapering, pointed, with sides
perpendicular or arched inward toward actinal surface; marginal plates massive,
plane, not tumid, separated by fasciolar grooves, either shallow or fairly deep;
superomarginals with or without large spinelets; inferomarginals regularly with
several enlarged spinelets, either appressed to ray or bristling, and in the form of a
perpendicular comb; exceptionally the inferomarginal spinules form two fairly
regular longitudinal series along ray; abactinal area with compact paxille, arranged

72 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

usually in regular transverse lines at sides of ray, and with polygonal or roundish
bases: actinal interradial areas small, with numerous small intermediate plates
which extend far along ray, and which bear compact groups of spinelets often
subsacculate; papule absent from midradial area and center of disk; armature of
adambulacral plates, a furrow series of rather long delicate cylindrical or flattened
spinelets in a straight or slightly curved series and on actinal surface, about two
longitudinal series of shorter more compressed spinelets, sometimes irregular in
disposition; anus present or absent; well-developed superambulacrals; tube feet
conical without sucking disks; no pedicellarie.

Remarks.—This genus differs from the closely related Bathybiaster in having a
rather numerous furrow series of slender subequal spinelets in a slightly curved
comb, the outer edge of which is frequently straight; the central spinelet of series
is not greatly enlarged, nor is the series very angular as in Bathybiaster. There is
not a flap of integument at the tip of any spinelet, as typically the case with the
central spinelet in Bathybiaster. Sometimes the actinal spines are nearly as saccu-
late as in Bathybiaster, but the marginal spines are much more prominent in Psi-
laster. In Bathybiaster the base of a paxilla is stellate; in Pstlaster roundish or

polygonal.
PSILASTER PECTINATUS (Fisher).

Pl. 7, figs. 5, 7; pi. 10, figs. 1-3; pl. 50, figs. 3, 3a-c.

Bathybiaster pectinatus Fisner, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 295.
Plutonaster abyssicola Lupwie, Mem. Mus. Comp. Zo0l., vol. 32, July 17, 1905, p. 30, pl. 3, figs.
13, 14; pl. 19, figs. 102-106.

Diagnosis.-—Rays five. R=73 mm.; r=15.5 mm.; R=4.7 r. Breadth of
ray at base, between first and second superomarginals, 16 mm.; midway along
ray, 13.5 mm. Rays moderately elongate, tapering continuously from base to a
sharply pointed extremity; interbrachial angles rounded; lateral wall, formed by
marginal plates, nearly vertical, or sloping inward toward the actinal surface, only
the upper end of the superomarginal plates forming a narrow border to paxillar
area; abactinal surface slightly inflated on disk usually plane on rays; actinal
surface convex; a low epiproctal cone usually present; a very small anal aperture.
Paxille of abactinal area small, not crowded, with about four central stubby spine-
lets surrounded by a peripheral series of ten to fifteen slender ones. Superomar-
ginal plates each with a sharp spinule near upper end, on edge of ray, forming a
series for two-thirds or three-fourths length of arm; sometimes two such spinules
present; inferomarginals with two spinules, one above the other, the upper forming
a series to tip of ray, the lower a parallel series for only about two-thirds length of
ray. Adambulacral plates with a furrow series of eight spinelets (seven in small
specimens) and on actinal surface two or three longitudinal series of three or four
each, invested in a pulpy saccular sheath. Actinal intermediate areas small, the
plates beset with compact groups of very fleshy, spatulate, round-tipped papille.

“ Vrom type; important variations exhibited by other specimens are mentioned, however.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 183

Description.—Paxille of abactinal surface small, not crowded except at center
of disk; arranged in transverse rows at sides of the paxillar area of rays; five or
six such series corresponding to two inferomarginals at base of the rays; no definite
arrangement along median radial line, nor in central portion of disk; size of paxille
decreasing toward extremity of rays, and on the disk, toward center; on the rays
the largest are along midradial line, and on disk about midway between center and
edge. Each paxilla consists of a compact, flat-topped, subcircular or irregularly
elliptical group of small, cylindrical, round-tipped membrane-invested spinelets, of
which two, three, four, or five (on larger paxille:) are stouter and occupy the center
of tabulum, being closely surrounded by ten to fifteen slenderer spinelets, which
frequently have the membrane thinner. Small papule are visible in the spaces
between the paxille, especially on rays.

The abactinal plates or bases of paxillee (viewed from inner or ccelomic surface
of abactinal integument) are generally circular in outline except on outer half of
ray, where they are longitudinally elliptical, and in center of disk and along mid-
radial line, where they are often slightly irregular. The plates are spaced, and are
much closer together in center of disk and along median radial line, where there
are no papulx, than along sides of abactinal area. On the latter areas they are
arranged in very regular transverse rows which assume an interradial direction on
disk, and the plates increase in size toward the midradial line, where for a width of
about three plates there are no papule and no regular arrangement of plates. The
plates decrease in size toward center of disk. The papule are single and so arranged
that there are usually six about each plate of the papular areas. The area without
papule in center of disk has a radius less than one-third distance from center to
edge of paxillar area.

Superomarginal plates, forty-six in number, from median interradial line to
extremity of ray, are nearly vertical in position, the upper ends forming a narrow
border to paxillar area. In some specimens the plates slope inward toward the
lower end so that the greater part of the plate can be seen from actinal side. Plates
are about two-thirds as long as high, very regular, with curved upper end, and are
covered with small squamiform spinelets or granules along the median transverse
or vertical line, these becoming slenderer and more papilliform toward the edges
of the plate, especially in the deep fasciolar grooves, where they are capilliform.
At about 1 mm. from upper end of the plate is an erect, slightly flattened, sharp
spinule attaining a length of 1 to 1.25 mm., and decreasing in length toward
extremity of ray. On first two or three plates they are much reduced in size.
They form a regular longitudinal series at border of the abactinal surface, extending
two-thirds to three-fourths length of ray. In typical specimens an occasional
plate has a second spinule below the first, and rarely three in a vertical series.
Several specimens from station 3308 have two spinules on nearly all the plates except
in the interradial angle (where the single spinule is not reduced in size) and toward

‘the outer end of series. Similarly, three spinules occur on a number of infero-
marginals, so that a side view of ray shows four or five spinules in a vertical series.

Inferomarginals correspond exactly to superomarginals in number and length,
the line separating the two series being straight; plates slope inward, but the exposed
surface is strongly arched, forming a rounded margin to actinal surface, the latter

74 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

passing gradually into the lateral face of ray. The dorsolateral margin, on the
other hand, is very abrupt. Plates are covered with squamiform, broadly lanceo-
late, ovate, obovate, blunt, or often sharp spinelets, which increase in size toward
lower end of plate, but are very slender and terete in the fasciolar grooves. About
1 mm. from the upper end of each plate is a flattened slender tapering spine, attaining
a length of 3.5 mm., the series of which extends the length of the ray, being absent
occasionally from the first one to three plates. These spines or spinules either
stand out from side of ray or are frequently appressed against it and may be
equal to or longer than superomarginal spinules. Spaced from the above spinule,
about midway between its base and lower end of plate, is a second similar but
shorter spinule, forming on the plate a vertical series, and a longitudinal series
for about two-thirds the length of ray. Some specimens have a few plates with
three spinelets in a vertical series. All the spinelets covering the marginal plates
are sheathed by membrane, which in alcoholic specimens causes these spinelets to
appear very compactly placed, with broad rounded tips. When the specimen is
dried the squamiform spinelets are seen to have narrower rough tips and to be
distinctly spaced. Terminal plate a trifle wider than long, rounded distally, con-
cave next to paxillar area, where three rows of paxillze touch its free edge. The
granuliform covering is rubbed off.

Adambulacral plates with a prominent angular margin to furrow. Armature
consisting of (1) furrow series of eight (seven in some specimens) delicate subequal,
membrane-invested, round-tipped spinelets, the two or three at either end of series
flattened with sides to furrow, while the four or two central spinelets have the edge
to furrow; the lateralmost spinelets are often shorter than the others; (2) on
the actinal surface are two or three longitudinal series of flattened, truncate or round-
tipped spinelets, four or five to each series. The outer spinelets are slightly wider
than the inner and all are subequal to furrow series. Actinal spinelets are sheathed
in a thick pulpy membrane which gives them a papilliform clavate appearance.
In dried specimens the spinelet is seen to be slenderer, but flattened, with a rounded
tip often wider than base. The first and second adambulacral plates are much
compressed, and each has two transverse series of spinelets, about twenty to twenty-
four on the first plate and about sixteen on the second.

Mouth plates elongated in the interradial direction, narrow, the median suture
being dilated near the middle. The spinelets are arranged in two parallel series
on each plate. The marginal series consists of about fifteen to eighteen short,
regular, slightly flattened spinelets, decreasing in length at outer end of plate,
but subequal all along inner half. These spinelets stand close together and are
appressed against a superficial series of about the same number of heavier, cylin-
drical, subglobose or clavate skin-covered spinelets, which increase in thickness
toward the outer end of plate. The inner end of the combined pair of plates is
truncate, and either two or four spinelets at the angle are slightly larger and heavier
than the others, are slightly bent at base, and are appressed against innermost of
the superficial spinelets. Typically the spinelets are very regular in arrangement,
the superficial series of companion plates being close together in some examples,
in others separated by a conspicuous furrow. The armature of mouth plates is

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 75

set at a considerably higher level when animal is viewed from actinal side than
is that of the adambulacral plates; that is, the plates are very prominent.

Actinal interradial areas small, the plates extending to twentieth or twenty-
second inferomarginal; or a little less to a little more than two-thirds length of
ray. The plates adjacent to adambulacrals are largest, and the arrangement is
in irregular rows between the adambulacral and inferomarginal plates, there being
more or less variation in details.¢ In the Californian specimens, which are smaller
than Ludwig’s material (from Gulf of Panama), the plates on the ray are not so
regular as in his figure 102, there seldom being two intermediate plates of the external
series corresponding to one in the series adjacent to adambulacrals. Usually there
is but one. The difference in size may account forthis. The plates are beset with
compact groups of very fleshy spatulate round-tipped papille, which become pressed
into various shapes by mutual contact. The calcareous spinelet within the papilla
is delicate, subcylindrical, truncated, sometimes flattened at tip. The number
of papille to each plate varies, and they are so compactly arranged that if the
prevalent slime is thoroughly cleaned off, the outlines of the groups in the inter-
radial areas are clearly distinguishable. At the outer end of the mouth plates are
a few isolated papille resembling in shape a flattened grain of corn attached by the
small end. In Bathybiaster pallidus Danielssen and Koren describe such groups
of spinelets as pedicellarizx. In the present species although the compact grouping
may suggest pedicellariz, it is not at all likely that the spinelets subserve any such
function. Structurally they are no more to be considered as pedicellarie than
are ordinary paxille.

Madreporic body small, broadly oval, situated a little to the outer side of a
point midway between center and margin of disk. Striations coarse and inter-
rupted, radiating from the adcentral side of plate. The ridges are occasionally
branched.

Anatomical notes.—Superambulacral plates well developed, but absent from
first two ambulacral plates and from about the last twelve which are small and
crowded against marginals. All except the two first superambulacrals extend
from the ambulacrals to upper end of inner surface of inferomarginals. The first
two meet the actinal intermediate plates. This varies probably with the size of
the specimen. A very small anal opening is present to which a short intestine
leads from the dorsal stomach. On the side of the intestine toward madreporic
canal, and on the opposite side (toward anterior ray) is a digitate saccular intestinal
ccecum which varies in shape in different individuals. In one example the ceca have
short very irregular branches, and in another four and five regular divisions, respec-
tively, the latter resembling glove fingers. Gonads in a tuft on either side of inter-
radial line, the aperature of each about 2 mm. from marginal plates in the abactinal
integument. Polian vesicles six, one on either side of madreporic canal, and one
in each of the other four interadii. Interradial septa uncaleified.

@ See Ludwig, Mem. Mus. Comp. Zodl., vol. 32, 1905, pl. 19, fig. 106.

76 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
Variations.—Certain variations due to size are shown in the following table:

Table showing variations in Psilaster pectinatus.

| vumb 7 Number of

: ; | Breadth of Sareea Nemberos inferomar-
| Station. R. I. Rir. ray at base. ginal rae eines ginal
| plates. spines.

| a!
3308 16 6.5 | 2.46:1 7 20 1 1
3308 29 10.0 | 2.9:1 10.5 27 1 2orl
| 3308 53 13.0 | 4:1 15.0 38 lor2 3 tol
| 3308 71 16.0 | 4.3:1 19.0 47 1, rarely 2 3tol
4387 73 15.5 4.7:1 | 17.5 46 1 2orl
3306 S4 17.0 | 5:1 17.5

| 58 2orl 3 tol

The greater number of specimens examined are from Bering Sea. Among
these the chief variations are in proportions, including length and width of ray
relative to disk; in the amount of inflation of abactinal integument; relative size
of epiproctal elevation; number of superomarginal spinules; the degree which the
exposed surface of marginal plates, or in other words the side wall of ray, departs
from the perpendicular; adambulacral spinelets of furrow series are seven or eight;
slight difference in size of actinal interradial areas; the tenuity of extremity of
rays; compactness of actinal intermediate paxille. There is also a trifling varia-
tion in the size of paxille of abactinal area; as a whole the specimens are remark-
ably uniform in this character.

Unfortunately there are only three adult specimens from off southern Cali-
fornia. These are pretty much alike. One specimen has a narrower paxillar
area on rays, but a larger disk and larger paxille than have the other two. Com-
pared with specimens from Bering Sea, these examples present no important dif-
ferences. The rays are a trifle thicker dorsoventrally, and the sides are a little
higher in consequence. The marginal plates are more vertical in the southern
specimens than in those from Bering Sea, where the side wall of ray generally
slopes inward on the actinal surface. But many from the latter locality have
the side wall nearly vertical. There are not enough southern specimens to deter-
mine whether the presence of a second superomarginal spinule is more frequent
in the northern individuals. The marginal spinules sometimes stand out at an
angle from ray, but are more often tightly appressed to side; this difference changes
the general facies in a misleading manner. Smaller specimens as a rule have.the
rays thicker near the tips than the large examples.

All the variations appear to be of slight importance—that is, are such as may
be expected in starfish species—and the Alaskan examples belong, seemingly, to
the same species as the California specimens. The latter show no important dif-
ferences from Panamic examples described by Ludwig under the name Plutonaster
abyssicola.

Young.—Young examples have much shorter, blunter, and distally thicker
rays than adults and half-grown specimens, the difference in proportions having
been given in the foregoing table. The marginal plates are more prominent on
the abactinal surface; each plate of both series (except near tip of ray of supero-
marginal series) bears a single spinule. The adambulacral plates have five to

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. ed

seven furrow spinelets. Actinal interradial areas are smaller than in large exam-
ples, the plates extending about one-half length of ray. Both the upper and lower
edge of ray is rather abrupt, not the upper alone as in adults. The epiprictal cone
is scarcely more prominent than in adult.

Type.—Cat. No. 22327, U.S.N.M.

Type-locality— Albatross station 4387, off San Diego and southeast of San
Clemente Island, California; 1,059 fathoms, mud.

Distribution.— Bering Sea to Bay of Panama,? in 1,033 to 1,625 fathoms, on
green ooze, green mud, sand, and hard bottom.

Specimens examined.—Total number, one hundred and nineteen, from the
following stations; six specimens have no label.

Specimens of Psilaster pectinatus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms. |
3307....-- Bering Sea, northwest of Umnak Island .......... 1,033 29 | U S. Nat. Mus.
3308...... Bering Sea, west of Pribilof Islands ..... 1,625 47 Do.
3604.2... Bering Sea, south of St. George Island ... 1,401 Do.
SOO en osc} ee cis Die eae fesse a ee ee es 1,044 32 | Do.
AOR Teco Off San Diego, California ................-.-----.- 1,059

4 | Albatross, 1904.

Remarks.—This species has the general appearance of a long-spined Bathy-
biaster, but the character of the abactinal plates, adambulacral armature, and
rather deep marginal fascioles assigns it to Psilaster. The double row of infero-
marginal and single row of superomarginal spines is very similar to the arrange-
ment of the much smaller spinelets in Bathybiaster, and the actinal spinulation is
decidedly sacculate. It was these considerations which led me to describe the
species under Bathybiaster. Since then, however, I have seen an abundance of
material of this species and have examined Bathybiaster vevillifer and B. robustus.

The character of the adambulacral armature is the most tangible feature which
separates the two genera. Both may have sacculate actinal spinelets, but Psilaster
has not the very angular furrow series with an elongate median spine bearing,
sometimes, a terminal flap of integument. The marginal spinules of Bathybiaster
are very inconspicuous and the vertical fascioles along the suture between the
plates are decidedly shallow.

Ludwig classifies this species in Plutonaster, a course which seems to me wholly
untenable, since it must necessarily ignore the sum of the structural characters in
favor of the presence of an anus. This is a good example of the bizarre results
which follow adherence to the erroneous idea that Astropectinide lack and ‘‘ Archas-
teride”’ possess an anus. It is also an instance of the fact that the presence or
absence of an anus is not even of generic value. It does not seem logical to ignore
all of a half-dozen important skeletai and anatomical features in favor of a single
oft disqualified character.

This species is readily distinguished from those described by Sladen, by the
presence of one or two prominent superomarginal spines forming a regular series

@ Plutonaster abyssicola=Psilaster pectinatus.

78 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

all along the upper edge of the ray, and by the fact that the bristling inferomarginal
spines form about two longitudinal rows along at least two-thirds length of ray,
and one row along the outer third.

Kehler ® has described the genus Phidiaster (Nicobar Islands, 888-930 fath-
oms), which is distinguished from Psilaster by the great development of the supero-
marginals. These plates, as well as the inferomarginals, carry each two large robust
spines, forming thus four longitudinal series along the ray, one dorsal, one ventral,
and two lateral. Another character is the presence on the inner angle of each
mouth plate of two enlarged teeth.

In some respects Psilaster pectinatus is intermediate between Psilaster and
Phidiaster. It has the superomarginal plates sufficiently large to satisfy the quali-
fications of Phidiaster, but has only one superomarginal series of spines, and the
lower of the two inferomarginal series extends only part way along the ray. The
teeth are not conspicuously enlarged. The characters of Phidiaster appear to be
somewhat trivial for generic value. The superomarginal spines vary in the genus
Persephonaster, being absent in the species formerly ranged under Psilasteropsis.
Kceehler has, himself, called attention to the fact that the presence or absence of
superomarginal spines can not be regarded as of generic importance in Persephon-
aster. Certainly the same remark applies equally well to the nearly related Psilaster
and Astropecten. The superomarginals of Phidiaster, judging by plate 9, fig. 3
(Keehler, 1909), are not larger than in Psilaster pectinatus, and some other species,
as, for example, Psilaster flore (Verrill). The superomarginals of Astropecten vary
greatly in relative size, and I have seen enough species of Persephonaster to know
that they do also in that genus. The remaining character, that of the teeth, is not
shared by typical Psilaster, although there are two (instead of four) more or less
enlarged median teeth at each mouth angle. These, in Psilaster, are usually bent
up and not directed across the actinostome as in Phidiaster. From an analysis
of these characters and owing to the intermediate position of Psilaster pectinatus
T am inclined to unite Phidiaster with Psilaster.

Genus THRISSACANTHIAS Fisher.

Thrissacanthias FisHer, Ann. and Mag. Nat. Hist., ser.8, vol.5, Feb., 1910, p.171. Type, Perse-
phonaster penicillatus Fisher.

Diagnosis.—Near Persephonaster Alcock, but differing in the arrangement of
gonads which are not confined to the interradii but extend for a short distance
along the ray as a number of separate tufts depending from the genital rachis;
adambulacial plates with one or two enlarged actinal spines; disk of medium size,
depressed, rays long; marginals opposite, with conspicuous spines; abactinal sur-
face with true paxille arranged in regular oblique transverse rows; no enlarged
radial series; papule all over abactinal surface except at tip of ray; actinal inter-
mediate plates extending far along ray; interradial areas rather small; spiniform

Senn pedicellarie on marginals; anus present; madreporic body large, not
\idden.

2 Deep Sea Asteroidea of the Investigator, 1909, p. 28, pl. 9, Hips le 2 oe

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 79
THRISSACANTHIAS PENICILLATUS (Fisher).
Pl. 17, fig. 4; pl. 18, figs. 1-5; pl. 53, figs. 1, la-e.
Persephonaster penicillatus FisHer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 297.

Diagnosis.—Rays five. R=176 mm.; r=35 mm.; R=5r. Breadth of ray
at base (between second and third superomarginals) 36 mm. General form flat-
tened; rays long, tapering from a narrow base and in small to medium-sized speci-
mens swollen a trifle above base; disk rather small, capable of slight inflation;
interbrachial angles rounded; abactinal integument thin, a trifle convex at base of
ray. Paxillee small, arranged in transverse spaced rows except on center of disk
and along midradial line, each with a cylindrical pedicel crowned with a brush-like
group of eight to twelve slender needle-like spinelets which are longer than pedicel;
papule everywhere except at tip of ray, conspicuous. Superomarginal plates
thirty-nine, strongly tumid, with a long, stout, tapering spine (sometimes two) in
center, and one or two smaller spines on upper edge; large spines sometimes bifid;
general surface of plate bristling with small spinelets; usually one or two pedicel-
lari present, formed of two to four small sharp spinelets in a circle, these very
inconspicuous and sometimes absent; occasionally present on inferomarginals.
Latter are opposite superomarginals; a few of proximal plates with a transverse
series of three stout tapering pointed spines, sometimes, but not always, a trifle
flattened and bifid; rest of plates including first two with two such spines; general
surface bristling with delicate spaced spinelets, increasing in size toward lower
end of plate and summit of transverse ridge; no true marginal fasciolar grooves; in-
stead plates slope up gradually from sutures without forming abrupt specialized
ridges, as in Astropecten, Psilaster, and others. Adambulacral plates with five or
six (in smaller specimens three or four) furrow spinelets rather delicate and com-
pressed, and forming a divergent palmate series, and on actinal surface an enlarged
usually flattened blunt spine with a variable number of smaller blunt membrane-
invested spinelets surrounding it or in two longitudinal rows. Mouth plates with
peculiar furrow series formed of an angular group of about five spinelets situated
at a higher level than enlarged tooth; that is, nearer peristome. Actinal inter-
radial areas small but plates extending far along ray, and armed with rather deli-
cate spaced spinelets, these sometimes forming a pedicellarian apparatus. Madre-
poric body large, free; gonads in several tufts (ten to twelve) extending a short
distance along ray; superambulacral plates very well developed.

Description.—The rays are decidedly long in this species and longer in large
than in small specimens. Abactinal integument thin, frequently collapsed on
outer part of ray onto the ambulacral ridge. Paxillee small to medium sized, dis-
tinctly and uniformly spaced; arranged in transverse well-spaced series on rays,
but without regular order in center of disk and along midradial line. The paxillx
decrease regularly in size distad and toward center of disk, and not so conspicu-
ously toward marginal plates, being largest on the radial areas at base of ray.
The transverse rows of paxille are quite regular, and between them is a double
row of conspicuous papule. Each paxilla consists of an elongate, elliptical base
set longitudinally with reference to long axis of ray, surmounted by a stout, cylin-

80 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

drical, convex-tipped pedicel whigh bears a crown of eight to twelve (type) very
slender, tapering needle-like spinelets, considerably longer than the pedicel, and
usually standing upright in a cylindrical coordinate group, highly characteristic.
Paxille in center of disk have irregular or circular bases. Small paxille are scat-
tered here and there on the radial areas with only three or four spinelets. In some
specimens the spinelets are thicker, due to the investing membrane. On the distal
portion of ray the tabulum or pedicel of paxillee becomes gradually much shorter,
and at tip of ray the paxille are convex spine-bearing plates. There is some varl-
ation in the diameter of the paxille but all specimens agree in having the char-
acteristic arrangement in transverse spaced rows. On an average three transverse
series correspond to one superomarginal plate. The paxille resemble, frequently,
pseudopedicellariw from the fact that the spinelets all converge and meet at one
point, forming a conical group.

The abactinal plates viewed from inner surface of dorsal integument are very
regular on the areas of regular paxille, where, as mentioned above, they are ellip-
tical in general form, with a suggestion of a lobe at either side in medium sized and
small individuals. In large examples the plates are very distinctly spaced all over
the dorsal surface, but in medium-sized specimens they usually touch, at least in
center of disk and along radial areas. The regularity of the plates is broken on
these areas; along the latter the plates become shorter and broader, often irregular,
while on the disk they are subcircular. The difference in appearance of the abactinal
plates in large and medium-sized examples seems to be due to the fact that beyond
a certain stage the paxille: do not increase in numbers in proportion to the increase
in size of the individual. Thus, an example with R96 mm. has a width of ray
at third superomarginal of 22 mm. and twenty-seven to twenty-nine paxille in a
row across the ray. The type, with R176 mm., and a breadth of ray at third
superomarginal of 36 mm. has only thirty-two to thirty-six paxille. As a result
the paxille are much more widely spaced in large specimens. The papule are
very conspicuous, and are distributed everywhere except at very end of ray. The
regular arrangement in double transverse rows is interrupted along radial line and
at center of disk, but papule are present in these regions. In many specimens the
papul are as large as the paxille in diameter.

Superomarginal plates, thirty-nine in number from interradial line, very tumid,
and not usually encroaching very conspicuously onto abactinal area proximally
of the middle of ray. Beyond this point the plates encroach more and more onto
abactinal surface, until considerably more of the plate belongs to the upper surface
than to side. Plates are subquadrate, except in interbrachial arc, where they are
much higher than long; abactinal margin of plates slightly rounded. There are
no fasciolar channels, but only a vestige of them, the surface sloping gradually
from the sutures toward the middle of the plate which bears a robust tapering
spine (or occasionally two), much longer than the plate (7 or 8 mm.) and directed
obliquely upward and outward. These spines are longest at about middle third of
ray, thence decreasing very gradually in length toward tip of ray. Many of them
are very curiously bifid for half or two-thirds their length, as if composed of two
fused spines, and, therefore, possessing two closely appressed points. Abactinal
margin of plate bears one or two robust, upright, tapering, pointed spinules or spines

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 81

(2 to 3.5 mm. long). When two are present, the adoral is commonly shorter;
in some examples these spinules are very short, scarcely more than 1 mm. General
surface of the plate bristles with very slender spaced capillary spinelets, largest
near the spines, these becoming papilliform in some specimens on account of a
pulpy membranous sheath. They are shorter than adjacent paxillar spinelets.
Near each upper corner of nearly all the superomarginal plates of type is a small
pedicellaria composed of usually four or five (sometimes only two or three) short
sharp spinelets, stouter than the rest, arranged in a circle about a common center,
or they may be more like the pectinate form characteristic of Cheiraster. The
position and number of these varies in different specimens. Sometimes there is
but one, near the upper marginal spine, and in specimens from the northern part
of range the pedicellarie are few, and many of the plates do not have any. Ina
small specimen (young) from station 2923 they are very conspicuous. Similar
pedicellariz occur less frequently on the inferomarginals, usually near or on the
intermarginal suture.

Inferomarginal plates corresponding in number to superomarginals, to which
they are opposite, forming a steep slightly arched bevel to actinal area. They are
also tumid, especially along the transverse axis, the tumidity passing into the
corresponding superomarginal without any conspicuous break at the suture between
the two. A few of the proximal plates bear a transverse series of three stout,
tapering, pointed spines, frequently a trifle flattened and bifid like those of upper
series; the rest of the plates, including usually also the first two, bear two such
spines, the upper the longer (8 to 9 mm.). Rarely on distal part of ray, where the
lower spine is occasionally much reduced in size, there may be either three or four
spines, due chiefly to the fact that a spine is split to its base, being in reality two
spines with a common articulating boss. These spines are all commonly appressed
to ray as in Psilaster. General surface of plates bristles with delicate spaced spine-
lets, which increase in size toward lower end of plates and toward summit of
tumidity. Pedicellariz, as described above, are occasionally present. Terminal
plate with apparently two spines.

Adambulacral plates set obliquely, broader than long at base of ray and longer
than broad near tip; furrow margin curved. Armature consisting of (1) a furrow
series of five, or less commonly, six (type) compressed, rather delicate, slightly
curved, blunt skin-covered spinelets, often capped with a knobby membranous
tip, arranged palmately, and usually graduated in length from the adoral to the aboral
end of series, or they may be graduated from the middle or the second spinelets.
There is considerable variation even in the same specimen in the number of furrow
spinelets. Some specimens have three or four; and the diameter of spinelets varies
according to the thickness of the investing membrane; (2) on actinal surface of
plate is an enlarged flattened blunt spine frequently with a shallow groove running
from tip half way to base, this surrounded by a variable number of smaller, tapering,
blunt or pointed skin-covered spinelets, arranged frequently in two longitudinal
rows, three or four to each, the spine standing in the inner series, or between the
two, in which case the spinelets appear to form a wide circle around it; outer
spinelets usually much flattened at tip and furrowed lengthwise as if incipiently

57444°—Bull. 76—11——6

82 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

bifid. Rarely a second spinule accompanies the enlarged spine, or none of the
spinelets may be flattened or furrowed, but tapering and fairly sharp, in which
case none of the marginal spines are flattened and bifid. Some (northern) speci-
mens have the enlarged ‘‘spine” very inconspicuous and only slightly larger than
the other spinelets. In small specimens there are but two or three actinal spine-
lets in addition to the spinule. Adambulacral armature is extremely variable,
both according to size and locality. ;

Mouth plates elongate, narrow, prominent actinally ; interradial length 13 mm.;
width of combined plates 5mm. At inner angle of combined plates are two stout
enlarged, somewhat curved, blunt spines, and the whole surface of the plates is
covered with short thick, blunt, very robust spinelets, which are largest near the
inner angle and decrease in size and thickness toward the margin adjacent to first
adambulacral. Furrow series very angular, and consisting of a group of spinelets,
about six in number, situated at a higher level than the tooth, and continued to
first adambulacral along the excavated short free margin in about four more short
pointed spinelets. The two actinal spines adjacent to teeth may be nearly as
large as the latter, making four teeth at angle, and one of the true furrow spinelets
is occasionally enlarged. In some specimens there is a regular and prominent series
or double series along the margin of the median suture, with a few slenderer spine-
lets along the sides, lower down, the actinal surface of plates being very convex.
The first adambulacral forms a short but wide companion plate, being much com-
pressed.

Actinal interradial areas rather small; intermediate plates small on rays,
larger on disk, extending in a single series 0.75 to 0.8 length of ray measured along
side; a second series extends 0.5 to 0.6 as far as first; a third about 0.25 to 0.33
as far as the second (to the third to sixth inferomarginal). Large specimens have
seven or eight plates in an interradial series opposite first inferomarginal. Super-
ficially the plates appear roundish and much as in Psilaster, Bathybiaster, and
allied forms. Viewed from the inner side, the plates are seen to imbricate slightly,
and are a trifle tumid and armed with a group of rather delicate, often flattened,
obtuse, grooved, occasionally bifid spinelets, similar near adambulacral plates to
their outer actinal spinelets, and forming thence all the transitions between these
and the lower spinelets of the inferomarginals. These spinelets are all sheathed in
membrane, which is frequently swollen at the tip; in some specimens the spinelets
have a papilliform appearance, and the calcareous part is slender and needle-like
without flattening or grooving. Not infrequently a central spinelet is enlarged,
and on some of the plates near the adambulacrals four or five spinelets form a
circular group, suggesting a sort of pedicellarian apparatus.

Madreporic body large (7 mm. in diameter), entirely uncovered, circular to
irregularly oval, the inner edge being about at the midddle of minor radius.

Color in life, madder brown.

Anatomical notes.—Very well developed superambulacral ossicles are present;
these are absent from first four plates and rudimentary near tip of ray. The gonads
ure not confined to the interradial region but extend a slight distance along ray,
there being ten to twelve independent clusters of tubules depending from each side

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 83

of a ray, forming a linear series. The genital stolon to which they are attached is
easily seen. Each cluster is composed of a single tube divided at base into two or
three long branches, the latter being subdivided into many, thick, irregular, often
grotesque side branchlets or pockets. Anal aperture present, connecting with a
large saccular intestinal cecum. Viewed from above this coecum is shaped like a
butterfly with rounded wings, there being a deep sinus in the outline, in the inter-
radius to right of madreporic body, and another shallower one in the left radius of
trivium (corresponding to anterior and posterior ends of butterfly). This coecum
is connected with spacious stomach by good-sized intestine. Tiedemann organs
small; Polian vesicles, one in each interradius. Tube feet pointed, without cal-
careous deposits; ampulle double. Interradial septa membranous.

Variations.—The principal variations have already been alluded to in the
foregoing description. These concern the number and relative thickness of paxille
spinelets; the relative size of the spinelets on upper edge of superomarginals, these
being usually smaller and two in number in northern specimens, frequently only
one in southern examples; the minute spinelets covering superomarginals, and the
number of pedicellariz, the former being thicker and the latter fewer in northern
examples, and more conspicuous in young than in very large specimens; the shape
of marginal spines, these being more often flattened and bifid in southern speci-
mens, and more often needle like in northern, the general appearance of actinal
spinulation, this varying from slender to papilliform, but never crowded; the
number of adambulacral spines and spinelets; mouth plates and their armature.
Some specimens have slightly wider and shorter rays than the type but the varia-
tion in this respect is inconsiderable.

Young.—The smallest specimen, from station 2923, has R=31 mm.,r=7 mm.,
and most of the adult characters well developed. The paxille are fewer, but in pro-
portion to size of specimen are larger than in adult. They have eight or ten spine-
lets, occasionally more. Marginal pedicellariz conspicuous; actinal intermediate
plates fewer than in adult. The general appearance is unmistakable after one
knows the adult form, but the superomarginals encroach conspicuously onto to
abactinal area. The general facies of the marginals reminds one of Cheiraster,
except for the fact that the plates of the two series correspond. The abactinal
paxille and distribution of papule are very different from Cheiraster.

Type.—Cat. No. 22329, U.S.N.M.

Type-locality — Albatross station 4380, off Los Coronados Islands, southwest
of San Diego, California, 530 to 638 fathoms, gray sand, green mud.

Distribution —Washington to Lower California (Los Coronados Islands) and
from 277 to 822 fathoms. Usually on green mud or fine sand.

84 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Specimens examined.—Thirty-four from the following stations:

Specimens of Thrissacanthias penicillatus examined.

|
|

Station. Locality. | Depth. | Nature of bottom. Number. Collection.
= _ —|— : =
Fathoms.
2300......| Off Heceta Bank, Oregon ..........------ | 277 pray sand. ........-.....------ 3 U.S. Nat. Mus.
2923......| Off Point Loma, San Diego, California ... 822 Teen MN sa s-.scsedecctee-= 5 Do.
2 Off Point La Jolla, San Diego, Callfornia..| 359 MM detest seen acters 8 Do.
2 Off Anacapa Island, southern California - . 388 green’mud).......-------.---.= 2 Do.
.| Off Sea Lion Rock, Washington.........- RTT oman Ose sacnees sem cocesee ten 2 Do.
Off Pigeon Point, California...........--- 296 | fine gray sand.. 2 Do.
| Off Destruction Island, Washington ..... | 516 green mud..-..... 1 Do.
| Off Cascade Head, Oregon.-.......-- at 345 TMU wees eae sae 2 Do.
| Off Point Arena, northern California | 455 fine gray sand 6 Albatross, 1904.
Off San Diego, California..........-..---- | 161-510 | fine gray sand, green mud, 1 Do.
| rocks.
4380..... Near Los Coronados Islands, Lower Cali- | 530-638 | gray sand, green mud........- al Do.
fornia.
ST a arora Off Santa Cruz Island, California... --.-.-- | 475-510 | black mud, broken stones...-.- 1 Do.
|

Remarks.—This species was originally described under the genus Persephon-
aster, although I was uncertain of the identification at the time. The radically
different arrangement of gonads which extend part way along the ray, on either
side, and the presence of a large actinal adambulacral spine are the two important
features which distinguish Thrissacanthias from Persephonaster.

Although the types of Persephonaster and Psilasteropsis have a different facies,
an analysis of the characters of several species-from Hawaii, the Philippines, and
Japan seems to warrant uniting the two genera. Typical Psilasteropsis has narrow
block-like inferomarginals which do not encroach far onto the actinal surface,
while in typical Persephonaster the inferomarginals are somewhat wider actinally
and the superomarginals possess stout spines. In other species of Persephonaster
the inferomarginals encroach conspicuously onto the actinal surface and the supero-
marginal spine is small. The arrangement of gonads, general character of adam-
bulacial armature, armature of inferomarginals, distribution of papule, armature of
mouth plates (especially the angular marginal series and presence of two teeth at the
inner angle of jaw), and arrangement of actinal intermediate plates are the same
in the two genera. Persephonaster is the older name. (See also Keehler, 1909qa.)

Tritonaster is probably a specialized offshoot of Persephonaster, distinguished
by the unusually small superomarginals, the very tumid inferomarginals (defining
ambitus), and by the rudimentary actinal intermediate plates of the ray.

Blakiaster also resembles Persephonaster, but is distinguished by the presence
of an odd interradial series of actinal intermediate plates, and lobed, imbricating,
low, convex tabulate abactinal plates, simulating paxille.

In Japanese waters there is a species, yet undescribed, which resembles penicil-
latus in general appearance, but differs in several important and many minor details.
This species has about three superomarginal spines in a transverse series; proximally
four or five inferomarginal spines with several auxiliary spinules; distally three; a
remarkable adambulacral armature consisting of a very angular furrow series of
three small spinelets set high in furrow (as in Henricia) the true margin being occu-
pied by four or five spinelets belonging to the first actinal series. The remainder of

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 85

the surface is covered with a few spaced spinules one or two of which may be enlarged.
This peculiarity of the true furrow spinelets is repeated in the furrow series of the
mouth plates. In penicillatus the furrow spines of mouth plates are also situated
high on the side of plate, which thus maintains an arrangement lost (or never
acquired) by the adambulacral plates.

Genus DIPSACASTER Alcock.

Dipsacaster AucocK, Ann. and Mag. Nat. Hist., ser. 6, vol. 11, 1893, p. 87; Journ. Asiatic Soc.
Bengal, vol. 62, 1893, p. 172 (no diagnosis). Type, D. sladeni Alcock.

Diagnosis—Rays five. General form depressed with broad disk and well-
developed actinal interradial areas; abactinal surface flat, not arched; marginal
plates large, the inferomarginals always broader than superomarginals and typically
extending laterally beyond them and forming a subserrate border to ray; fasciolar
channels deep and conspicuous, the ridges being correspondingly highly developed;
inferomarginals with a tuft of spines at outer end, or these may be exceptionally
absent, the superomarginals sometimes with one to three small tubercles; covering
of plates ranging from capillary spinelets to polygonal granules; paxilla highly
characteristic, composed of a tall pedicel springing from a round or stellate base and
crowned with a glomerular tuft of very many slender crowded spinelets; papule
typically distributed all over the abactinal surface, in sixes orfives about the plates; no
internal independent ossicles connecting plates; actinal intermediate plates carinated
and imbricated, bearing a paxilliform group of spinelets, there being fasciolar chan-
nels leading from inferomarginals to adambulacrals; adambulacrals not compressed,
with a palmate or pectinate furrow series of cylindrical or much compressed spines,
and a variable number of smaller spinelets on exposed surface of plate, the latter
usually not regularly arranged; mouth plates prominent actinally, rather broad,
with numerous spinules on exposed surface, and a marginal series resembling those
of adambulacrals; madreporic body typically very large, hidden by many large
paxille springing from its surface. Anus always present, connected with an
unbranched sac-like intestinal coecum, lying in the left radius of trivium (or adjacent
interradius of bivium), the latter connected with stomach by a short intestine;
superambulacral plates always present; tube feet pointed, without calcareous
deposits; gonads extending far along ray, the genital tubules depending from a
genital stolon and decreasing in size distad.

KEY TO THE SPECIES OF DIPSACASTER HEREIN DESCRIBED.

a', Inferomarginal plates with a tuft of enlarged spines at outer end or a series along aboral margin;
border of rays subserrate; papule distributed all over abactinal surface; madreporic plate very
large bearing many paxillz on its surface.

b'. Paxille conspicuously larger in central portion of disk and along median area of rays, their spinelets
descending the pedicel in bristling array, so that they resemble minute bottle-brushes; rays broad
near tips; abactinal plates strongly stellate along median radial area, with two or three papuli
to each area instead of one; actinal intermediate areas broad, far along ray-....-- eximius, p. 86.

b?. Paxillz not conspicuously larger in central portion of disk, etc., their spinelets grouped at top of
pedicel; rays evenly tapered to tip; abactinal plates along median radial area either round or
stellate, but not surrounded by many papule, never by more than eight; actinal intermediate
areas narrow beyond middle of ray.

86 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

c!. Abactinal plates lobed; marginals broader, covered with compact, rigid, polygonal granules
(superomarginals) and squame (inferomarginals); superomarginals with one or two tubercles;
adambulacral furrow spines compressed with broad side to furrow; actinal intermediate plates
covered with squamiform spinelets.....--.-.-------------++++++++rteeeeeeeeee borealis, p. 91.

c. Abactinal plates not lobed; marginals narrower, the superomarginals without tubercles and
covered with delicate capillary sp.nelets; inferomarginals with delicate imbricating narrow
squamiform spinelets; adambulacral furrow spines strongly compressed with edge to furrow;
actinal intermediate plates with slender spinelets.........-.------------- letmophilus, p. 95.

a?, Inferomarginals entirely without specialized spines; border of ray even, not serrate from the form of
inferomarginals; papule absent from center of disk; madreporic plate median sized, hidden, but

with only one or two paxille on its surface.....-.---.-----+-++++++++2----- 220s anoplus, p. 97.

DIPSACASTER EXIMIUS Fisher.
Pl. 11, figs. 1, 2; pl. 13, fig. 2; pl. 14, fig. 1; pl. 16, fig. 3; pl. 52, figs. 1, la-c.
Dipsacaster eximius Fisuer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 296; Zool.
Anz., vol. 35, March 29, 1910, p. 546.

Diagnosis.~—Rays five. R=124 mm.; r=55 mm.; R=2.26 r. Breadth of
ray at base, 60 to 65 mm.; at middle of ray, 46 mm. General form flattened, rays
broad, leaf-like in shape, tapering gradually, but with an outwardly curved contour;
extremity bluntly pointed; interbrachial angles wide, abruptly rounded; abactinal
surface but slightly inflated, a slight depression on interradial lines; actinal surface
subplane. Abactinal paxille very regularly arranged in chevrons, largest along
midradial area and center of disk; each with a large tuft of very numerous slender
spinelets which extend down pedicel for half its length, or more; regular arrangement
interrupted on area of largest paxille. Papule generally distributed, six around
each lobed plate, except along radial line where there are two or three papule to
each area, and twelve to twenty about many of the plates. Marginal plates massive,
the inferomarginals defining contour of ray, and armed on outer end with a trans-
verse or vertical row of four to six short stout tubercles; superomarginals usually
with a single short tubercle; occasionally more. Adambulacral plates with five or
six strong furrow spines, the median longest and all more or less compressed, with
edge uppermost; and on actinal surface two or three rows of three or four spinelets
each; usually only the inner actinal row has any regularity. Actinal intermediate
areas large, and maintaining a considerable breadth far along ray; intermediate
plates strongly carinated bearing a cluster of spinelets of which the peripheral are
slenderest, those in center not typically much enlarged although occasionally so on
some specimens.

Description.—The rays in this species have in large examples, a very character-
istic broad leaf-shape. That is, they taper only slightly from base to outer third,
thence the edges curve and taper rapidly to the blunt extremity. Paxille large;
largest in center of disk and along median radial areas, thence decreasing gradually
and uniformly in size toward the marginal plates. They are very regularly arranged
in oblique transverse rows along either side of the median radial area, forming
chevrons pointing toward center of disk; but on latter area and along radial line the
paxillee, due to their becoming larger, have no regular arrangement. Opposite or
corresponding to first four superomarginals, in large specimens, are eight to ten
transverse rows of paxille. Each paxilla springs from a lobed plate (described

2 From a large specimen from station 3112.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 87

below), the shaft with spinelets being 3.5 to 6 mm. high in larger specimens. This
shaft or pedicel is surmounted by a globular or elongate crown of very numerous,
slender spinelets sheathed in delicate membrane, the whole forming a dense glomeru-
lar tuft, the central or uppermost spinelets bluntly tipped, but those on the side
mucronate, less crowded, and usually extending in a bristling arrangement down the
sides of pedicel, for nearly to over one-half its length. In the latter case the paxilla
resembles a minature bottle-brush. In a dried specimen the centrally situated
spinelets are much sharper, due to a shrinkage in the membranous envelope.

The abactinal plates or bases of paxille are strongly lobed on center of disk and
along median radial area, the lobes becoming less prominent toward margin of abac-
tinal area. Some plates have six, some four, some five lobes, and they are very irreg-
ular. On the regular lateral rows of plates there is more uniformity, but here some
plates have five divisions while the majority have four. Near center of disk numer-
ous plates are rounded or irregular without evident divisions. In another specimen
examined the plates are much more regular especially along the lateral rows. Here
they touch or slightly overlap by lateral lobes, which does not take place in the first
example. Some plates near the margin of interradial area are nearly square.
Papule are distributed all over abactinal paxillar area, from tip of ray to center of
disk. There are usually six about each plate, but where the plates are not arranged
in regular rows there are often more, especially in the case of the median radial series.
When viewed from the outer surface, the papular areas are seen to be larger here, and
to contain two or three papule, instead of one, as is usual. Thus a carinal or radial
plate may be surrounded by twelve to eighteen or twenty papule in groups of two or
three, these papulx being of course common also to several adjacent plates.

Marginal plates massive and regular, the inferomarginals defining contour of
ray, and extending laterally beyond superomarginals for a variable distance which is
sometimes (as in type) equal to width of latter. Superomarginals, over thirty
(thirty-two to thirty-four) in large specimens, are slightly wider than long, and
shorter in proportion to width in interbrachial angle than farther along ray. They
are completely confined to abactinal surface, and form an arched bevel to margin of
paxillar area. The exposed surface which is slightly tumid, is covered with small
but robust polygonal flat-topped spinelets in middle, these rapidly becoming slen-
derer toward edge of elevated ridge, where they are either blunt or sharp; and are
longer on outer than on inner edge of plate. Nearly all the plates bear a tubercular,
short, polygonal stumpy spine either near middle or in outer aboral quarter of plate.
In some specimens there is a group of two or three such tubercles, which do not ex-
tend conspicuously above the general covering of plate, except in rare cases, but
are larger on outer part of ray than on proxinal portion.

Inferomarginals correspond exactly to superomarginals, are much wider than
long, and form a conspicuous border to actinal surface. On outer end of each plate
(or rather the specialized ridges) is a transverse series of four to six stout tubercular
spines the upper or next to upper longest, forming an armature on margin of ray. In
some specimens (usually small to medium-sized) these spines are tapering and rather
sharp, but usually on large examples are truncate and thicker at tip than at base;
and the tip is also usually truncated obliquely, due to the spines being bent toward
tip of ray. Near base of ray the thick spines may have the tip curiously hollowed

SS BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
o

out like a cup, instead of being cut square off or slightly rounded. General covering
of plate consists, on actinal surface, of ovate or oblong lanceolate spinelets, which
become capillary in fasciolar furrows; outer end of plate bristles with slender, sharper
lancet-like spinelets those on sides of the broad fasciolar grooves, frequently with fine
mucronate tips. Spinelets adjacent to the spines are enlarged. The fasciolar
grooves between plates are very deep, the specialized ridges being about as thick,
when denuded, as the adjacent grooves. These ridges, when viewed from actinal
surface, extend laterally beyond the bottom of groove a distance equal to the width
of the plate, measured from bottom of groove to inner (actinal) edge.

Terminal plate small, granulous, as wide or wider than long, not notched toward
paxillar area. |

Adambulacral plates fairly massive with a curved furrow margin. Armature
consists of (1) a furrow series of six strongly compressed, round-cipped spines, the
two centrally situated the longest (3.75 to 4 mm.) and the others graduated in length.
Commonly the edge of spine toward furrow is broader than that from it. Frequently
there are but five large spines (rarely but four), the sixth being much smaller than
the rest and placed back slightly from the edge at either end of series. Rarely
seven furrow spines are present. (2) On actinal surface are two or three irregular
longitudinal series of much smaller and slenderer spinelets which decrease rapidly
in size as they recede from furrow. Series adjacent, to furrow spines consists of about
three or four tapering pointed spinelets, one-half to two-thirds the length of median
furrow spines. The outer one or two series are very irregular; sometimes there is
no regular serial arrangement at all. The average plate is best shown by the figures.

Mouth plates large and prominent actinally, the combined pair broadest at
about the middle. Armature consists of (1) a furrow or marginal series of seven to
nine, similar in character to median spines of adambulacral furrow series, which begin
at about middle of plate and form a fairly straight series to inner angle, increasing in
size as they proceed inward, the inner two or three being broad, flat, and knife-like
and in common with the others the edge away from furrow is often thinner than that
toward it. (2) Actinal surface is covered with slender spaced spinelets which
increase in size toward the suture and inner angle.

Actinal intermediate areas large, the plates being arranged in regular series
parallel with furrow, and also in regular series leading from marginal to adambulac-
ral plates. The breadth of ray near extremity is largely due to the intermediate
plates, of which four longitudinal series extend 0.39 length of ray (measured from
interradial angle); three series, 0.65 or slightly more; two series, 0.7 to 0.8; and the
series adjacent to adambulacrals extends about half way between end of latter and
tip of ray or about 0.85 total length. In the interbrachial angle there are additional
short series; for instance, in a large specimen, a fifth series extends nearly 0.25 length
of ray, a sixth to the third or fourth inferomarginal; and counting along interradial
line, there are nine or ten plates between the first inferomarginal and third adambu-
lacral (see pl. 13, fig. 2). The plates are strongly carinated, the keel running trans-
versely, each being surmounted by a prominent paxilliform tuft of spinelets, of which
the peripheral are very slender and pointed, while the central are usually much
stouter, often clavate and three to five sided in section, with flattened flaring tips
cut obliquely, which appear to be bent outward toward marginal plates. In such
specimens the spinelets of series adjacent to adambulacrals are heavier, often sub-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 89

prismatic also, and square tipped. More often the central spinelets are only slightly
heavier than the peripheral; or one or two may be enlarged, and the rest graded in
size up to these. Between these actinal paxille are deep channels leading from the
marginal grooves to the furrows. The plates, from which the carinations spring,
are seen to imbricate very strongly when viewed from the ccelomic side, the outer
end being often prolonged. Considered from the external surface the latter lies
under the inner end of the adjacent plate (in interradial direction).

Madreporic body large (about 9 or 10 mm. in diameter), undulating in outline
and hidden by large paxille situated on its surface (about eighteen). It is situated
a little adcentrally to the middle of the minor radius.

Color in life: Abactinal surface orange yellow, actinal surface lighter yellow.

Anatomical notes.—Superambulacral plates are present as slender rods, slightly
flaring at tips, which pass from outer end of ambulacral plates to the first or inner
actinal intermediate plates; absent from first two plates, and very small neartip of ray.
Gonads, in the form of closely placed dichotomously branched tufts, depend from the
genital stolon, extending in a series half the length of ray. Each genital tube con-
sists of two main branches and two or more terminal branchlets. In the specimen
examined they pack the celom full. These genital stolons are situated on either
side of the median radial area slightly less than midway to margin of paxillar area.
Anal aperature prominent, with a narrow periproct. It is connected by a short
intestine with the large single sack-like intestinal ccecum lying in the left radius of
trivium. The lining of the cecum is traversed by small folds or corrugations.
Directly under the anal aperture of ccecum is a slit about 2 mm. long opening into a
very short intestine, which leads into the stomach. The latter has much folded
walls, and these appear to be no differentiation into dorsal and ventral divisions.
Hepatic cceca are large with a spacious interior, which is pinnately divided into simple
side pockets with more or less folded walls. A large Polian vesicle in each interra-
dius. Tiedemann organs, two to each interradius, large, and with many fine
divisions. Interradial septa single, without calcareous bodies. Stone canal very
large. Tube feet large, conical, without calcareous rods in the walls.

Variations.—In medium-sized examples—that is, with R equal to about 70
mm.—the rays are not so broad on the outer part; consequently they seem to
taper more evenly. The actinal interradial areas are also smaller, and on outer part
of ray there are not so many intermediate plates. The abactinal paxille are not so
crowded, nor of course so large. Irrespective of size there is a slight variation in
marginal plates. The inferomarginals project laterally beyond the superomarginals
in a varying amount, and the spines or tubercles are bluntly pointed in some speci-
mens, in others clavate with truncate or concave tips. Some specimens have as
many as three superomarginal tubercles, especially on outer part of ray, where they
are longer than proximally. The actinal intermediate paxille are rather variable,
the central spinelets being usually slender (but stouter than the peripheral) and
slightly spaced on the keel or tabulum of plate; a very few specimens (two or three)
have the central spinelets conspicuously thicker than the peripheral ones, polygonal
in section, with truncate tips, these being graduated in size from one or two central
spinelets to the peripheral. The spinelets on surface of mouth and adambulacral
plates are also variable in arrangement.

90 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
The following table will show some of the variations due to size:

Variations in Dipsacaster eximius.

| Second
Number | series of ;
of super- | actinal Third Fourth

Station. | R. Me Rer. omargi- | interme-| series.a | series.a
nals. diate
| plates.e
= —
3112 124 55 2.26:1 32-34 0.78+ 0.65+ 0.39
4334 108 45 241 32 0.8 0.5 0.24
4321 7 31 2.5 :1 28 0.7 0.41 0.17
2937 30 |' 11 ane 24 0.34 0b Oc

2 The figures denote how far along ray, measured from interradial angle or first plate, the intermediate plates extend. Thus
the second Ae of actinal intermediate plates in the first specimen extends 0.78 length of ray, the third row 0.65, the fourth 0.39.

+ To second inferomarginal.
¢ Does not extend beyond first inferomarginal (four plates In all).

Young.—The smallest specimen from station 4333 has R=13 mm., and is easily
recognized as belonging to eximius, and as distinct from D. anoplus by reason of
the spiny inferomarginals extending conspicuously beyond superomarginals, by the
wide fasciolar grooves, character of adambulacral and actinal intermediate plates.
The abactinal paxille are high delicate and more spaced as well as relatively fewer
than in the adult. The papule are generally distributed as in adult, but not gonads
have developed. Adambulacral furrow spinelets four, and only four to six very
delicate spinelets to each intermediate plate. The spinulation is open and lacks
the compact appearance of D. anoplus. (See pl. 16, fig. 3.)

Type.—Cat. No. 22328, U.S.N.M.

Type-locality— Albatross station 4334, Los Coronados Islands, southwest of
San Diego, California, 525 to 514 fathoms, green mud.

Distribution.—From north of Monterey Bay (Pigeon Point) to the vicinity of
San Diego, California (Los Coronados Islands, Lower California) and from 206 to
525 fathoms, usually on green, black, or yellow mud.

Specimens examined.—A total of twenty-two specimens from the following
Californian localities:

Specimens of Dipsacaster eximius examined.

Station. Locality. | Depth. Nature of bottom. Number. Collection.
| Fathoms:
2896......| South of Santa Rosa Island................. | 376 yellow: mud).. 252...-c-se~es 2 U.S. Nat. Mus.
ee | Off Point La Jolla, San Diego............... 359 MUG beer ees se oe eee 5 Do.
Man Raita en Between Santa Catalina and Oceanside... 464 Breen MUG sa. \eic eee 1 Do.
Mass: | Off Anacapa Island...................-..... S88) |: ~-. GO so eve onc es dene cee 1 Do.
a112...-.. | Off Pigeon Point, north of Monterey Bay...| 296 fine gray|sand..........:...4 3 Do.
MST oo. 22: [OM Ban Diego ee... 2202-22 wcce tess ee 1) 5 2065) s|iblack mud’. --.<.---es eee 2 Albatross, 1904.
S558 oc wobeleas an Cn eee asee eee ees eenes sees 301-487 | green mud.................. 1 Do.
S334,...3. Near Los Coronados Islands, southwest of | 525-514 |..... CO ntcavncacs scicceecee aes Type. Do.
San Dtego.
MME Seon wees 0 Oe Setecah cere G ess sce ace 625 Gray SANG Feet oe ears z Do.
423...... Between Santa Barbara and San Nicholas | 339-216 shells, black pebbles, green 1 Do.
Islands. | sand.
|
MIT occas Off Santa Cruz Island....................... | 475-510 | black mud, broken stones... 4 Do.
= |
@ Error.

> Probably erroneous, or specimen incorrectly labeled.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 91

DIPSACASTER BOREALIS Fisher.
Pl. 12, figs. 1, 2; pl. 13, fig. 1; pl. 14, fig. 2; pl. 52, figs. 2, 2a, b.
Dipsacaster borealis FisHer, Zool. Anz., vol. 35, March 29, 1910, p. 546.

Diagnosis.-—Similar to D. eximius in general form, but rays more evenly
tapered. Rays broad at base; interbrachial arcs abruptly rounded; abactinal sur-
face nearly plane. R=117 mm.; r=51 mm.; R=2.3 r. Breadth of ray at base,
60 mm. (greater than r). Abactinal paxille regularly arranged as in erimius, but
smaller, and not conspicuously larger along median radial area and center of disk;
in latter area they may be even a trifle smaller; spinelets not so numerous as in
eximius and not extending far down pedicel of paxilla. Papule generally distrib-
uted, six around each lobed plate; median radial series of plates not conspicuously
lobed as in eximius, but placed close together; not surrounded by more than six
papule. Marginal plates massive, the inferomarginals defining contour of ray;
superomarginals covered with low granules compactly placed, and on distal portion
of ray, bearing one or two inconspicuous tubercles; inferomarginals covered with
imbricating scale-like granules becoming more or less flattened tubercles on outer
aboral edge; the covering much more compact and resistant than in eximius; spe-
clalized ridges of inferomarginals thicker, and fasciolar furrows narrower than in
that species. Adambulacral plates with four to six long, stout, round-tipped furrow
spines, the four largest flattened with side to furrow; on actinal surface about six to
nine spinelets in two series, or without regularity. Actinal intermediate areas
smaller than in eximius, especially on rays; the plates strongly carinated, bearing a
compact group of broad scale-like spinelets, surrounded by a row of slenderer curved
spinelets.

Description. By comparing the largest specimen with that of D. eximius, the
rays, although very broad, are seen to be less arched in contour; that is, they taper
more evenly. The paxille are decidedly less compact than in eximius and more
uniform in size across the disk or across the ray. They are arranged in regular
oblique transverse rows except along the median radial area and center of disk. But
unlike those of eximius they are not larger on these areas, being, if different, slightly
smaller in center of disk. Thus when the two species are placed side by side the
abactinal area has a different facies. It is true that the paxille along the median
radial area are the largest on ray, but they grade so evenly into the lateral paxille
that the difference is not so noticeable as in eximius. The paxille themselves are
individually smaller than in eximius and the spinelets fewer. ‘These are grouped in
a compact bunch at the tip of the shorter pedicel, are seventy to eighty in number
on largest paxille, and do not extend very far down the pedicel. The median spine-
lets are blunt, but those on sides are pointed and successively shorter. The spine-
lets are relatively shorter. Probably a comparison of the figures (pl. 52, figs. 1a, 2a)
of representative paxille drawn on the same scale will show the striking difference,

_as well as the fundamental or “‘generic”’ similarity. The paxillee covering madreporic
body are two or three times larger than the others. Opposite first four superomar-
ginals are eleven to thirteen rows of paxille.

@ From type.

92 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Abactinal plates or bases of paxille small, and nearly the same size as in D,
eximius, those of lateral regular series if anything a trifle larger and more regular.
These commonly have five or six very short lobes, which become absolete at tip and
toward margin of area where the plate changes from a circular to an elongate con-
tour. The plates of the median radial region are rather more regularly arranged
than in eximius and are closer together. The carinal or midradial plates are
circular with very slight lobes on some of them, and adjacent plates touch them so
that the papular ‘areas are small, with single papule. (In eximius the carinal plates
are spaced and strongly lobed with larger papular areas than ordinary.) On
the disk, especially toward the center, the plates are much smaller than in eximius.
They decrease rapidly in size toward the center where they are scarcely larger than
the diameter of pedicel or paxilla shaft, circular in outline and widely spaced.
Along the course of either large abactinal adradial muscle band the plates form two
regular longitudinal rows as shown in figure, and their shape is also slightly altered.
Papule are distributed all over the abactinal surface, as is characteristic for this
genus. They are usually in sixes about each plate, and are never more than one to
an area on either side of the mid-radial series of plates, as is the case in D. eximius at
about the middle of the ray.

The arrangement of the midradial row of plates and adjacent papule, as well as
the form of plate, and their size toward center of disk, constitute some of the most
reliable differences between this species and eximius. These are best shown by
the figures (pl. 52, figs 1c, 2b).

Marginal plates massive, the inferomarginals defining the margin of ray when
viewed from above. Superomarginals, thirty-four in number from interradial line,
are wider than long and form a slightly arched bevel to abactinal surface. They
are not quite so strongly arched as those of D. eximius, nor are the plates so tumid;
and in proportion to length are a trifle wider. The plates are covered with low,
flat-topped close-set hexagonal and pentagonal granules, largest near center or
upper end, becoming slenderer toward edge of plate, where they assume the form
of robust, blunt spinelets, never slender and sharp as in erimius. Beyond middle
of ray on each plate one to three granules are usually enlarged into short, bluntly
pointed tubercles, and are situated on aboral edge of plate, being directed toward
edge of ray. On the more distally situated plates the granules increase in size
toward this tubercle, which increases in length toward the end of ray but is never
conspicuous. The aboral transverse edge of the specialized ridge of plate becomes
in this region slightly more abrupt than the adoral edge which is rounded. This is
due to the crowding of the large granules toward the aboral edge. The whole
covering of the plate is more compact, firm, and graniliform than in ezimius. The
fasciolar grooves between the plates (or specialized ridges of them) are not so broad
as in eximius, and unlike the latter species, are not evident unless the plate is
denuded.

Inferomarginals correspond to superomarginals and encroach conspicuously
onto actinal area, about which they form a slightly arched border, the outer ends
forming a rounded serrate edge to ray. The specialized ridge of each plate, when
denuded, is much thicker than the adjacent furrows, and is relatively thicker
and the furrows are relatively much narrower than in erimius. The plates are

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 93

covered with robust squamiform granules, which increase in size toward the aboral
edge of plate, the squamules being pointed outward and aborally. On the
outer part of the aboral edge (on margin of ray) are a number (six to twelve) of
pointed squamiform or tubercular spines directed distad, and forming a fringe
overhanging the proximal border of adjacent plate. They are stoutest and bluntest
on the very edge of ray and are graduated in size actinally, sometimes extending
nearly to inner edge of plate on the distal portion of ray. Frequently they form
two irregular series. The squamules of plate are graduated in size up to these.
Tn interbrachial angle where the plates are much shorter the spines form a median
row. On the abactinal end of the plate are small granuliform spinelets similar to
those on adjacent portion of superomarginal; and on the edge of the fasciolar
grooves are slender blunt spinelets. In general, the covering of the inferomarginals
is much more compact, heavy, and squamiform than is that of eximius. The
difference is very striking on comparison of the two forms.

Adambulacral plates about as long as broad, with a rounded furrow margin, and
separated by fairly wide sutures. Armature consists of (1) a furrow series of four,
five, or six robust round-tipped spines, which are much longer than plate, and are
graduated in size toward either end of series. These spines, at least the four central,
are usually more or less flattened with the broad side to furrow, and the tip is
usually expanded. (The same spines in eximius are strongly compressed, tapering,
and with edge to furrow.) Commonly, the spine, at either one end of the series or
the other, or both, is tapering, much smaller (two-thirds to three-fourths length of
large spines), and stands back from the margin, slightly out of line. When there
is a smaller spine at both ends of series, there are usually six spines in all. Near
extremity of ray there may be but three spinelets in furrow series. (2) On actinal
surface of plate are six to nine rather small-spaced, tapering, blunt spinelets, which
form two irregular series, or are without definite order. Those nearest margin are
longest. The lateral members of the furrow series frequently stand back far
enough to be in line with the actinal spinelets. The latter, however, are smaller and
distinctly tapered. The first plate is not much modified, and has six or seven fur-
row spines and a variable number of actinal spinelets. The outer edge of plate
abuts against the side of the mouth plate. It is not so compressed as in eximius.

Mouth plates prominent, broader at inner than at outer end, with a curved
margin at mouth of furrow. Furrow spines in a curved series, eight or nine in
number, increasing in size toward the inner end of plate; all are heavy, usually
with thick tips, the inner ones being compressed in varying degrees. The spines
are not so thin and knife-like as in eximius. A series of very small spinelets borders
the edge next to first adambulacral, and another, consisting of short but thick
and blunt spines, follows the median suture border. These lack regularity in size
and shape, but are often quadrate in section, and much thicker than the same
spinelets in erimius. A few intermediate in position between the two series may
be present.

Actinal intermediate areas considerably smaller than in erimius, especially on
ray. A single series (that next to furrow plates) extends 0.8 length’ of ray meas-
ured along its side; second series, 0.5; third series, 0.23; fourth series, less than

94 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

0.18; while a fifth series does not extend beyond the third inferomarginal. (Com-
pare with table under eximius.) Between first inferomarginal and third adambu-
lacral are seven plates. The series meeting the first and second adambulacrals do
not reach the marginals. Intermediate plates bear a marked tabulum or carination
crowned with a very compact group of numerous (about fifteen to twenty) round-
tipped, flattened, scale-like spinelets directed toward marginal plates; surrounding
this elliptical or ovoid group of imbricating spinelets is a peripheral series of about
fifteen to twenty-five slightly curved, round-tipped, slender, flattened spinelets
resembling flower petals. Indeed, the whole group greatly resembles an asymmet-
rical miniature chrysanthemum blossom. The central spinelets are much less
numerous in erimius and never of the flattened imbricating squamiform shape.
The furrows running interradially between the actinal paxille are narrower in
the present species than in the foregoing.

Madreporie body very large (13 mm. in diameter) with fine centrifugal stria-
tions. It is subcircular, with a scalloped edge, and its center is about midway
between inner edge of superomarginals and center of disk. The surface is uneven,
and is entirely obscured by about thirty to thirty-five very large paxille. One
of these exceeds the rest in size.

Anatomical notes—The internal anatomy does not differ in any very impor-
tant respects from that of D. eximius: superambulacral plates the same, as also
the details of the water vascular system. The intestinal cecum is the same shape,
but lies in the left bivial interradius (very nearly the position in eximius); anus
present, of conspicuous size; connections as in eximius. Gonads arranged as in
eximius, but in specimen examined they extend farther along the ray, beyond
middle.

Variations.—The variations exhibited by the few available specimens are
slight and concern chiefly minor details of ornamentation already mentioned.
One specimen from station 3331 has the rays a trifle shorter and broader at the
base than the rest. Despite small differences in spine counts, the type facies is
maintained by all the specimens, and one can distinguish them at a glance from
exumvus.

Type.—Cat. No. 24341, U.S.N.M.

Type locality — Albatross station 3331, north of Unalaska, 350 fathoms, mud
(four specimens).

Distribution.—Bering Sea and south of the Aleutian Islands, 121 to 351
fathoms.

Specimens examined.—Seven specimens, besides four from type locality, one
from station 3330, same locality, 351 fathoms, black sand, mud; two from station
3337, south of Unimak Island, 280 fathoms, green mud; one from station 3489,
west of Pribilof Islands, Bering Sea, 184 fathoms, green mud; one from station
3500, south of the Pribilof Islands, 121 fathoms, fine gray sand.

Remarks.—As detailed in the foregoing description, this species differs from
Dipsacaster eximius in the shape of rays in large specimens, in having smaller
paxille, which have fewer spinelets, and these not descending far down the pedicel;
in not having the paxille conspicuously enlarged on center of disk and along middle
of ray; in the shape and position of median radial plates and in number of papulee

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 95

surrounding them; in covering of marginal plates; in having thicker specialized
ridges to inferomarginals, with correspondingly narrow fasciolar furrows; in the
special spines on both series of marginals; in having differently shaped adambu-
lacral furrow spines; and in having much smaller actinal intermediate areas on rays,
as well as a different armature to the intermediate plates. Still other differences
have been alluded to in the course of the description.

A fairly well-marked variety, or perhaps representative species, is found in Japan-
ese waters. The principal external differences are the following: The Japanese
form has the paxille still smaller on middle of disk, a larger madreporic body (pos-
sibly not constant), smaller and less compact granules on superomarginal plates;
thinner specialized ridges to inferomarginals (more as in eximius) with wider
fasciolar grooves. The furrow spines are nearly as in borealis, but there is frequently
an additional spine. Thus the number is usually six or seven, and occasionally
eight. The most important difference is in the abactinal plates, so far as my
limited examination can determine. In the Japanese form the plates of ray are
also lobed, but the median radials or carinals are not so closely placed as in borealis.
They are also frequently decidedly lobed. The plates along the course of the
muscle bands are not modified. The lobing is apparently more distinct in medium-
sized than in very large specimens of Dipsacaster.

DIPSACASTER L#TMOPHILUS Fisher.
Pl. 12, fig. 3; pl. 15, figs. 1, 2; pl. 52, figs. 3, 3a, 3b; pl. 53, fig. 2.
Dipsacaster letmophilus Fisner, Zool. Anz., vol. 35, March 29, 1910, p. 547.

Diagnosis.—Rays five’ R=118mm.;r=45mm.; R=2.6r. Breadth of ray at
base 47 to52mm. Rays tapering gradually to a bluntly pointed extermity; general
form flattened; interbrachial ares wide; paxille regularly arranged as in preceding
species, and similar; slightly larger in center of disk than elsewhere. Abactinal
_ plate not lobed, except rarely on median radial area; papule generally distributed, in
sixes about each plate. Marginal plates massive; superomarginals nearly quadrate
beyond proximal fourth of ray, covered with close-set delicate spinelets like those
of paxille and without specialized spines or tubercles; inferomarginals narrower
than in preceding two species, covered with delicate imbricating narrow squamiform
spinelets which become enlarged into five to seven small tapering spines on outer
aboral end of plate, causing a decidedly serrate appearance to edge of ray; fasciolar
furrows rather wide. Adambulacral furrow spines seven very strongly compressed
and blade like, their edges to furrow; tip much broader than base, rounded and
thin like a knife; on actinal surface ten to eighteen slender tapering or clavate
spinelets, decreasing in size outward. Actinal intermediate plates extending far
along ray; intermediate areas about as extensive as in borealis, but a single row of
plates extending 0.9 length of ray measured from interradial angle.

Description.—Rays are narrower than in either of the two preceding species,
but broader than in D. nesiotes Fisher. Paxille regularly arranged along sides of
area, slightly larger in middle of disk than elsewhere (except on madreporic plate) ;
along mid-radial area not markedly larger than along either side contiguous to this
area. Paxille very similar to those of D. borealis, the numerous spinelets (eighty

96 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

to over one hundred on large paxillx) slender and blunt, the most centrally situated
ones forming a flat-topped group, those about periphery forming one or two irregular
tiers graduated in length (see pl. 52, fig. 3b). Ten or eleven transverse rows of
paxille correspond to first four superomarginals.

Unlike the two preceding species, the abactinal plates are not lobed, except
occasionally very slightly on mid-radial areas. (In D. nesiotes the abactinal plates
are circular, but placed more closely than in the present form.) Plates of the
regular lateral paxillar rows are subcircular, but the slight flange above the true
base of the plate or paxilla is frequently rather irregular, On the radial areas the
plates are more widely spaced, uneven in contour, with slight traces of lobes some-
times, and without regularity in arrangement. Papule distributed all over paxillar
area; in sixes about plates, but occasionally on the irregular median radial area
there are seven or eight.

Marginal plates typical of genus; inferomarginals defining contour of ray,
forming a serrate edge. Superomarginals, thirty-five in number from interradial
line, are nearly quadrate except in interradial angle, where they are much wider than
long. General surface is covered with slender and delicate, pointed spinelets about
the size of those on paxille, forming a close nap very unlike either D. borealis or D.
erimius, but similar to D. nesiotes. No specialized tubercles on superomarginals.

Inferomarginals correspond to superomarginals, and do not encroach onto
actinal surface so much as in D. borealis. Surface is covered with delicate, imbri-
cating, narrow squamiform spinelets, which increase in size toward the outer aboral
edge of plate, there forming a tuft of five to seven small tapering spines, as in D. nesi-
otes. The abactinal end of plates is covered with slender spinelets like those of
superomarginals. The fasciolar furrows are wide and lined with delicate terete
spinelets. In proportion to length, the plates are much narrower (and are actually
so) than in erimius or borealis, but are shorter and wider than in nesiotes. Com-
pared with the latter species, the fasciolar grooves are much wider and the
specialized ridge not so thick.

Adambulacral plates with a furrow series of seven very strongly compressed
bladelike spines, their edges to furrow, and graduated toward either end of series.
These spines are much broader and thinner at tip than at base, the edge at tip being
rounded, laminar, and knifelike. On actinal surface are about ten to eighteen
slender, tapering or clavate spinelets decreasing in size outward. Those nearest
furrow are subequal and largest, and commonly form a longitudinal series of five
to eight.

Mouth plates elongate, the combined pair very tumid and the whole surface
bristling with spaced spinelets similar to those on actinal surface of adambulacrals.
Free margin with seven to nine compressed spines like those of adambulacrals, these
increasing in size toward the inner end of plate.

Actinal interradial areas fairly large, the intermediate plates about as numerous,
and intermediate areas as extensive on rays, as in borealis. The series adjacent to
adambulacrals extends 0.9 length of ray; the second, 0.6 to 0.65; the third, 0.3;
the fourth, 0.15; a fifth extends to fowrth inferomarginal; between third adambu-
lacral and first inferomarginal is an interradial series of nine plates, the additional
plates being added rapidly between first and fourth inferomarginal. The first row

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 97

extends very slightly farther along ray than in either of the two preceding forms.
Actinal plates bear a low carination (of narrow elliptical form) crowned with small
papilliform spinelets, those in center being clavately thickened and fifteen to thirty
in number, the peripheral, about twenty or less according to size of plate, being
slenderer.

Madreporie body about midway between extreme margin and center of disk,
11 mm. in diameter, and hidden by about twenty-five large paxille.

Anal aperture of conspicuous size, though hidden by paxille; tube feet large,
pointed, without calcareous grains or rods in walls.

Type.—Cat. No. 24672, U.S.N.M.

Ty pe-locality.— Albatross station 3340, south of Alaska Peninsula, between
Unalaska and Kadiak (lat. 55° 26’ N.; long. 155° 26’ W.), 695 fathoms, mud; one
specimen.

Distribution.—Known only from the type-locality.

Remarks.—This species differs from both of the foregoing forms in the following
characters: Form of abactinal plates; width, covering, and special armature of
marginals; armature of adambulacrals. From D. eaimius it further differs in the
size and character of abactinal paxille; shape of rays; in distribution of papule
along mid-radial line; in having narrower superomarginals; form of outer end of
inferomarginals due to different armature; in having narrower actinal intermediate
areas on rays; and a different armature, as well as less prominent carinations to
actinal intermediate plates. From D. borealis the present species further differs
in having slightly narrower rays (probably not constant); larger paxille in center
of disk; in armature of actinal intermediate plates which have less prominent cari-
nations; in armature of mouth plates. D. letmophilus differs from nesiotes, of the
Hawaiian Islands, in having much broader rays; longer paxille pedicels in pro-
portion to width of crown of spinelets; more distantly spaced abactinal plates;
marginals which correspond plate for plate and do not alternate on outer part of
ray; wider superomarginals and inferomarginals, especially the latter; a less plane
actinal surface to inferomarginals; less thick specialized ridge; broader fasciolar
grooves; more prominently carinated actinal intermediate plates; larger actinal
intermediate areas on rays. The armature of adambulacrals and mouth plates is
very similar in the two forms, as well as the finer spinulation of marginals and
absence of special superomarginal spines or tubercles.

DIPSACASTER ANOPLUS Fisher.
Pl. 16, figs. 1, 2; pl. 52, figs. 4, 4a.
Dipsacaster anoplus FisHer, Zool. Anz., vol. 35, March 29, 1910, p. 547.
Diagnosis.—Rays five. R=25 mm.; r=12.5 mm.; R=2 r. Breadth of ray
at base, 13mm. General appearance greatly resembling a species of Leptychaster*
(as L. pacificus or L. anomalus). Rays tapering more or less gradually to a

bluntly pointed extremity; sometimes slightly swollen on outer part; interbra-
chial angles wide. Abactinal paxillar area fairly compact; paxille arranged in

2See Remarks on p. 100.
57444°—Bull. 76—11——7

98 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

oblique rows at sides of median radial area; pedicels high, pestle-shaped, with a
crown of thirty to forty spinelets, of which about one-half form a peripheral series;
abactinal plates feebly lobed, or round and crowded at center of disk; papule absent
from center of disk and in young specimens also along median radial area; papule
in fives and sixes about each plate. Marginal plates entirely without enlarged
spines or tubercles, and with deep and wide fasciolar grooves and narrow specialized
ridges; inferomarginals usually not extending beyond superomarginals. Adambu-
lacral plates about as long as broad, with a furrow series of five or six subcylin-
drical bluntly pointed spinelets, and on actinal surface twelve to eighteen spine-
lets decreasing in size outward; those adjacent to furrow series form an irregular
series of five or six, the rest without definite order as a rule. Actinal interradial
areas extensive; actinal intermediate plates extending two-thirds the length of ray,
strongly carinated and bearing a paxilliform group of rather long, slender spinelets.
Madreporic body hidden; not so large as in other species of genus. Gonads extend-
ing far along ray in sexually mature forms. Anal aperture present, opening out
from an undivided sac-like intestinal cecum.

Description.—Abactinal paxillar area fairly compact; paxille arranged in
oblique transverse rows at either side of median radial line, decreasing abruptly
in size from inner third of minor radius toward center of disk and gradually toward
extremity of ray. Each paxilla consists of a slightly lobed or roundish base,
immersed in the dorsal integument, from which arises a comparatively high round-
tipped pedicel, broader at extremity than at base, crowned with a divaricate group
of slender spinelets. A large paxilla has about fifteen to twenty spinelets in the
peripheral series and an equal number forming a group in the middle. The spine-
lets are about half as long as the pedicel. Small specimens have fewer spinelets.

Abactinal plates (or base of paxille) feebly lobed, or in youngest examples
round. In the type there are five or six very short fobes, but in small examples
the plates are indefinite in outline. In center of disk they are always round and
crowded. Papule absent from center of disk, and along median radial area are
either absent or interrupted in distribution in small examples; but in type the
papule extend all the way across ray; papule in fives and sixes about each plate.

Marginal plates without specialized spines of any sort, thus resembling Lep-
tychaster, and with wide, deep fasciolar channels between the raised ridges. Supero-
marginals, seventeen in number from interradial line, do not encroach much onto
abactinal surface, but are confined largely to side wall of ray. Each plate is wider
than long, and the raised ridges when denuded are less in thickness (or width, count-
ing the lesser dimension) than the intervening channels; sometimes a ridge is only
one-half as wide as the furrow. Each plate is covered with short, slender, and
blunt papilliform spinelets, becoming slenderer and longer in the fasciolar grooves.
Terminal plate, considerably broader than long, concave toward paxillar area, and
with two thimble-shaped spines on either side of the terminal papilla.

Inferomarginals extending but a trifle if at all beyond the superomarginals
in largest specimens; in small examples, usually somewhat, toward end of ray.
The plates encroach conspicuously onto actinal area, to which they form a strongly
arched margin; and viewed from the side they occupy considerably more than

half of side wall of ray. The plates are covered with spinelets similar to those of
superomarginals, only thicker.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 99

Adambulacral plates about as long as broad, not compressed, with a rounded
furrow margin and wide sutures between the plates. Furrow spinelets, five or
six, usually five, subeylindrical, bluntly pointed, the lateral slightly the shortest,
the median about as long as plate. On actinal surface of plate are twelve to eighteen
spinelets, those adjacent to furrow spinelets largest and forming an irregular series
of five or six. The others sometimes suggest two irregular series, but are usually
without order. The first plate is shorter than the rest.

Mouth plate very prominent actinally, the free margin occupying about one-
half the length; median suture, narrow elliptical, the outer end of plates diverging
slightly from interradial line. Marginal spinelets eight to ten, slender and pointed,
the innermost abruptly larger than remainder. A series of slender blunt spine-
lets follows the margin of median suture, and parallel with it an intermediate
series, while along the margin adjacent to first adambulacral are several very
slender, smaller spinelets, with two or three odd spinelets out of line on surface of
plate.

Actinal interradial areas extensive. The actinal intermediate plates are
strongly and narrowly carinated, and a single series extends two-thirds length of
ray measured from interradius, a second series about one-third, a third series to
the third inferomarginal, and between first inferomarginal and second adambulacral
are four intermediate plates in an interradial series. The intermediate plates
bear a paxilliform group of rather long slender spaced spinelets, the peripherally
placed ones being directed over the deep fasciolar channels between the raised
tabula of plates.

Madreporic body slightly nearer margin than center of disk. It is smaller
than in typical Dipsacaster and is entirely obscured by paxille, one or two of which
arise from its surface. The striations are fine and radiating.

Anatomical notes.—Superambulacral plates present, except on first ambulacral
plate, and on the distalmost. Gonads extending far along ray, the genital tubes
largest in interradial regions, thence decreasing in size distad. The arrangement
appears to be as in typical Dipsacaster except that the genital stolon is nearer
marginal plates. Immature specimens may be found in which the gonads are
still confined to the interradial region. Anal aperture present, and easily seen.
It connects directly with an unbranched sac-like intestinal eeecum, lying in the
left radius of the trivium. Its inner surface is rugose. The opening into the very
short intestine is slit-like, and is situated just below the anal aperture, on a low
prominence. <A single Polian vesicle in each interradius. Tube feet pointed,
without sucking disks, and without deposits in walls.

Variations—In some specimens the rays are broader and shorter than in
others, and in these the inferomarginals extend laterally beyond the superomar-
ginals; while there is also variation in the form of the tip of the ray, this being
more rounded in a few examples than in the type. These small differences cause
a considerable difference in the general facies of the species. Unfortunately most
of the specimens are undoubtedly young, on account of the condition of the gonads.
The variation in the distribution of the genital tubules is due to age, the gonads
appearing first in the interradial region and thence progressing along the ray.
Fully developed gonads are found in individuals with R= 13 mm., but some exam-
ples equally large have undeveloped gonads. In such an event one relies upon

100 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

the general facies, structure of marginals, characteristic terminal plate, and the
tall slender paxilla, as well as the ‘characteristic furrow armature to determine
the species.

Type.—Cat. No. 24675, U.S.N.M.

Ty pe-locality.—Albatross station 3347, off Cascade Head, Oregon, 345 fathoms,
mud (2).

Distribution.—Washington to vicinity of San Diego, California, and between
300 and 800 fathoms, usually on mud.

Specimens examined.—Forty-five specimens, including type, from the following
stations:

Specimens of Dipsacaster anoplus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
28302 ~ 222 Between Santa Catalina and San Clemente 414 Pray SANG S222 os a ececceen nasa | 26 U.S. Nat. Mus.
Islands, California.
OIG Sanna Off San Diego, California............------- 339 1 Do.
Ol cadadn lames WO sen ceern cece seer ctccesedecsssasannmene 313 2 Do.
2980...... South of Santa Cruz Island, California. ..... 603 2 Do.
9075..---- Off Sea Lion Rock, Washington.........-.-- 859 1 Do.
S347 2 ce <e Off Cascade Head, Oregon 345 2 Do.
A317 eo cee 3 Off San Diego, California .... ...| 161-510 | fine gray sand, green mud. d Albatross, 1904.
4351/5. .<~5|nen-= OO sececne nscee ant ea sees ce osea ence ae 423-488 | green mud..............-.-- 1 Do.
4380...... Near Los Coronados Islands, Lower Cali- | 530-638 | gray sand, green mud....... 2 Do.
fornia. |
4405...... Between San Clemente and Santa Catalina | 654-704 | green mud.............-..-- 1 Do.
| Islands, California.
MU en ela eau Opec nae een ae pera ocean 334-600 | rocks, shells, fine gray sand. - 1 Do.
4415...... Between Santa Catalina and Santa Barbara | 638-302 | green mud.........-....---- 1 Do.
| Islands, California.
4416...... Between Santa Barbara and San Nicholas | 448-323 | dark green mud........----. 2 Do.
Islands, California. |
4477. .s. Off Santa Cruz Island, California .........- 475-510 | preenmMuds =. -ccoosaee eet | 2 Do.

Remarks.—This species bears a striking resemblance to Leptychaster (for
instance a form about intermediate in external appearance between L. anomalus
and L. pacificus) principally on account of the unarmed marginals and the general
facies of the actinal surface. The structure of the gonads, character of paxille as
well as the distribution of papule and presence of a conspicuous anal aperture
remove the species from that genus. At the same time it must be admitted that
its relegation to Dipsacaster is open to criticism. The marginal plates, especially
the inferomarginals are unlike those of Dipsacaster, and the absence of papule from
the center of disk is an important difference. The madreporic plate is relatively
smaller than in typical Dipsacaster, and although hidden, only one or two paxille
arise from its surface, instead of many. The general form of the paxille, form of
intestinal cecum, well developed anal aperture, and distribution of gonads have
been the characters which have led me to place the species in Dipsacaster. The dis-
tribution of gonads would prevent the species being ranked under either Patagiaster
isher or Leptychaster Smith (including Parastropecten). Ludwig’s genus is founded
upon evidently immature specimens.

The present species undoubtedly brings Dipsacaster much closer to Leptychaster,
and it may be regarded as an extremely aberrant member of the former.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 101

Genus DYTASTER Sladen.

Dytaster SuavEN, Narr. Challenger Exp., vol. 1, 1885, p. 608; Challenger Asteroidea, 1889, p. 60.
Type,@ D. nobilis Sladen.

Crenaster > PERRIER (not d’Orbigny), Annales des sci. nat., 1885, p. 71; Expéd. sci. du Travailleur
et du Talisman, 1894, p. 306.

Diagnosis.—Disk rather small; rays very long and attenuate often subcarinate
along the median line abactinally, very sightly flexible. Marginal plates rectangular;
the superomarginals thin and lamelliform, confined more or less to the lateral wall
of the ray, which is usually vertical. Plates of upper series directly superposed on
the corresponding plate of the inferior series, and the contingent horizontal margins
of both plates are consequently straight. Plates of both series usually armed with
one more or less prominent thorn-like spine, but this may be aborted or may be
absent, especially in young forms. The general surface of the plates of both series
covered with short papilliform spinelets or granules. Abactinal area with small
rounded scale-like plates bearing small paxille more or less well developed; no
definite order of arrangement. Papule numerous or not, generally distributed (?)
or absent from outer half of ray. Actinal interradial areas fairly well developed,
with numerous intermediate plates arranged in more or less definite columns.
Armature of adambulacral plates consisting of a longitudinal furrow series of short
spines standing parallel to one another and one or more longitudinal series of
papiliform spinelets or granules on the actinal surface; in some forms the first
series are well-developed spinelets and nearly as large as those of furrow series; and
in some there is a single enlarged spinelet among the first series. Madreporic body
very large, compound, placed near the marginal plates and covered with numerous
paxille. Pedicellariz (subvalvular) may be present on the abactinal and actinal
interradial areas. Anus present, subcentral. Actinostome and ambulacral furrows
wide; tube feet pointed. Gonads (always?) extending, in numerous dichotomously
branched tufts, far along ray in a series parallel with marginal plates.

DYTASTER GILBERTI Fisher.
Pl. 17, figs. 1-3; pl. 53, figs. 3, 3a.
Dytaster gilberti Fisher, Bull. Bur. Fisheries for 1904, vol. 24, p. 300, June 10, 1905.

Diagnosis.—Related to D. exilis Sladen. Rays five. R=114 mm.; r=22
mm.; R=5.2 r. Breadth of ray at base 22 mm.; at tenth superomarginal, one-
fourth length of ray, 17.5 mm. Rays elongate, robust, tapering at first very
slightly, then more rapidly, to a pointed, attenuate, extremity; abactinal integu-
ment inflated on disk, plane on rays; lateral wall of ray vertical on proximal half,

4No type was mentioned by Sladen in his original description of the genus in the Narrative, but
he fortunately mentions that ‘‘at the same station (No. 325) was dredged the genus Dytaster.’’ The only
Dytaster taken at that station was D. nobilis, upon which he evidently based the generic diagnosis.
The genus should date from 1889.

> This name has no status in nomenclature, irrespective of the status of the group to which it has been
applied. Crenaster has been employed previously in zoology by d’Orbigny. (Prodrome de paléon-
tologie, vol. 1, 1850, p. 340.)

102 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

eradually arching inward and upward on outer half, so that marginal plates form a
steep bevel; interbrachial arcs rounded. Ambulacral furrows wide; tube feet
large and pointed. Paxille small, without regular order, with fifteen to twenty
short papilliform spinelets on larger ones; certain scattered paxillee with two,
three, or four jawed pedicellarie; papule absent from terminal 0.4 of ray. Supero-
marginal plates fifty, subquadrate, slightly tumid, bearing a single tapering spine
on upper edge, forming a longitudinal series along dorso-lateral margin of ray
except very end; a similar spine on middle of each inferomarginal, forming a
similar series on ventro-lateral border; marginals covered with fine nap-like spinu-
lation, increasing in size toward middle of plates, and lower edge of inferomarginals.
Adambulacrals long and narrow with ten short furrow spinelets, and about two
rows of much shorter ones on actinal surface; distal most plates with an enlarged
central spine. Actinal interradial areas small, the plates extending to sixth or
seventh inferomarginal and covered with spaced spinelets. A few pedicellarie
consisting of three or four clavate, curved, often scoop-shaped jaws, evidently
modified spinelets, to each interradial area. Madreporic body very large, hidden
by about forty large paxille. Superambulacral plates present; gonads extending
far along ray.

Description.—Abactinal paxillar area nearly flat on ray, slightly convex on
disk, and on outer third of area becoming rapidly very narrow, so that for a con-
siderable distance near the extremity only three or four longitudinal rows of small
paxille separate the superomarginal plates of either side. Paxille rather small,
without definite order, and rather crowded on disk but distinctly spaced on ray;
at base of ray, about four paxille opposite each superomarginal. Each paxilla
with a pedicel broader at tip than at base, the former being often elliptical when
viewed from above; paxilla crown consisting of a nearly flat topped, subcircular
or four-cornered group of short, cylindrical, papilliform, obtuse, equal-sized spinelets,
or the peripheral slightly slenderer. There are fifteen to twenty spinelets to a
paxilla at base of rays, even more on disk, and about twelve to fifteen at outer part
of ray. These stand vertically, the peripheral series flaring a trifle on disk and
basal portion of rays, but on outer part of ray the peripheral series radiate widely.
The paxille are a trifle smaller along midradial line than on either side, and are
largest on interradial areas of disk and adjacent portions of the rays. Scattered
rather widely over disk and rays, but occurring especially adjacent to superomargi-
nals, certain paxille bear a single two, three, or four jawed pedicellaria. The jaws
are very much stouter than the spinelets, but do not extend much above their
general level, and are curved toward each other, being scoop-shaped at the tips.
These pedicellarie are very inconspicuous. Sometimes they replace all the spinelets
of a pedicel, but usually there is a more or less interrupted peripheral series sur-
rounding them like the calyx of a flower.

Abactinal plates crowded, often slightly overlapping, subcircular, irregular,
or with slight indication of lobing, especially on disk. Papule small, rather irregular
in position, absent from distal 0.4 of ray, and on disk from numerous narrow areas

radiating from center. There are usually about four papule around the paxille;
occasionally five, often only three.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 103

Marginal plates thin. Superomarginals, fifty in number from median inter-
radial line to extremity of ray, nearly quadrate except in interbrachial arc, where
they are higher than long; plates confined to side wall, except that the upper end
(representing the thickness of plate) forms a very narrow border to the paxillar
area. At the base of the ray the plates are vertically oriented, but beyond the
basal 0.4 they slope upward and mesially (together with the inferomarginals), thus
forming a steep bevel. At the end of ray the inferomarginal plates with their
spines define the contour of ray. Each plate is slightly tumid and bears a rigid,
erect, robust, tapering spine, at about the middle of the abactinal end, and approxi-
mately as long as the height of plate. These spines decrease in size distad, and
are apparently lacking on the last few plates. General surface of plates covered
with very small papilliform spinelets, which are spaced and slenderer at edges of
plate than in center. No fasciolar channels. Terminal plate longer than wide,
notched toward paxillar area, and apparently covered with fine spinelets, although
these are rubbed off.

The inferomarginals correspond plate for plate with superomarginals (except
at extreme tip of ray, where there is one additional plate) and the slight tumidity
of the latter merges into that of the former without a break at intermarginal
suture, and the spinelets are also the same, but increase in size toward inner and
lower edge of plate, being more widely spaced on the actinal surface. The plates
encroach slightly onto actinal surface, forming an abruptly rounded border, about
as much of a plate belonging to lateral as actinal surface. In center of each stands
a tapering rigid spine, forming a series all along ventro-lateral margin, and extend-
ing outward at right angles. These are a trifle longer than corresponding supero-
marginal spine. On first two inferomarginals (by a slip of the pen, erroneously
stated superomarginal in original description) is an additional spine between the
regular spine and actinal margin of plate. The superomarginal and inferomarginal
series of spines are very regular.

Adambulacral plates much longer than wide, with a slightly curved furrow
margin. Armature consists of (1) a furrow series of ten (eight or nine sometimes)
fragile, slightly tapering, round-tipped, faintly compressed spinelets, which are
slightly bent at the base and graduated toward either end of the series. (2) On the
actinal surface just behind the furrow series is a longitudinal row of about nine
much shorter, thick, clavate spinelets, and on the outer edge of the plate is a series
of delicate, slender tapering, papilliform spinelets which follow the contour of the
plate and are smaller than the other series. On the outer part of ray the plates
become very narrow, and although the spinelets are mostly rubbed off, two or
three plates near the tip still retain an enlarged actinal spine. Only the distal
most plates seem to have this.

Mouth plates large, prominent actinally, and the united pair broadest at
middle, but only slightly narrower at outer end; lateral margins toward furrow
slightly excavated, and the end toward actinostome truncate. (Unfortunately the
armature in the specimens examined has been largely rubbed off.) Marginal series
much like that of adambulacral plates, and consisting of about fourteen spines,
compressed and curiously expanded above the slender base, from this expansion
tapering to the extremity. These spines seem to increase in size toward the inner

104 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

angle, where there are six larger and heavier teeth. General surface of plates
covered with small papilliform spinelets similar to those of actinal intermediate
plates, and increasing in thickness at inner angle.

Actinal intermediate areas small, the intermediate plates, which are slightly
convex, extending to sixth or seventh inferomarginal, those adjacent to the
adambulacral plates being largest. The plates are armed with spaced, radiating,
small, papilliform spinelets. In a smaller specimen than the type several plates of
each area bear a peculiar pedicellarian apparatus consisting of three or four shorter,
thickened, clavate spinelets. These jaws are curved toward each other, and the
tip is usually broadened and scoop-shaped. The type has only one or two to
each area. ;

Madreporic body large, 8 mm. in diameter, situated a little more than half its
own diameter distant from the margin, and hidden by about forty large ornate
paxille, which stand flush with the general surface. These paxille are larger than
any on the general abactinal surface and their spinelets are heavier and more
clavate, those situated on the periphery being slenderer, however, and apparently
sharp or even mucronate.

Anatomical notes—Superambulacral ossicles present, though very small;
absent from first ambulacral, and beyond middle of ray, or else rudimentary.
Gonads in very numerous small tufts, these distributed in a linear series parallel to
marginal plates, and extending beyond middle of ray (for about two-thirds length
in type). Upper and lower ends of ambulacral ossicles very broad, the lower
end with a median keel. Tube feet large conical pointed, without deposits. Anus
present.

Type—Cat. No. 22331, U.S.N.M.

Ty pe-locality— Albatross station 4397, off San Diego, California, 2,196 to
2,228 fathoms, gray mud; two specimens (all that have been examined).

Remarks.—Among the species dredged by the Challenger, D. gilberti appears
to be nearest D. erilis, but differs from that form in having shorter and broader
rays, the sides of which are high, and arched inward toward the dorsal surface
distally, in having less crowded paxille with a greater number of spinelets, more
numerous marginal plates, the ventral series of which do not encroach upon the
actinal area to any great extent; a rather less numerous series of furrow spinelets;
in having abactinal pedicellarie. There is no carination of the rays as in D. evilis
carinata. The inward arching of the sides of ray mentioned above is not so pro-
nounced in a smaller mutilated specimen from the same station. The shape of
the ray is, however, characteristic. The sides are straight or a trifle swollen above
the base and do not have the arcuate taper of grandis (=madreporifer Sladen,
1889, pl. 3). The paxillar area is broader at base and narrower at extremity than
in spinosus, and therefore much wider than in ezilis.

Allowing for variation D. demonstrans Ludwig * appears to be very close to
the present species. The marginal plates and proportions are practically the
same, but the paxillar area does not appear to narrow so abruptly distad, nor to

@ Mem. Mus. Comp. Zodl., vol. 32, 1905 (July 17), p. 41, pl. 5, figs. 23-25; pl. 18, figs. 97-99; pl. 19,
fig. 107; pl. 20, figs. 108-115.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 105

be so wide at base of ray, judging by figure. The paxille are slightly larger in
gilberti and the papule probably not so numerous, there being rarely six about
the plates, this being the rule in demonstrans. There are a few other minor differ-
ences. Whether demonstrans can be retained after a series of gilberti is dredged
remains to be seen. J think it probable that the former will be found to be iden-
tical with the latter, although at present the two forms appear doubtfully distinct.

Family LUIDIJD Verrill, 1899.

' Luidiine SuapEN, Challenger Asteroidea, 1889, pp. 175, 244.
Luidiide Verriw, Trans. Conn. Acad., 1899, p. 201.—Fisuer, U.S. Fish Comm. Bull., 1903,
pt. 3, 1906, p. 1031.
Astropectinide (part) Gray, 1840, and succeeding authors to Verrill.

Phanerozonia near the Astropectinide, but without anus, intestine, or intes-
tinal ceca; superomarginals aborted, represented by paxille; inferomarginals,
adambulacrals, and actinal intermediate plates forming regular transverse series,
the first two being correspondent in number; papule compound, the extremity
forming a tuft of papille.

Remarks.—Sladen places the Platasterias* of Gray in this group. There are
several morphological features of importance concerning which we have no infor-
mation. Gray’s description directly contradicts Sladen’s statement, but the
former, judging by figures, is evidently inaccurate and misleading. He says, for
instance, that there are no marginal plates. The correspondence in position of
actinal, adambulacral and inferomarginal plates is the only feature which Plata-
steries has in common (according to descriptions) with Lwidia. But the absence
of an intestine and ccecum, and the character of the papule are more important,
I think. Sladen states that Platasterias has superomarginals, so that the genus
may really belong in the Astropectinid near Ctenophoraster. Until definite infor-
mation is forthcoming it may remain provisionally a member of this family, although
I think that it will ultimately be placed in the Astropectinide. The presence of
superomarginals certainly would exclude the genus from the Luidiide. If, how-
ever, Platasterias lacks an intestine and ccecum, and has papule similar to those of
Luidia, then the feature of the superomarginal plate would be outweighed, and
should be dropped from the family diagnosis. At present, however, there is no
evidence upon which to base a conclusion.

Genus LUIDIA Forbes.

Luidia Forses, Mem. Wernerian Soc., vol. 8, 1839, p. 123. Type L. fragilissima Forbes=L.
ciliaris (Philippi).

Hemicnemis Miurer and Troscuet, Monatsber. Akad. Wiss. Berlin, April, 1840, p. 105.
Type, ZL. ciliaris (Philippi)).

Petalaster Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 183. Type, P. hardwickii Gray.

Diagnosis.—Rays five or more, long, usually narrow, and depressed; disk

small; abactinal surface covered with true paxille, frequently in regular longi-
tudinal rows at sides; superomarginal plates paxilliform, similar to adjacent

@ Proc. Zool. Soc., 1871, p. 136, pl. 9 (Tehuantepec, Mexico).

106 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

paxille but frequently larger; inferomarginals with a transverse series of spines
or spinelets; fasciolar channels deep and wide; actinal interradial area very small;
actinal intermediate plates extending nearly to tip of ray in a single series (excep-
tionally two or three series are present, and plates rarely absent), the inferomar-
cinals, intermediate plates and adambulacrals corresponding in a transverse series.
Papule compound, i. e., with distal portion subdivided into numerous papillee.
Pedicellariw frequently present, sessile, two or three jawed as a rule, either conical,
tong-shaped or low and bivalved. Tube-feet in two rows, without true sucking
disks. Superambulacral plates present, well developed. Anus and_ intestinal
ececum absent. Gonad in numerous tufts forming a linear series, all along either
side of abactinal integument.

KEY TO THE SPECIES OF LUIDIA HEREIN DESCRIBED.

a'. Lateral abactinal paxilla with a quadrate or subquadrate tabulum.
b'. No pedicellariz; abactinal surface drab gray or greenish gray in life.........-- foliolata, p. 106.
b?. Pedicellarise (bivalved) on inferomarginal plates (abactinal end) and on superomarginal paxille,
and trivalved upright pedicellarie on actinal intermediate plates; abactinal surface reddish in

life, sometimes mottled with lighter ...-..--.----------+----+-+++++22e+2e-222---- ludwigi, Pp. IS:
a2. Paxille with stellate crown; granuliform abactinal two-jawed pedicellariz; slender two-jawed
actinal intermediate pedicellarie; rather prominent lateral spines........-- asthenosoma, p. 116.

LUIDIA FOLIOLATA Grube.
Pl. 19, figs. 1-3; pl. 21, figs. 3-5; pl. 54, fig. 3.

Luidia foliolata Grune, 43. Jahresber. d. schles. Ges. vaterl. Cultur, Breslau, 1866, p. 69 (no
locality).—Lupwia, Mem. Mus. Comp. Zodl., vol. 32, 1905, p. 80.—FisHer, Proc. Wash.
Acad. Sci., vol. 8, 1906, p. 121.

Luidia foliata SuapEeN, Challenger Asteroidea, 1889, pp. 247, 742.

Diagnosis.—From a large specimen, station 4457. R=222 mm.; r=28 mm.;
R=8.64r. Breadth of ray at base, 32 mm. Rays five, long, evenly tapered to a
sharp but not attenuate extremity; abactinal surface of rays compact, well arched
but subplane along median area; interbrachial angles acute; rays constricted at base.
No pedicellarie. Abactinal paxille quadrate along either side of ray forming four
regular series; much smaller and irregular on median portion of ray. Abactinal
plates cruciform. Inferomarginals broad with a transverse row of three flat-
tened lanceolate pointed or blunt spines on outer half of plate; other spinelets
slender, not squamiform and chaffy. Adambulacral plates well spaced; armature
as follows: One saber-shaped furrow spine and on actinal surface three to five
slender, slightly tapering, bluntly pointed spines, accompanied by a variable num-
ber of much smaller spinelets. Actinal intermediate plates in a single longitudinal
series on ray, each bearing a tuft of spinelets but no pedicellarie. Mouth plates
with a compact group of numerous small spinelets on face toward actinostome,
and numerous spines on exposed surface. Color, olive drab or pinkish gray, usually
the former.

Description.—Abactinal paxille quadrate along either side of ray, forming four
regular series and decreasing in size toward median area of ray where the paxille
are small, irregular, and not arranged in series. Superomarginal paxille largest,

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 107

closely appressed to upper end of inferomarginal plates, bearing on tabulum twenty
to thirty polygonal, flat-topped granules decreasing in size toward periphery, which
is surrounded by about forty slender and, unless seen from side, inconspicuous
spinelets, the whole having the exposed surface very flat. The other paxille
resemble these, but differ in size as indicated above.

When the abactinal body wall is treated with caustic potash and viewed from
the inner side the plates or bases of paxille are seen to be very strongly cruciform;
that is, with four lobes. By these processes or arms, which are fairly stout, unta-
pered, truncate, or notched at tip, the plates overlap regularly along the four lateral
rows of paxille, irregularly over median area of disk, where the plates may have
three or five lobes each. On the lateral plates the transverse axis is raised into a
keel, which, as the plates are in regular transverse rows, forms a transverse ridge
about the width of four paxille, the lobes overlapping strongly. This keel is absent
from superomarginal plates, which have rather wider lobes than the adjacent series,
that lobe impinging on the inferomarginals being shorter and rounded. Infero-
marginals lobed also, but those on the longitudinal axis (of ray) very short. Plates
on central portion of ray and on center of disk very small in comparison to regular
lateral ones. Everywhere on the abactinal surface, in the relatively large spaces
between the plates, emerge the papulz, four about each plate. These papule are
compound; that is, the summit is broad and subdivided into many papillose subdi-
visions, forming a flat-topped group resembling a decalcified low paxilla.

Inferomarginal plates separated, as usual, by deep fasciolar grooves lined by
densely placed capillary spinelets on the outer but not on the inner part. Exposed
surface armed with a transverse row of three robust, short, flattened, lanceolate,
blunt, or pointed spines on the outer half of the plate, these forming three longitudinal
series all along edge of ray; median spine usually longest (4 mm.) equaling about
one-half width of actinal surface of its plate; occasionally innermost longest;
outermost spine of series decreasing more rapidly than median as series proceeds
along ray, so that at outer part of ray these are reduced to small spinelets; median
and inner spines frequently become broader in proportion to length on outer part of
ray, where they very gradually decrease in length. On first dozen plates there are
usually four or five spines in each transverse row. Remainder of actinal surface
beset with diverse sizes of flattened, blunt, or truncate, rarely pointed, spinelets;
two or three in median line larger than rest, more flattened, broader, and usually
very blunt or truncate, forming continuation of series of marginal spines. Spinelets
become more slender toward edge of exposed surface, grading very abruptly into
capillary spinelets lining grooves. These actinal inferomarginal spinelets are much
longer and larger than the flattened chaffy squamiform granules of L. brevispina
Litken. Abactinal end of plate with a number of flat-topped granuliform spinelets
intergrading on the one side with spinelets of actinal surface and on the other with
those of superomarginal paxille. Usually just external to outermost marginal spine
is a small flattened, sharp, lanceolate spinelet, the series of which dies out before the
middle of the ray is reached. Terminal plate strongly concave toward paxillar area,
crescent-shaped from above, and covered with granules enlarged distally.

108 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Adambulacral armature consisting of a saber-shaped furrow spine, and on
actinal surface three to five slender slightly tapering, bluntly pointed spines, accom-
panied by a variable number of much smaller spinelets. Of these the innermost
spine stands in line with furrow spine and is longer than latter. At base of ray the
remaining four or five usually form two transverse series, being accompanied by a
number of delicate spinelets on the edge of the plate, or farther along ray one, and
finally both of the outer series disappear, and either one of the first series, even,
becomes reduced in size. The actinal adambulacral spines decrease in length
toward outer end of plate. They are frequently more or less flattened. One or two
of the spinules on outer end of plate is usually larger than the rest.

Actinal intermediate plates each bear a tuft of spinelets, four or five in the center
being considerably longer than the remainder, and resembling the more slender of
the actinal inferomarginal spinelets. Some specimens have the actinal intermediate
spinelets shorter and clavate, and only one or two in the center slightly enlarged,
the tuft then appearing much more compact.

Mouth plates narrow but prominent. At the inner end of each plate is a diverg-
ing group of about twenty small papilliform spinelets on the edge between the
actinostomial and furrow surfaces of plate. These spinelets form a several ranked
series or triangular group between exposed surface of plate and peristome. On the
exposed surface are two series of numerous spines extending to outer end of plate
and thence decreasing in size; sometimes only one zigzag series; spines adjacent to
margin much larger. Margin (adjacent to first adambulacral) has many very small
spinelets, partly covering the deep fasciolar groove between the plates. The first
adambulacral has a small group of spinelets, similar to those on inner face of mouth
plates, between the furrow spine and first actinal. The next three plates usually
have one to three spinelets in this situation. The small spinelets along the trans-
verse margins of first seven or eight plates are more numerous than on succeeding
ones.

Madreporie body entirely hidden by paxille. It is situated in an interbrachial
angle, the outer margin being the width of two rows of paxille from inferomarginals.
Body convex and small.

Color in life a curious neutral olive drab or pinkish gray, yellowish white on
sides and actinal surface. Often some of the paxille are whitish, causing a mottled
appearance. The color changes very little in alcohol.

Anatomical notes.—Superambulacral plates present, well developed, absent from
first ambulacral ossicle. Unpaired oral plate or ‘‘odontophore” prominent, with
a well-developed keel, to which is attached lower end of the free edge of interradial
septum. Interradially the lower ends of ambulacral ossicles of two adjacent rays
are connected by a strong muscle band, the ends themselves being separated. This
allows considerable movement of mouth plates. No anus. Intestine and intestinal
caecum absent. Hepatic cceca extend two-thirds length of ray. Stomach occupies
most of disk, not divided into dorsal and ventral parts. Gonads large, extending
in numerous tufts depending from abactinal body wall, in linear series nearly to tip
of ray. The individual bundles of tubules are rather small in specimens examined,
consisting of four tubules twice dichotomously divided, making sixteen terminal

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 109

tubules. Each bundle is attached to a cross ridge formed by the carinate transverse
lobes of the cruciform plates. Tube feet very large, with pointed tips, and without
calcareous deposits in the walls.

Variations.—Small and medium-sized specimens differ from the above in hay-
ing relatively shorter rays and less conspicuous spines. For instance: R=106 mm.;

=16 mm.; R=6.6 r; again, R=45 mm.;r=9 mm.;R=5r., and soon. In the
first of these two specimens, the marginal spines are rather broader and shorter
than in very large examples, and in the latter specimen they are only about one-half
as long (relatively) and much broader in proportion. In the proximal part of ray
there are but three actinal adambulacral spines and in the distal portion only two.
Among the many examples of this species available, some are as small as R=8 mm.
These do not show any marginal spines, and indeed the latter do not appear in a rec-
ognizable form until R=20 mm. Here they are scarcely to be differentiated from
the other spinelets, being merely broader and flatter. I find no difficulty in sepa-
rating even these small specimens from brevispina on the character of the latter’s
chaffy inferomarginal spinelets or granules and the adambulacral armature.

Two unusually large specimens were taken at stations 4203 and 4219, in Queen
Charlotte Sound and Admiralty Inlet, British Columbia. In general features they
agree very well with Californian examples. R=270 mm.; r=30mm.; R=9r. In
that from 4203 the paxillar granules are not so flat topped nor so crowded as in the
large Californian examples and the paxille along central area of ray and on disk are
subcircular and slightly convex, the slender marginal spinelets being in evidence. In
Californian examples the central paxille are usually crowded, four or five sided, flat
topped, and very compact in every way. In these northern specimens the marginal
spines are a trifle slenderer and longer; that is, are more tapering and less obviously
lanceolate, the base being the broadest part of the spine. The two spinelets which
form on the inner half of the plate a continuation of the series of marginal spines
are flattened, chisel shaped, and much more conspicuous than in Californian speci-
mens, where the same spinelets are usually lanceolate. The other actinal spinelets
are longer and slenderer than in the Californian examples. The same is true of
spinelets of the intermediate plates. The adambulacral armature is essentially
alike in specimens from the two regions. These differences are much less marked
in smaller specimens, although the northern examples have slenderer and usually
slightly longer spines and spinelets.

Ty pe-locality.—Unknown.

Distribution.—Southeast Alaska (Kasaan Bay) to San Diego, California, and
probably to Mazatlan, Mexico; found in 10 to 189 fathoms, usually in less than 80
fathoms; so far as known not found above lowest tide marks.

\
110 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
Specimens ecamined.—Four hundred and fifteen, from the following localities:
Specimens of Luidia foliolata examined.
Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
Gulf of Georgia, B. C......-..---.----- 87 Ane Sands... .2s2.500 5-6 -e 3 4 U.S. Nat. Mus.
Near Flattery Rocks, Washington .... 37 fine|pray Sand®s- - --ceern-2--- =~ 1 Do.
..| Off Gray’s Harbor, Washington....... 58 POOKY arese'cpars <aierota minaa(eecreceentota 5 Do.
..| Off Columbia River, Oregon .......-.- * 68 gray) Sand ss. ccavescess-seeeee 5 Do.
.| Off Tillamook Rock, Oregon .......--.- 29 fine gray sand...............-- 2 Do.
20 ea laanas CO Se oeceac eee encsaenne te 1 Do.
53 fine gray sand, mud......-.... 3 Do.
96 sandiandimud) oss-5.cee.s ons 2 Do.
a 44 fine gray.sand.-.-.~.2..22<..<- 1 Do.
2051...... South of Santa Cruz Island, California. . CS | easss GO Sesicecer ce emer eens, 5 Do.
2056......| South of Santa Rosa Island, California. . 520 ew lessee G0 Soccasccciec sane cus tom miele al Do.
2070...... Santa Barbara Channel, California. ..... OF al scare GO sacseser sseceseeee een 1 Do.
2O7E ow lanes C0 santeacsceec one ceeeeeen-cesece<e 73 greenmudicstacics.stoscscccs 1 Do.
3046...... Off Gray’s Harbor, Washington......- 48 fine gray'sand--- =< 52.2...-s<<\- 16 Do.
co ree eet. pe ee le nr ete eae BOle teers Os aac et eee ae eee oa 15 Do.
3048...... [egy ea eh Oe Pee ee Ry Alixocky2c settee ueete ea eee B De.
3049...... Off Shoalwater Bay, Washington..... 43 fine black sand................ 2 Do.
SOSiscoo Off Heceta Bank, Oregon.............. 59 coarse broken shells. .....--..- 2 Do.
B0odssecealescas C0 eee cee eee eee ener near oe GAT eee Oe saan coer eset eee 2 Do.
SOBRE Sa enloscce CO ceccan ae ste eee em neneiacseees an ss 53 LOCKS Sect secue eseeee ee eae 3 Do.
9058....... Off Cape Foulweather, Oregon......... 38 coarse gray sand, shells........ 1 Do.
8059...... Off Siletz Bay, Oregon ................ 7 MUGS occa secserec<sseetonee 5 Do.
3060...... Off Tillamook Rock, Oregon .......... 28 Drown UMNd soc cee eee eee 1 Do.
BO Lasescefassoe DO waaceaesatasawadeaeeesceee Secc.+ 23 fine!black'sand2’s. -ss- ses ceee 2 Do.
8062...-..].-<.. GO w ee eamcce sees eaene saeco sa esnee Ee el Reo OAs aeet a= Jans cemaeeseece 1 Do.
W00G eee nc sees C0 poe eren eee nee eeneaee Sees e. 42 fine gray sand. ----- <-cee ose 7 Do.
S0GEiee ne elas NO re cone eens aenitesincc eset sas Cp | Bates GON cane nea Seneca 7 Do.
OB F aoa se Off Tsiltcoos River, Oregon............ 61 greenmud) sand. -s22:2- --- 2 4 Do.
3063. -...- Off Siuslaw River, Oregon............- 32 fine gray, sand’ S22 5252 .-ee- eee 3 Do.
3084. . | Off Heceta Head, Oregon .............. 46,0 eo Oe Siocon eon ecaseascaeaee 2 Do.
3006......| Off Orford Reef, Oregon.............-- 33 | stones, broken shells.......... 3 Do.
OL10.=e-08 Between Golden Gate and Monterey 39 MOCK c:cc ses occa swoeceeeeencene 3 Do.
Bay, California.
20 Say sand. cn snoesenee see ceee 4 Do.
62 DON Ss ems wea aaa 10 Do.
43 fine black sand................ ll Do.
--..-.| Off Santa Cruz, California.............. 48 greenInud, sands... ssc eec 2 Do. '

3147. Northwest of Santa Cruz, California... 56 browmamud + sae cane ee 14 Do.
3148. eT coma OWA atan capes acteteeee eee 8 Do.
3149. Off Pescadero Point, California......... 45 BrECMI MUG. wat eccec eee eee os 4 Do.
$152...... Drake’s Bay, California -. 36 fine! prayisand =~ -=1...2 seasons 4 Do.
bSscceatlusses DO cee tee eon ane anearne ee caer aee 32 BLCENIM UG! seinen cee teeaeeee 8 Do.
BIB 6s one ek WO ose nee Oence eee ee eee 20 black sand, mud.............. 2 Do.
S155; cca o aloes Oe emcinaaeceaacecdeteea ne secot oes 35 green Mud sencernc ase eee asec 3 Do.
BlO7E a. ceslocons Oeaa-saceoees SSie lessee Ova cateescnchesecaceees 2 Do.
Daihes caus Off Russian River, Cal 76 TOCKy; Sand. --ce seco cesses 2 Do.
3175...... Off Point Reyes, California .. 57 brown) se aces ne costes 4 Do.
DMO. cawetl ots C0 Joep cmeceennssate ees 37 SIA SON ons eres aweece eb aeees 7 Do.
3350...... Off Point Arena, California............ 75 fine sand, mud................ 3 Do.
3456...... Straits of Fuca, Washington . -| 136 gtay/Sand 225. sec sass 2 Do.
BAD so airs laren a OO aware dodeaweace stones ecsc eek Bove eee WO eve cevaceececce sees 1 Do.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER.

Specimens of Luidia foliolata examined—Continued.

111

Station. Locality. Depth.
Fathoms.
M671es Monterey Bay, California..........-..- | (66
Port Simpson, British Columbia...... eeeeeeees
Puget Sound, Washington. ............-|......--.-
Departure Bay, Nanaimo, British Co- |..........
lumbia.
4208.25. Queen Charlotte Sound (off Fort | 25-30
Rupert, Vancouver Island, British
Columbia).
4219... <<<. Admiralty Inlet (vicinity of Port | 16-26
Townsend, Washington).
4243. ....- | Kasaan Bay (Prince of Wales Island, | 42-47
southeast Alaska).
MOAB ee mare ieee! Uthenmera sine caste atin toe eee ear 123-101
55-58
75-134
2 164-85
4420...... Between Santa Barbara and San| 32-38
Nicholas Islands, California.
4431...... Off Santa Rosa Island, California... ... 30-41
4438...... Monterey Bay (north of Humpback | 41-46
Rock), California.
4441...... Monterey Bay (between Mussel Point | 35-28
and Humpback Rock).
4442...... Monterey Bay (5 miles northeast of | 26-31
Monterey).
4452...... Monterey Bay (northeast of Point | 49-50
Pinos).
4453......). 20.4 10 Seance ae se ann ane e se asta 49-51
4457...... oe eee ee eee a 46-40
4460......| Monterey Bay (11 miles off Moss Land-| 55-67
ing).
4464...... Monterey Bay (southwest of Moss! 51-36
Landing).
BAO Seen alae CO manenctcassseeenasdaa seme eeen ead 31-21
4467...... Monterey Bay (8 miles west-northwest | 54-51
of Moss Landing).
Monterey Bay (north of Point Pinos)..| 54-63
Monterey Bay (northwest of Point | 71-59
Pinos).
Monterey Bay (north of Point Pinos)..| 54-65
Monterey Bay (northeast of Point | 43-34
Pinos).
4476...... Monterey Bay (southwest of Moss | 25-39
Landing).
BT ieee ts DO sees cae eee anne eee: Secs 19-11
4480...... Monterey Bay (10 miles west-north-| 76-53
west of Moss Landing).
4483...... Monterey Bay (6 miles west of Moss | 45-44

Landing).

Nature of bottom.

Volcanic sand, gravel, broken
shells, sponge.

green mud, sand, broken shells.

Pee MIC seen nc ae

gray green mud, coarse sand,
shells.

fine gray sand, mud...........

mud, gray sand, rocks
fine gray sand

gray sand, mud
hard sand

| soft green mud

do

Rerbe|

14

eB

ne

nw

Collection.

U.S. Nat Mus.
U.S. Nat. Mus., T.H.
Streets.
U.S. Nat. Mus., Dr.
| C. B. Kennerly.
| U..S. Nat. Mus.,
| Geo. W. ‘Taylor.
| Albatross, 1903.

Albatross, 1904 (Cali-
fornia).
Do.
Do.
Do.

112 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Specimens of Luidia foliolata examined—Continued.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
4492......| Monterey Bay (off Hill’s Land- 26 soft green mud .....-..------- 1 Albatross,1904 (California).
ing), California.
ASO l eo casclennon WO snesn acceso ee sase- seo ens 12-11 hard! sand. .--2- ses e5 cess 1 Do.
4535......| Monterey Bay (3 miles off Pyra- 71-54 hard gray sand..........-.---- 4 Do.
| mid Point).
4548...... Monterey Bay (between Italian 46-54 coarse sand, shells, rocks.....-. 1 Do.
ledge and shore).
4551......| Monterey Bay (near Itallan}| 56-46 |..... COs osc eos scenes nee em 1 | Do.
| ledge).
4552......| Monterey Bay (70 Fathom Bank)| 73-66 green mud, rocks.......-.-.--- 1 Do.
60-80 |...-. CO rece enisiscc aa aeleie teers 10 Do.
66-69 j-.-.- GO ew atete eee ce aeilasacmreiterats 3 Do.
Monterey Bay (Portuguese | 53-54 | rocks.......-.-.----+++-+++---+ 9 Do.
Ledge).
4558...... Monterey Bay (Humpback 40-28 |..... Otte ee eee sence se nena ae 10 Do.
Rock). |
4560...... Monterey Bay (off Soquel 10-12 fine gray sand, rocks........-- 2 Do.
Point).
15 coarse sand, shells, rocks...... 1 Do.
10-11 hhard(sand so sees <a aseaeialai 12 Do.
Mazatlan) Mexico (2) <<< 5... sccc|ocoanscwas|cowesienc n= ssiicenee= esse emma 3 Stanford University.
San Pedro, California........... IB ecedactoens-seeesooee cee snieclees 6 University of California.

Remarks.—This is a constantly five rayed short spined species with quadrate
paxille along either side of the abactinal surface of rays. It is entirely without
pedicellariew, and is hence sharply differentiated from L. quinaria and L. limbata
of Japan, and appears to be more closely related to L. brevispina Liitken, of Mazatlan
and the Hawaiian Islands. From this form it may be readily distinguished by the
longer and more numerous inferomarginal spines, those of brevispina being very
short and confined to the extreme margin of ray. There are only one or two of them.
The other armature of inferomarginal plates consists, in brevispina, of numerous
broad rounded squamiform granules which imbricate strongly, and along border of
plates a row of very slender spinelets. The general appearance of the actinal surface
in the two species is entirely different. The paxille are usually convex in brevispina,
and the adambulacral armature and that of mouth plates is also very different.
There are no pedicellariz in brevispina. Compared with L. clathrata of the eastern
coast of the United States, foliolata is seen to have much heavier and less chaffy
spinelets on the inferomarginal plates, the latter being narrower and slightly longer
than in clathrata (and brevispina). The marginal spines of foliolata are more numer-
ous and heavier than those of clathrata. Both clathrata and brevispina lack the
numerous minute spinelets on inner face of mouth plates. From the specimens at
my disposal (one brevispina from Mazatlan and twelve clathrata from Cameron,
Louisiana) I would unhesitatingly place brevispina much nearer to clathrata than to
foliolata. The character of the abactinal paxille, marginal, adambulacral, and oral
armature is of the same general character in the first two forms, which appear to be
perfectly distinct, however.

As the list of localities will show this species is very numerous in moderate depths
onasoft bottom. In Monterey Bay it is about the commonest starfish. The species

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 113

is more active than any form with which I am acquainted. It progresses with ease
and comparative rapidity by half walking, half swimming. The large tube feet are
waved back and forth and the rays undergo a slight undulatory movement, asthey
are very flexible. When kept in stale water or placed in alcohol, the creature will
frequently constrict off the arms near the base. These do not fall off at once, but
soon break at the line of weakness where one may observe the paxille pulled apart
as if by some violent muscular contraction.

Remains of ophiurans and several Dentaliwm shells were found in about a dozen
stomachs examined, both of large and small specimens.

LUIDIA LUDWIGI Fisher.
Pl. 20, figs. 2, 3; pl. 21, fig. 2; pl. 54, figs. 2, 2a.
Luidia ludwigi Fisuer, Proc. Wash. Acad. Sci., vol. 8, Aug. 14, 1906, p. 122.

Diagnosis.—Rays five. R=107 mm.; r=13 mm.; R=8.2 r. Breadth of ray
at base, 15 mm. Rays slender, very gradually tapering to a pointed extremity;
interbrachial arcs acute; general form depressed as in other species of genus, but
abactinal surface well arched; sides of ray rounded; abactinal area with three or
four regular series of quadrate paxille on each side, the superomarginal with small
two and three jawed pedicellariw; inferomarginal plates rather narrow, arched,
with one to three, usually two, lateral spines, and three to six actinal spinules
larger than spinelets of general surface, and on upper end a pedicellaria similar to that
of adjacent paxilla; actinal intermediate plates of interradial areas and proximal
half of ray, each with a rather prominent three-jawed pedicellaria; adambulacral
plates with a curved furrow spine, three actinal spines, and one or two smaller
spinules.

Description —Abactinal paxillar area is rather crowded and paxille of four or
five regular lateral series are quadrate, although in the fourth, fifth, and sixth
series (according to size of specimen) many paxillee are subcircular or not obviously
quadrate. Superomarginal paxille slightly smaller than those of adjacent series;
paxille thence decreasing in size toward mid-radial area, where they are arranged
without regularity and are roundish or irregular in outline. In some small speci-
mens paxille are not so obviously quadrate in lateral series, being subcircular in
outline, but nevertheless arranged regularly. Crown of spinelets not so flat as in
foliolata, but rather convex, especially in small examples. Superomarginal paxille
with about thirty-five short clavate spinelets in a radiating coordinate group, and
most of them also with one or two small two or three jawed valvate pedicellaria,
slightly longer than spinelets; next series with about forty spinelets, those in center
of tabulum stouter than the peripheral, as in superomarginal paxille; small mid-
radial paxillee with about twenty spinelets, the peripheral series being much slenderer.
Paxille decrease in size toward center of disk, being smaller there than on basal
portion of ray.

Abactinal plates four-lobed as in foliolata, the lobes being a little shorter and
thicker. Superomarginal plates with much broader lobes than adjacent abactinal
series, being nearer a lozenge-shape with rounded corners. Plates imbricate by

57444°—Bull. 76—11——_8

114 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

lobes forming regular transverse and longitudinal series (except on area of irregular
paxillae which have a three to five lobed base). On the distal part of ray the plates
all have wider and shorter lobes, even those of the lateral series, which acquire an
arcuate quadrate contour and may rarely be arcuately pentagonal. | Papule four
about each plate, large on the lateral portions of ray, very small in center, and
irregular there. Papule are compound. Numbers of them are invaginated into
body cavity and gorged with granular substance as if actively excretory. This
is significant when considered in connection with the absence of an intestinal cecum
and anal pore. '

Inferomarginal plates relatively narrower than in foliolata. Fasciolar channels
deep, about as wide as the thickness of the intervening elevated ridges. Outer or
abactinal end of each plate with a two-jawed (rarely three-jawed) pedicellaria
similar to that of adjacent superomarginal paxilla, and with one or two, usually two,
tapering sharp spines, of which sometimes the inner, sometimes the outer, is the
longer; the longer (about 4 mm.) equal to about width of its plate; more rarely
three shorter subequal spines in transverse series on outer end of plate. The
spines form a prominent marginal fringe to ray. On actinal surface of plate three
to six much shorter spinules form a transverse series in line with lateral spines, or a
zigzag, or even double series, while margin of plate bears numerous unequal spinelets,
becoming more capillary in fasciolar grooves. Only the outer part of the furrows
is lined with the minute filtering spinelets. Considerable variation exists in the
details of arrangement and number of auxiliary spinules. They are usually blunt
and more or less flattened. From the actinal surface the marginal spines are
decidedly conspicuous.

Adambulacral armature consisting of a curved saber-shaped furrow spine,
and on actinal surface three tapering, bluntly pointed, sometimes quite slender,
spines, of which one, the longest, often strongly compressed and also saber-shaped,
stands behind furrow spine, and the other two form a slightly oblique longitudinal
series just behind first actinal spine; or two spines, the adoral the shorter, stand in
a longitudinal series just behind furrow spine, and the third just outside of the
aboral (longer) spine of the series; one to three small slender spinelets occur on
outer part of plate, frequently three at base of ray forming a longitudinal series,
or one on adoral edge of plate, back of outer adoral spine. At base of ray the
three inner actinal spines form a triangular series on margin of plate (the furrow
spine being higher in furrow), and behind them are usually two or three other
spines. ‘Two spines are sometimes associated to form a sort of incipient pedicellaria.

Actinal intermediate plates of interradial region and proximal half of ray paxilli-
form, surmounted by a prominent three-jawed pedicellaria which is surrounded
at base by numerous slender spinelets in a calyx-like whorl. Each pedicellaria is
conical and one and one-half to two times as high as its width at base. The inter-
mediate plates extend nearly to tip of ray, but on its outer third are very small
and usually without spinelets. They can be seen readily if ray is soaked in caustic

potash solution. Besides the primary row of intermediate plates, in each inter-
radius there are one to five additional plates.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 115

Mouth plates narrow, with six or seven marginal spines and seven or eight
superficial ones, forming together a double series on the raised exposed surface of
plate parallel with median suture. Inner spine of superficial series largest, and
like the rest, slender, pointed, tapering. All spines decrease in size toward outer
end of plate. Innermost marginal spine situated nearer peristome than is the
enlarged inner superficial spine.

Madreporie body between second and third lateral rows of paxilla and hidden
by them.

Color in alcohol, abactinal surface dull Indian red, actinal surface and lateral
spines, whitish; some small specimens with abactinal surface darker Indian red,
mottled with whitish, that is, with whitish paxille here and there. This fades into
brownish red, mottled with bleached yellow. The color washes off under a jet of
water.

Anatomical notes.—Gonads in a series along either side of abactinal wall of ray
as in preceeding species. Each tuft has about four short thick finger-like divisions.
No anus; no intestine; no intestinal cceca. Superambulacral plates well devel-
oped, absent from first ambulacral plate. Tube feet large, with bluntly pointed
tips; no calcareous deposits in their walls.

Variations.—The most conspicuous difference exhibited by small specimens
(R=27 mm.) are as follows: usually but one lateral spine (occasionally two) on
outer end of inferomarginals, auxiliary spinules less conspicuous, lateral paxille
sometimes rotund—less obviously quadrate; actimal intermediate pedicellarie
confined to interradial regions, or not extending far onto ray; abactinal pedicel-
larie slightly more conspicuous than in large specimens. Spines and spinelets are
slenderer in small and medium sized than in large examples.

Type.—Cat. No. 21928, U.S.N.M.

Type-locality— Albatross station 2970, vicinity of Santa Barbara Islands, in
29 fathoms, on fine gray sand and mud.

Distribution—Monterey Bay, south to San Pedro, California, in 15 to 50
fathoms.

Specimens examined.—Thirty-three from the following Californian stations:

Specimens of Lwidia tudwigi examined.

Station. Locality. Depth. Nature of bottom. Number. | Collection.
Fathoms.
30 |\COSYE@ INN oe ede nae eee oes 1 U.S.Nat.Mus.
31 |hxaitned eds ie he ee DL | 1 Do.
26 | gray sand, pebbles .............-.... 2 Do.
29 fine gray sand, mud.... 6 Do.
30-41 mud, gray sand, rocks . 8 Albatross, 1904.
42-40 gray sand, shells. ...... 1 Do.
49-50 green mud, fine sand.............-.- | 2 Do.
28-24 | sand, fine, mud, shells............... 1 University of Cali-
fornia.
TUG BBN coe etdeen Wau sdemnanead 2 Do.
rocky {tt spoteicn--- -<o- eoacuncnane 7 Do.
sandy, kelp 1 Do.
sand and cobbles.................... } 1 Do.

116 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Remarks.—This species has the general form of L. lorioli Meissner (Mazatlan),
but has longer arms, which are more attenuate distally. L. ludwigi lacks the con-
spicuous sharp spinules which are present in many of the lateral abactinal paxille
of lorioli, and the latter has no abactinal pedicellarie, such as are very character-
istic of the present species. Another character which separates ludwigt from
both lorioli and bellone Liitken is the presence of prominent pedicellariz on the
actinal intermediate plates of interradial region and proximal half of ray. Details
of adambulacral armature differ in all three forms. JZ. ludwigi differs from L.
quinaria in having much longer narrower rays, no scattered and abundant abactinal
pedicellari over the midradial region, and in having three-jawed, not two-jawed,
actinal pedicellarie. The abactinal pedicellarize of quinaria are low, and of the
bivalved form of some Goniasteride. The adambulacral plates also have two-
jawed pedicellarie in guinaria.

LUIDIA ASTHENOSOMA Fisher.
Pl. 20, fig. 1; pl. 21, fig. 1; pl. 54, fig. 1.
Luidia asthenosoma Fisurr, Proc. Wash. Acad. Sci., vol. 8, Aug. 14, 1906, p. 124.

Diagnosis.—Rays five. R=86 mm.; r=9 mm.; R=9.5 r. Breadth of ray
at base, 10 to 11 mm. Rays long, narrow, pointed, very gently tapering, with a
slightly convex abactinal surface usually sunken along midradial line. General
form much flattened; sides of rays rounded; inferomarginal plates narrow, not
encroaching much upon actinal area, but forming rather the margin of ray, each
with a transverse row of three sharp acicular spines; ambulacral furrow wide and
shallow; tube feet long, in two series; actinal and marginal spines rather long and
bristling, the adambulacral armature, consisting of a transverse row of three promi-
nent spinules, forming a series in line with that of inferomarginal plates; actinal
intermediate plates usually with a rather short, two-jawed, blunt, papilliform
pedicellaria. Paxille stellate, the superomarginal about twice as large as others.
Scattered two or rarely three jawed abactinal pedicellariz are present. Prominent
two-jawed pedicellaria on margin of each mouth plate.

Description.—Abactinal paxille with a stellate crown; those of supero-
marginal series larger than rest, and each corresponding to an inferomarginal plate,
to upper end of which it is closely juxtaposed. Crown of superomarginal paxilla
longitudinally oval (as in sarsi), the others subcircular. Adjacent to superomar-
ginal paxille are about two regular longitudinal series of lateral abactinal paxille,
about two of which correspond to one superomarginal paxilla. Paxille diminish in
size very rapidly toward median line of ray and become less regular in arrangement
as they approach it. Superomarginal paxilla has slightly convex tabulum armed
with about thirty slender denticulate spinelets, of which about ten are scattered on
surface of tabulum and the remainder about the periphery, the whole forming a
diverging group. The superomarginal and other lateral paxille sometimes have a
blunt two-jawed pedicellaria similar to but larger than those scattered over the
midradial area (see below). The adjacent paxille have about twelve peripheral
and three to five central spinelets, while those in midradial region have about ten
peripheral and three or four central, very much smaller, spinelets, the whole paxilla

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. ai

being notably smaller. Many of small paxille of midradial area also bear in center
of tabulum, surrounded usually by a few small peripheral spinelets, a small obovoid
two-jawed valvate pedicellaria, resembling a split granule. Viewed from above,
the pedicellaria is elliptical in shape when closed. Each jaw is hollowed on inner
face and occasionally is larger, springing from a very low paxilla and emerging
between the others. Rarely there are three jaws. Jaws of pedicellaria much
thicker and more robust than any paxilla spines.

Plates, or bases of paxille, are cruciform along either side of area, leaving large
quadrate spaces for the papule. Superomarginal plates very much larger than
adjacent abactinal plates, and cruciform in shape also, but with very much broader
lobes, the arms on the longitudinal axis being longer than those on the transverse.
Inferomarginal plates also four-lobed (as seen from the inside), there being con-
spicuous roundish spaces between the infero and superomarginal series, correspond-
ing to the papular spaces on dorsal surface. Plates decrease very rapidly in size
toward radial line, are irregular, frequently have five lobes, and on median radial
area are roundish and crowded, without papule between. Thus along the radial
area of ray there is a band of varying width free from papule.

Inferomarginal plates relatively very narrow, transversely arched, encroaching
but slightly upon actinal surface, forming rounded margin to ray; chord of width
equal to one and one-half times that of adambulacral and actinal intermediate plates
combined. Fasciolar grooves deep and wide, slightly wider (i. e., measured on long
axis of ray) than corresponding dimension of specialized elevated ridge of plate.
Each plate with a transverse series of three robust, tapering, sharp spines, of which
the outer is often slightly the longest, but frequently the middle one or the two are
subequal; inner (actinal) spine of series is sometimes much slenderer than other ,
two, and only one-half or two-thirds length of longest spine; latter attains a length
of 5.5 mm., or slightly over one-half width of abactinal paxillar area, or nearly
twice the width of plate (that is, chord of width). General surface of plate covered
with slender, almost capillary spinelets, which become finer and more densely placed
in fasciolar grooves, and upper end of plate sometimes bears a pedicellaria similar
to those of abactinal surface.

Adambulacral plates with a slender saber-shaped furrow spine and forming
a linear series with it on actinal surface, two slender, tapering, pointed spines, the
inner of which is the stouter and slightly the longer. Except in small specimens,
one to three very slender spinelets stand on adoral side of outermost spine, which
decreases in size toward extremity of ray more rapidly than the inner.

On most of the actinal intermediate plates of proximal two-thirds of ray is a
small two-jawed pedicellaria accompanied by two or three capillary spinelets; when
former is absent its place is taken by about three to five capillary spinelets; jaws of
pedicellaria blunt, oblong to obovate, 0.5 mm. high; three or four pedicellarie in
interradial region, but very few spinelets.

Mouth plates more like those of Astropecten than most species of Luidia.
Exposed surface of combined plates, ovoid, prominent; suture between plates fairly
wide. Armature consisting of a slightly tapering, bluntly pointed tooth, and back
of that on margin a large two-jawed pedicellaria nearly as long as tooth. Two

118 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

shorter spines may take the place of the pedicellaria. Tn line with the tooth a series
of about ten superficial spinelets follows edge of suture, decreasing In size toward
outer end of plate; and along curved margin adjacent to first adambulacral arefour
or five slender spinelets, the second from inner end of series often thelongest. This
series is separated from the superficial by a shallow groove.

Color in life, reddish brown (burnt sienna) on abactinal surface; marginal spines
lichter, often whitish; actinal surface whitish.

Anatomical notes.—Superambulacral plates present, well developed, absent from
first ambulacral ossicle. Gonads extend, im a series along either side of abactinal
intecument, to tip of ray; individual tufts short, with few thick tubules. No intes-
tine; no intestinal cecum; no anus. Stomach not divided into dorsal and ventral
compartments. Tube feet large, pointed, without calcareous deposits in their walls.

Variations and young.—A specimen with R=31 mm. differs from the foregoing
description in having two marginal spines to plate, the third being sometimes repre-
sented by a much smaller spinule. Outer adambulacral spinelet is reduced in size,
and the others are relatively longer than in the bigger specimens. Actinal inter-
mediate pedicellarie are confined to basal half of ray, although a few are scattered
on distal portion. Abactinal pedicellarie few or absent. A very small specimen
(R=16 mm.) has two series of marginal spines along ray, the outer spines usually
the longer; actinal pedicellarie do not extend far onto ray; paxille small and
spaced; no abactinal pedicellariz.

Specimens vary considerably in respect to the abactinal pedicellariz. In the
type they are numerous, but some large examples lack them on the superomarginal
and adjacent paxille.

Type.—Cat. No. 21929, U.S.N.M.

Type-locality.— Albatross station 3148, off central California in 47 fathoms on
brown mud.

Distribution.—From north of Monterey Bay, California, south to Los Coronados
Islands, Lower California, 11 to 339 fathoms.

Specimens examined.—Seventy-one, from the following stations (California,
except as noted):

Specimens of Luidia asthenosoma examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms. |
Pea aaws Off Point Conception..........- 233 THUG eee woca sa eeaee ase ee 1 U.S. Nat. Mus.
2002 552353 Santa Barbara Channel......... 53 fine gray sand, mud........... 1 Do.
wa leadaee Cortes Bank, off San Diego-..... 60 FOCKS SAU Cancers eee erenecees 1 Do.
2030...... Near Los Coronados Islands, | 60 MUGS oe cis nasa cee sensene 1 Do.
Lower California.
34 gray sand, shells.............. 1 Do.
36 Une pray;Favdc. soos occ cee ee 1 Do.
30 coarse mndian acu see cscees 2 Do.
52 Ainejgray/sandes oss -. see cease 1 Do.
S106 6 a2 < Between Golden Gate and 43 rocks, broken shells........... 2 Do.
Santa Cruz.
3148......| Northwest of Santa Cruz ....... 47 DIOWMINUG so. -ae cee ened 2 Do.
37a. Seo. Monterey Bay.................. 68 gray sand 1 Do.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 119
Specimens of Lwidia asthenosoma examined—Continued.

|

Station. Locality. Depth. Nature of bottom. Number. Collection.
|
| Ip i | fin’ on ri ne ae.
| Fathoms.
S310 s.-3~- Off’ San Diepos—--< oe scans. 5-- 71-75 green mud, fine sand.......... 1 Albatross, 1904.
| do 110-199 | green mud, broken shells al 1 Do.
do 62-138 | gray sand, broken shells....... | Z Do.
4342....... Near Los Coronados Islands, 53-66 | fine gray sand................ 1 Do.
Lower California.
aSage dul Goi eee eee ek vt, | 55-155 |..... dosssot meres te 3 Do.
4347...... Off San; Diego. - ---*.--- 5... | 55-58 | fine gray sand, mud........... 1 Do.
AGSG eens elena GO seosenase ere nee eee =o 89-107 | coarse sand, rocks... 3 1 Do.
4421.2... | Off San Nicholas Island......-. 291-229 | gray mud, rocks -| 3 Do.
4493...... Between Santa Barbara and | 339-216 | green sand, shells............. | it Do.
San Nicholas Islands. | |
30-41 fine gray sand, green mud...-.-. 5 Do.
green mud, fine sand.......... 1 Do.
dark green mud............--. 2 Do.
soft dark green mud.......... ll Do.
PTeeM MUG = 25-22. ss5e sens ace 1 Do.
green mud, gray sand. on 1 Do.
wery Dard 2-22 0---.5-5 E= 2 1 Do.
6 Do.
1 Do.
3 Do.
1 Do.
1 Do.
3 University of California.

Remarks.—This fragile creature bears a close resemblance to L. sarsi Diiben
and Koren, of northern Europe and the Mediterranean, and may be looked upon as
a north Pacific representative of that species. None of the specimens are as large
as L. sarsi is known to grow. The Californian species differs from sarsi in having
very small, abactinal, two-jawed (rarely three-jawed), granuliform pedicellariz
scattered along the medioradial area, with larger ones sometimes on the regular
lateral paxille, and on upper end of inferomarginal plates. A comparison of speci-
mens shows that the spinelets of the abactinal paxille are much longer and more
delicate in asthenosoma and the inferomarginal spines are longer. In sarsi the central
paxillar spinelet or spinelets, in contrast to those of asthenosoma are stouter than the
peripheral. The actinal pedicellarie of sarsi taper slightly toward the tip, but in
asthenosoma they either do not taper or the tip is broader than the base, the pedicel-
laria being obovate. Lidia elegans Perrier from off the east coast of the United
States, is closely related to sarsi, and differs from asthenosoma in the structure of
the abactinal paxille, which have the central spinelets slightly stouter than the
peripheral, and all the spinelets shorter and more robust (as L. sarsi). The abactinal
pedicellarie are fewer than in asthenosoma, while the actinal pedicellarie are longer
and slenderer than in either sarsi or asthenosoma (twice or three times as long as
wide). Moreover, the pedicellaria taper very decidedly, especially on the inter-
radial areas and proximal portions of ray.

The rays of L. asthenosoma drop off very readily.

120 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Family BENTHOPECTINID Verrill, 1899, emended.

Pararchasterine SLADEN, Challenger Asteroidea, 1889, pp. XxVili, 4.
Benthopectinide +Pontasterine VERRILL, Trans. Conn. Acad., vol. 10, 1899, p. 217.
Notomyota Lupwia (Cheirasteride + Benthopectinidx), Sitzungsber. k. preuss. Akad. Wiss.,

vol. 23, 1910, p. 435.

Phanerozonia with more or less obviously alternating, convex, spiniferous
marginal plates, an odd one above and below in the interradial angles of three
genera; with small disk and long slender rays, and with the papul confined to a
limited, often specialized, area at base of ray and adjacent portion of disk (or all
of disk and up to three-fifths of ray). Especially characterized by the presence
of a pair of muscle bands extending along the dorsal wall of the rays, from the
region of the third to eighth superomarginal plates to tip. Abactinal plates with-
out definite arrangement, either simple and flattened, with one or few spinelets, or
in the form of low paxille, both kinds sometimes occurring in the same species.
Adambulacral plates with a salient inner angle bearing a series of spinelets; one
to several conspicuous actinal (or subambulacral) spines. Actinal intermediate
plates few and interradial areas small. Pedicellarie when present pectinate, fas-
cicular, or rarely bivalvate.. Anus present. No superambulacral plates. Tube
feet with sucking disks. Intestinal cecum thick,saccular, four-parted.

KEY TO THE KNOWN GENERA OF BENTHOPECTINIDS.

a, No unpaired interradial marginal plates; dorsal muscle bands sometimes attached to proximal
ambulacral ossicles by a tendon.
b'. Paxillee with numerous spinelets or granules.
c’. Papularia not two-lobed, usually more or less distended.
d', Pedicellarize when present two-jawed and occurring on the adambulacral plates; the latter
withibwo spines on their siriace.s.cn- <6 seco as coe eee eee ects Pontaster Sladen.
@, Pedicellarie fasciculate, occurring on dorsal, marginal, and actinal intermediate plates.
Pectinaster Perrier, p. 122.
c. Papularia not swollen, more or less distinctly two-lobed.
d'. Pedicellariz usually present and pectinate; actinal intermediate plates always present on
disk; no tuft of rather long paxillar spinelets near first superomarginal.
e. Dorsal muscle bands not unusually large and not attached to enlarged crest of a proximal
ambulacral ossicle; adambulacral plates with one enlarged subambulacral spine.
Cheiraster Studer.
e. Dorsal muscle bands large and attached by a prominent tendon to a specialized knob or crest
of either one or two proximal ambulacral ossicles, as well as to adjacent superomarginal
plates; adambulacral plates with two or more enlarged subambulacral spines; marginals
very spiny...-.-..... Bae estates ei Se sc cen etn Neer Luidiaster Studer, p. 127.
@. With a tuft of rather long paxillar spinelets near the first superomarginal plate; primary plates
of disk each with one large spine; pedicellarize unknown; adambulacral plates with two
subambulacral spines; actinal intermediate plates absent; innermost marginal mouth spine

Motsprolonged!s54.5s0: 5 esa eee eben eee er eee Marcelaster Koehler.
b*. Paxillse with usually but one spinelet, that on the primary plates of disk becoming a large spine;
papularia flat, two-lobed; no pedicellariee ..............-------------eeee eee Gaussaster Ludwig.

a’. Odd interradial marginal plate present in some-or all of the interradii. Dorsal muscle bands not
attached to a proximal ambulacral ossicle.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 121

b!. Abactinal plates of papular areas low tabulate and strongly stellate, the larger bearing a central
conspicuous spine surrounded by a circle of more or less elongate spinelets; superomarginals
with two or three and inferomarginals with two to five long bristling spines; odd interradial
marginal plate not much if any more prominent than adjacent superomarginals.

c!, Papule extending one-third to three-fifths length of ray; dorsal muscle bands weak; mouth
plates with more than nine marginal spines each; odd interradial marginal usually absent in
HOME OLGHE INLerradiln seen ee - eek ee ee pe ee ed Nearchaster Fisher, p. 132.
c?, Papulee extending one-seventh length of ray; dorsal muscle bands strong, prominent; mouth

plates with less than eight marginal spines; odd interradial present in all interradii.
Myonotus Fisher, p. 140.

6?. Abactinal plates flat or somewhat convex and, though frequently lobed, not strongly so, bearing
a central spine only, or several spinelets in addition, or one to several small spinelets; superomar-
ginals usually with only one large spine and inferomarginals with one or two primary spines; odd
interradial marginal prominent and usually present in all interradii-. Benthopecten Verrill, p. 142.

Remarks.—Members of this family have a very characteristic outward “ habit’
and are easily identified with their group. This is due to the structure of the mar-
ginal and adambulacral plates, with their bristling armature, to the limited distribu-
tion of papule, and to a less extent to the simple abactinal plates which commonly
have aslender central spine and a variable number of additional spinelets. They are
often in the form of simple paxille, but all gradations between paxille and flat
immersed plates occur. Another feature which lends a highly characteristic appear-
ance to members of this family is the form of the pedicellariz. When these occur
they are of the double pectinate variety, or a modification of this type. That
is, a comb of modified spinelets is placed opposite another similar comb, each
occupying separate but closely juxtaposed plates, or the spinelets may be grouped
(fasciculated) on one plate or two neighboring plates. The double pectinate
pedicellariz are strictly characteristic of this family.

Previous to Sladen’s Challenger report (1889) the members of this family were
not set apart by even subfamily distinctions from other Archasteridx; in fact some
species of Cheiraster and the type of Acantharchaster were first described as “‘ Archas-
ter.” Sladen erected the subfamily Pararchasterine for the reception of Parar-
chaster and Pontaster. In 1894 Verrill split this subfamily, recognizing the Bentho-
pectininee for Benthopecten, and Pontasterine for Pontaster, Cheiraster, Pectinaster,
and Acantharchaster. In 1899 the Benthopectinins was raised to family rank by
Verrill, and the Pontasterins was placed in his new family Plutonasteride, near
Plutonaster and allied genera. Believing that Sladen’s grouping of genera more
nearly expressed the relationship, I placed the Pontasterine under the Benthopec-
tinide in 1906. For reasons stated below the Pontasterine has been abandoned in
this report.

The order Notomyota Ludwig (1910) is coextensive with my Benthopectinide
and is founded primarily on the characteristic dorsal muscle bands of the rays.
Ludwig recognizes two families, the Cheirasteride (Pontasterine Verrill) and
Benthopectinide, the latter including only Benthopecten and Pararchaster restricted,
hence having the limits originally set by Verrill. These two families are character-
ized by the absence or presence of unpaired interradial marginal plates, and by no
other features of importance. In the new genus Nearchaster, one or more, occasion-
ally all, the interradii have the odd marginal (not always in both series of the same

122 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

interradius, however). The abactinal plates are structurally like those of Lwidiaster
and Cheiraster rather than of Benthopecten, to which the presence of odd inter-
radials would naturally ally the genus. The dorsal muscle bands are very weak
in Nearchaster and well developed in Myonotus, and are nearer to Benthopecten
than to Luidiaster in the mode of attachment of the proximal ends. The very
spiny marginal plates suggest Tuidiaster rather than Benthopecten. Even Ben-
thopecten may lack some of the odd interradial marginal plates as regularly occurs
in B. mutabilis. In other words, Nearchaster stands nearly halfway between the
two groups, and furnishes convincing evidence of the close relationship of Bentho-
pecten on the one hand and Cheiraster and allies on the other. I have not main-
tained the two families so ably expounded by Ludwig, because I do not think a
sharp line of separation can be drawn.

Whether the Benthopectinide should be regarded as a distinct order, the
Notomyota, is wholly a matter of opinion. There are unfortunately no rules to
determine what shall constitute ordinal characters. The Porcellanasteride are
probably just as deserving of ordinal rank since they possess a set of special char-
acters which set them apart from the other Phanerozonia. The fact that the
Notomyota is not here adopted simply means that the writer has preferred to use
characters of wider application. It may well be that the solution of the present
unsatisfactory condition of the orders of Asteroidea will be the recognition of
numerous smaller groups; about four in the Phanerozonia and two in the Spinulosa.

I have followed in the main Ludwig’s arrangement of genera and species. I
am not at all sure of the validity of Marcelaster. I have not adopted Pararchaster,
since it is separated from Benthopecten by a rather trivial character, or at least by a
variable and difficult one.

For a systematic and historical account of the genera and species of the Noto-
myota, Ludwig’s paper should be consulted. Two new genera (with four new
species) have been added to his list, one new form to Pectinaster, and two to
Benthopecten.

Genus PECTINASTER Perrier.
Pectinaster PerrieER, Ann. sci. nat., vol. 19, 1885, art.8, p. 70. Type, P. filholi Perrier. Expéd.

5

sci. du Travailleur et du Talisman, 1894, p. 278.—Lupwia, Sitzungsber. k. preuss. Akad.
Wiss., vol. 23, 1910, p. 448.

Pontaster SuADEN (part). Challenger Asteroidea, 1889, p. 23.

Cheiraster Lupwie (part), Asteroidea, 1905, p. 1.—Fisuer, Zool. Anz., vol. 35, 1910, p. 551.

Diagnosis.—Benthopectinide lacking odd interradial marginals, with the
abactinal plates a very simple form of low paxille, bearing numerous spinelets
and with entire (not two-lobed) more or less swollen papularia; pedicellarize when
present fasciculate, confined to a single plate, or standing on two adjacent plates,
occurring on actinal intermediate plates, or also on the marginals and abactinal
paxillee; marginal plates small, usually elongately oval or subtriangular in form, the
two series having a tendency to alternate more or less, causing the horizontal
margin that touches the companion series of plates to be more or less definitely
angulated or to form two facets; actinal interradial areas with very few interme-
diate plates; these small and without order. Armature of adambulacral plates
(1) a semicircular furrow series of small spines, and (2) one, usually larger and

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 123

conical, on the actinal surface. Well-developed dorsal muscle bands, attached to
one or two proximal ambulacral ossicles and adjacent superomarginals.

Remarks.—The genera Cheiraster, Pontaster, and Pectinaster are very closely
related and have been combined in the past under the generic title of Cheiraster
(Ludwig, 1905; Fisher, 1906,1910). Ludwig (1910) has lately used the papularia,
pedicellariz, and subambulacral spines as primary characters for distinguishing
the groups. I find also a very useful feature in the mode of attachment of the
proximal end of the two dorsal muscle bands.

In Pontaster and Pectinaster the papularia are entire and swollen, while in
Cheiraster and Luidiaster they are flat and two-lobed. Pontaster is distinguished
from Pectinaster, in which the pedicellariw are fasciculate and on any plates but
the adambulacrals, by having two-jawed pedicellariz on the adambulacrals only.
LIuidiaster and Pectinaster differ from Cheiraster in having the dorsal muscle bands
attached to one or two of the proximal ambulacral ossicles (aear where they abut
against the superomarginals) as well as to the adjacent superomarginals; Cheiraster
has the muscles attached to the superomarginals and abactinal plates, but not to
the ambulacrals. Cheiraster has only one subambulacral spine, Luidiaster has
two or more.

The following species is a typical Pectinaster, with highly specialized papu-
larium and the maximum distribution of pedicellariz.

PECTINASTER AGASSIZI EVOPLUS (Fisher).
Pl. 28, figs. 1, 2; pl. 55, figs. 4, 4a; pl. 57, fig. 1.

Cheiraster agassizi evoplus Fisner, Zool. Anz., vol. 35, March 29, 1910, p. 55).

Diagnosis.—Ditlering from Pectinaster agassizi (Ludwig) in having abactinal
and intermarginal pedicellariz and usually larger papular areas with more numerous
papular pores. Rays five; R=72 mm.; r=11.5 mm.; R=6.4+r. Nearly related
to Pectinaster filholi (Perrier) and Pectinaster mimicus (Sladen). Rays long,
slender, evenly tapered from the base to an attenuate extremity. General form
very much as in P. mimicus, but rays shorter and slightly slenderer. Abactinal
surface with numerous small, roundish, low, spaced paxille; larger ones with a
delicate central spinule and a crown of eight to fifteen small spinelets, interspersed
with more numerous smaller ones lacking the spinule, the former extending abund-
antly to tip of ray; a variable number of small pectinate or, more rarely, fascicular
pedicellariw on the abactinal surface. Papulew confined to an oval papularium at
base of ray, forty to seventy-five pores to each area in adult specimens. Acces-
sory supermarginal spinules on first five or six plates, one to three (rarely four)
accessory inferomarginal spinules on proximal third or half of ray, only one beyond,
and merely the single primary spine on distal fourth of ray; adambulacral plates
with five to nine furrow spinelets, and one actinal with a small companion; actinal
intermediate plates extending as far as the sixth to tenth inferomarginal. Pedi-
cellariz numerous on interradial areas, and present on suture between marginal
plates, on inferomarginals, and also sometimes between inferomarginals and inter-
mediate plates, and between inferomarginals and adambulacrals.

Description.—Larger paxillze of abactinal surface slightly elevated, the summit
bearing an acicular, slender, very delicate spinule about 0.75 to 1.5 mm. long, sur-

124 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

rounded at its base by about eight to fifteen radiating, very small, papilliform
spinelets; the numerous smaller paxille, scattered between, with from one to eight
equal, upright or slightly radiating similar spinelets, one often standing in the
center of a circular group. Individual paxille clearly visible to naked eye; said
to be invisible in mimicus. The larger paxille with a central spinule extend entire
width of arm and to its tip. Scattered all over abactinal surface are numerous pedi-
cellaris formed of two to six or seven thickened spinelets, about intermediate in
lenzth between the central spinule and smaller spinelets. The spinelets of pedi-
cellariee are thicker than any others of abactinal surface, and form either a coor-
dinate (“fascicular”) group, or two opposing combs, usually two plates entering
into the formation of a pedicellaria. Pedicellariz are found on both large and
small plates, are largest on disk, and extend to end of ray in some specimens; in
others they are comparatively scarce, and lacking on outer part of ray; in a small
number of examples only four or six are present, on disk. The typical form from the
Galapagos to Gulf of California has no abactinal pedicellariz.

Papularium broadly elliptical, convex, so situated that a line drawn across
base of ray at interradial angle bisects it. It is about as long as half the width of
ray at this point. Papule increase in number with age; in large specimens, forty
to seventy-five to each area. The abactinal wall is much thicker on the papular
areas, since the plates are here in the form of vertical rods bearing a small crown
of spinelets externally (two to six or seven, with sometimes a central spinule of
inconspicuous size). The rods are compressed and oriented so as to appear to
radiate from larger central plates as shown in the figure (pl. 57, fig. 1). On
account of the form of the plates of papularium that area is convex also on the
inner or coelomic side of the body wall. Its plates are rather closely bound
together by tissue. Elsewhere the plates are thin, subcircular, and distinctly
spaced, the difference in size between the large and small plates being very
noticeable.

Superomarginal plates, thirty-four in an example with R=72 mm., confined to
side wall of ray, much longer than high on outer part of ray, where they are nearly
opposite the inferomarginals. Proximally, however, the plates are nearly or quite
as high as long and are alternate with the inferomarginals. Surface of plates,
sometimes with exception of a small area beneath the spine, is covered with spaced
short papilliform spinelets. Each plate bears near middle of upper edge a needle-
like sharp spine, the fourth from interradius longest (about 4 mm.). The first
four or five plates have below this an accessory lateral spine, which in the first three
or four plates is nearly or quite as long as the primary spine. A variable number
of the superomarginals, beginning with the fourth to tenth, have a small pedicellaria
composed of two to four stubby spinelets standing usually over the intermarginal
suture, some of the spinelets, therefore, on the inferomarginals. The number of
pedicellarie is variable; sometimes only a few distal plates have them, at others
fully two-thirds of the plates.

Besides a lateral prominent primary spine, slightly longer than those of superior
series, the first five or six plates bear two to four unequal accessory spinules on the
inner part of the plate; then beyond these a single spinule to outer third of ray.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 125

General surface of plates covered with delicate papilliform spinelets, similar to those
above, leaving partially or wholly naked a small area about the spines. Many of
the plates have on the lower margin a small pedicellaria, composed of three or four
spinelets. This is occasionally shared with the adambulacral or actinal intermediate
plates, according as one or the other of these is adjacent. These pedicellarie
frequently occur far along the ray, but are sometimes very few in number, though
not usually so.

Terminal plate tubercular, slightly wider than long, of a quadrate form, but
indented on the side toward paxillar area. Distally it bears a tuft of spinelets.

Adambulacral plates with a prominently curved furrow margin bearing a series
of six to eight (sometimes as few as five) slightly curved, blunt spinelets. The
maximum number of furrow spinelets is probably seven, the eighth and sometimes
a ninth continuing the series on the adoral margin of the plate. These two extra
spinelets are really members of the furrow series so far as appearance goes. In
center of plate is an upright, prominent, sharp, conical spine with usually an acces-
sory spinule, situated between it and the adoral margin. On outer edge of plate
one or two short pointed spinelets sometimes form part of a pedicellarian appara-
tus as mentioned above. At base of these are sometimes two actinal spinules,
instead of one and a small companion.

Mouth plates with furrow series of seven or eight spinelets, of which the two
or three inner are larger than the rest and the innermost largest of all. About mid-
way between inner and outer ends of plate is a prominent, tapering, actinal spinule.
A graduated uneven series of five to seven spinelets follows the median suture
border but is very irregular, and sometimes there is no order at all in the disposi-
tion of spinelets, so that the armature of mouth plates is of little value for classifi-
cation purposes.

Actinal interradial areas with numerous small crowded irregular plates, those
adjacent to inferomarginals being smallest and most crowded. In each interradial
area there are twenty-five to thirty rather large pedicellariw, each with from two
to five spiniform jaws. These extend some distance along ray on the single series
of intermediate plates, sometimes straddling the suture between an inferomarginal
and an intermediate plate. In interradial area the pedicellarie occupy either one
or two plates, but are conical in general form. Smaller intermediate plates
without pedicellariz usually have instead one or two slender spinelets.

Madreporic body situated about its own diameter from inner edge of supero-
marginal plates and surrounded by large paxille.

Anatomical notes—No superambulacral plates. Dorsal muscle bands promi-
nent and extending from about the sixth or seventh ambulacral ossicles (to which
the tendon is partly attached) to end of ray. Anus present; intestinal cecum
saccular, with four divisions as in other members of this family; no very distinct
dorsal stomach, merely five interradial constrictions forming a dorsal division;
hepatic coca short. Gonad confined to interradial areas. Tube feet with small
sucking disks.

Young.—A specimen with R=20 mm. (and r=4 mm.) has the general appear-
ance of the adult, the central spinule of the larger paxille being especially conspicu-

126 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

ous. Abactinal pedicellarie present on the disk; about ten papula to each area;
a few intermarginal pedicellariz; only two or three actinal interradial pedicellariz,
and none on inferomarginals or adambulacrals. Furrow spinelets, five to seven;
enlarged actinal mouth spine relatively larger than in adult; furrow spinelets of
mouth plates about five. Madreporic body nearly touches the marginals, only one
row of paxille intervening.

Type.—Cat. No. 27798, U.S.N.M.

Type-locality —Albatross station 4387, off San Diego, longitude of Point Con-
ception, 1,059 fathoms, green mud.

Distribution.—Known only from southern California, 984 to 1,059 fathoms;
undoubtedly ranging southward, where intergradation with P. agassizi is probable.
Typical agassizi ranges from the Galapagos Islands to the Gulf of California
(Ludwig, 1905).

Specimens ecamined.—Albatross station 2919, off Cortez Bank (latitude of San
Diego, California), 984 fathoms, gray mud, 218 specimens; station 4387, off San
Diego on longitude of Point Conception, 1,059 fathoms, green mud, 9 specimens;
total, 227 specimens.

Remarks.—This form is distinguished from mimicus, which was dredged by the
Challenger in the Arafura Sea, northwest of the Aru Islands, by the presence in
adult and moderately young specimens of abactinal, intermarginal, and much more
abundant actinal pedicellarie. The papularia are larger and contain many more
pores. For instance P. mimicus with R equalling 78 mm. has but a dozen pores to
each papularium, while evoplus has forty to seventy-five. In P. evoplus the furrow
spinelets are slenderer and curved, and the spine on the actinal surface of each
adambulacral plate is not so prominent as in mimicus, while the proximal infero-
marginals are more heavily armed than in mimicus.

Ludwig regards P. forcipatus (Sladen) as synonymous with P. filholi Perrier.
Judging from Sladen’s figures and description the papularia of filholi are small and
the abactinal pedicellariz have shorter spinelets and are differently formed (less
pectinate than usually in evoplus), the paxille have stouter spinelets, and the
central spinelet is short and conical. The paxille of evoplus are more nearly like
those of mimicus. Other differnces to be considered are: secondary superomar-
ginal spinule in evoplus, none in filholi; more numerous and prominent secondary
spinules on proximal inferomarginal plates in evoplus, and the furrow spinelets
longer, slenderer, and curved. Judging from Sladen’s figures the actinal pedicel-
lari of filholi have shorter and more pointed spinelets than in evoplus, and the
intermarginal pedicellariz of the latter are absent in filholi (although that species
proximally has a few pedicellariz on the vertical sutures between the superomar-
ginals).

Through the kindness of Professor Ludwig I have received for comparison a
specimen of P. agassizi from Albatross station 3366 (near Cocos Island). In this
example R=67mm. Although the form here described is close to agassizi, the differ-
ences already mentioned seem sufficient to warrant its separation as a distinct race.
The proximity of the known ranges of the two forms (as well as the range in depth)
points to their probable intergradation. In the specimen of agassizi examined
the papular areas are only one-half to two-thirds as large as in evoplus and

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 127

contain only about thirty pores, whereas in the latter race there are seldom as
few as forty in an equally large specimen. However, Ludwig records and figures
a papularium with fifty papule (R=65 mm.) in agassizi. There are more numerous
actinal pedicellarize in evoplus, and in agassizi, as already recorded, no abactinal or
intermarginal pedicellariz. In other characters the two species are practically
alike. Whatever slight differences in spine length and counts one might mention
is more than compensated for by the individual variation in evoplus.

It will not be surprising if mimicus, filholi, agassizi, and evoplus are nothing
more than races of one widely distributed species. They are all very closely

related.
Genus LUIDIASTER Studer.

Luwidiaster Sruper, Sitzungsber. naturforsch. Freunde, Berlin, Oct. 16, 1883, p. 131. Type,
L. hirsutus Studer.—Lupwia, Sitzungsber. k. preuss. Akad. Wiss., vol. 23, 1910, p. 451.
Acantharchaster VERRILL, Proc. U. S. Nat. Mus., vol. 17, 1894, p. 268. Type, Archaster dawsoni

Verrill.—FisHer, Zool. Anz., vol. 35, March 29, 1910, p. 549.
Cheiraster Perrier (part), Expéd. sci. du Travailleur et du Talisman, 1894, p. 275.—FisHer
(part), Bull. U. S. Fish Comm. for 1903, ph. 3, 1906, p. 1042.

Diagnosis.—Benthopectinide without odd interradial marginals and with the
abactinal plates a simple form of low paxillze, bearing a prominent central spine,
surrounded by accessory spinules; these interspersed with secondary smaller plates,
usually without central spine, bearing instead a few spinelets of varying length,
always short on outer part of ray; papule not sharply circumscribed, but occupying
entire basal part of rays, and adjacent portion of disk, and distal part of area more
or less two-lobed; pedicellariz pectinate, occupying two plates and occurring on
abactinal surface, sometimes also on actinal intermediate and marginal plates;
actinal intermediate !plates few, spinous; marginal plates more or less alternate,
very spiny; adambulacral plates with salient inner angle, and bearing a divergent
group of furrow spinules and a transverse actinal row of two or more long spines.
Dorsal muscle bands attached proximally to a crest of one or two ambulacral plates,
as well as to adjacent superomarginals.

Remarks.—In this genus the primary abactinal plates of the papular region are
roundish with faint indication of lobing. The plates are raised into a low tabulum
bearing a variable number of spinules and a central movable spine.

This group differs from Cheiraster, its nearest relative, in having the dorsal
muscle bands attached proximally to a special crest of an ambulacral ossicle, near
the point where the latter abuts against the superomarginal. The muscle is attached
also to the abactinal plates and superomarginals by the same tendon but the origin
is on an ambulacral plate. In Cheiraster inops and C. snydert the attachment to
the ambulacral is wanting. In Cheiraster the papularium is more definitely circum-
scribed. In L. dawson the papulze extend all the way across the ray and occupy
the radial portions of disk, except at the very center.

Ludwig includes five species: dawsoni, hirsutus, teres, vincenti, and gerlachei.
To these I would add horridus, described by me as a Cheiraster in 1906. The test
as to whether all these are really congeneric will be the determination in each form
of the mode of attachment of the dorsal muscle bands. I do not think gerlachei
is congeneric with dawsoni, and I am skeptical also concerning teres and vincenti.
I do not think oryacanthus is synonymous with dawsoni, as Ludwig states.

128 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
LUIDIASTER DAWSON I (Verrill).
Pl. 25, fig. 2; pl. 26, fig. 3; pl. 27, fig. 2; pl. 55, figs. 3, 3a; pl. 56, fig. 5; pl. 119, fig. 2; pl. 120.
Archaster dawsoni Vern, in Appendix C, On Some Marine Invertebrata from the Queen

Charlotte Islands, By J. F. Whiteaves, Report of Progress; Geol. Survey of Canada for

1878-79, 1880, p. 194b.
Acantharchaster dawsoni, Verrtty, Proc. U. S. Nat. Mus., vol. 17, 1894, p. 269.—FisHEr, Zool,

Anz., vol. 35, March 29, 1910, p. 549.

Luidiaster dawsoni Lupwia, Sitzungsber. k. preuss. Akad. Wiss., vol. 23, 1910, p. 452.

Diagnosis —Rays five. R=124 mm.; r=22 mm.; R=5.45 r. Breadth of
ray at base, 25 mm. A narrower-rayed example: R=107 mm.; r=21 mm.;
R=5.l r. Breadth of ray at base, 25 mm. In the first specimen the rays taper
evenly to the tip; in the second, abruptly at first, then very gradually. Inter-
brachial ares of first specimen, angular; of the second, distinctly rounded. Disk
small. Abactinal surface paved with small unequal plates in the form of low simple
paxille; the larger of these plates bearing a large central articulated spine sur-
rounded at base by a circle of small Spinal (usually); » the smaller paxillee without
central spine and with comparatively few divergent papilliform spinelets. Papule
confined to disk (except center and interradial lines) and basal part of rays. Double
or triple pectinate pedicellariz of conspicuous size on dorsal surface, and one to
three still larger in each interradius of actinal surface. Each pedicellaria with
upward to twelve curved papille or spines to each group. Anus present. Mar-
ginal plates conspicuous, alternate, those of the upper series smaller than the lower,
rounded, with a central eminence bearing one or two large bristling movable spines
surrounded by spinules. Inferomarginals with about three large spines. Adam-
bulacral plates with a salient furrow angle bearing a divergent group of furrow
spines (about five) and a transverse actinal row of two to four long slender blunt
spines. Actinal intermediate plates very few and confined to interradial region.

Description.—Abactinal area plane, or slightly swollen at base of rays. Abacti-
nal surface paved with larger round plates interspersed with smaller more irregular
ones. Viewed from the cclomic surface these plates are seen to be irregularly
arranged and close together, but not in actual contact. Papule are absent from
center of disk, a narrow interradial line, and distal two-thirds of R. They are pres-
ent, therefore, on basal part of ray and adjacent portion of disk. On this area there
is some semblance to a formation of longitudinal rows by the plates of the medio-
radial region, and here also there is a greater difference between the two sizes of
plates. The smaller plates act as connectives between the larger ones, and the large
papule pass through the interspaces. The plates form low, simple paxille (proto-
paxille). The tabulum is only very slightly elevated. The large paxil, corre-
sponding to the large plates, each bear a long, sharp, movable spine (sometimes two),
encircled by several shorter unequal spinules (two to six or eight). On the edge of
the tabulum is a circle of numerous small spinelets. Sometimes the auxiliary spi-
nules are very few or lacking and their place is taken by small spinelets. The smaller
paxille bear one to twelve small, slender (or thicker, papilliform), acute spinelets.
The dorsal spines are largest in center of disk, being there about 5 mm. long and
fairly stout. Beyond the papular area they decrease markedly in size, as may be

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 129

seen from the photographic figure. There is considerable variation in the density
and length of these spines, but they are never so long nor so robust as the supero-
marginals. The difference in size between the large and small paxille is better
marked on disk than on the nonpapulated area of ray. On the former area, in
large specimens, one may find numerous secondary plates, with a small central
spinule, in the process of forming primary plates. On the proximal half of ray are
from three to eight scattered pectinate pedicellarize of conspicuous size, similar to
the actinal pedicellarize described below. Each pedicellaria is subcircular in shape,
about 2.5 mm. wide, and consists of two or three primary plates juxtaposed, each
plate with five or six stout, curved, blunt, specialized spinules, meeting over the
common suture. (See figure.) Sometimes the pedicellariz lie between the upper
curved ends of two superomarginals.

Marginal plates of two series, alternate. Superomarginals not encroaching
upon abactinal area except far along ray. Each plate strongly convex, almost tabu-
late, bearing one or two stout, movable, tapering spines, the uppermost the longer
when two are present. These spines, which form a bristling series all along ray, are
as long as the combined length of two superomarginals, and the base of each (or of
the pair) is encircled by upward of twelve or fifteen slender auxiliary spinules. Out-
side of these on the sides of the eminence are numerous well-spaced spinelets, but
the border of the plate isnaked. In this way the plates of both series have the raised
portions separated by naked channels. The first superomarginal spine is shorter
than the succeeding ones.

Inferomarginal plates wider than superomarginals, and with central portion
similarly elevated, this elevation having a transverse trend. Each bears a trans-
verse series of three (occasionally four) rigid (though movably articulated), tapering,
pointed, outstanding spines, subequal, or the uppermost or central the longest.
Usually the spines are slightly stouter and often the longest is a trifle longer than the
corresponding superomarginal spine. Surrounding these is a variable number of
auxiliary spinules, about one-third the length of the spines, together with scattered
blunt spinelets. Terminal plate small, wider than long, densely covered with spines.

Adambulacral plates wider than long, with a fairly wide suture between. Fur-
row margin curved, with five to seven furrow spinules, the three or four central ones
as long as width of plate, slender, tapering, and blunt; the lateralmost spinelet very
short asarule. Actinal surface of plate with a transverse series of two or three long
terete, truncate or blunt, rather slender spines, which are nearly as long as infero-
marginal spines. On outer end of plate is often one or two much smaller spinules,
and along the adoral side of the series are a number of small papilliform or squamous
spinelets. In some specimens the skin covering the plates and spines is thicker than
in others. In the former the spinelets often appear to have swollen tips.

Mouth plates spade-shaped, with broadly rounded actinostomial margin and
rather convex surface. Marginal spines, six to eight, either tapering or untapered
and blunt, the inner one or two being considerably longer than the others, which
diminish rapidly in size. Seven to ten spines stand on the exposed surface of each
plate, the outer three or four being often much smaller. The larger spines vary from

57444°—Bull. 76—11——9

130 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

short, thick, and blunt to long and slender. They appear to be always shorter than
the corresponding adambulacral spines.

Actinal intermediate plates few and confined to the disk. Each bears a number
of small spaced spinelets and usually one or two upright slender tapering spines,
similar to but smaller than those of adambulacrals. In each interradius there are
one to three (usually one or two) large circular pectinate pedicellarize composed of
two opposite series of eight to twelve convergent strongly incurved papilliform
spinelets. Rarely there are three groups, in which case there are fewer spinelets to
each. The plates bearing the pedicellarian combs usually have also one or two
ordinary spines, and numerous papilliform or granuliform spinelets.

Madreporic body convex, with very irregular striz. It is situated nearer margin
than center of minor radius.

Color in life: some fresh alcoholic specimens from station 4791 showed unfaded
patches of bright vermilion, where alcohol had not had free access to the surface.
The tube feet seemed to be brightly colored also.

Anatomical notes—No superambulacral plates. Tube feet in two rows; sucking
disks well developed; ampullx double; one Polian vesicle in each interradius except
that of madreporic canal. Interbrachial septa membranous. Gonads confined to
interradial region and opening near marginal plates. Anus present. Intestinal
cecum large, with four thick lobes. Passage from ccecum into dorsal stomach short
and narrow. Stomach not clearly divided into dorsal and ventral portions. Retract-
ors of stomach well developed; cceca extending only a short distance into arm. No
calcareous particles in tube feet or walls of stomach.

The two muscle bands on the ccelomic surface of the dorsal integument are
remarkably well developed. The tendon is attached to aspecial crest of the seventh
and eighth ambulacral ossicles, and also to the inner surface of the adjacent (third)
superomarginal. The tendon is about 9 mm. long, and the heavy muscle band which
immediately becomes attached to the dorsal integument tends slightly inward,
broadens a little and at about one-third R, joins the muscle of the other side. The
two run parallel, and closely joined to tip of ray, occupying the width of the narrow
paxillar area. Professor Ludwig has kindly sent me preparations of L. hirsutus,
which show that the muscles are similarly attached in that species. This character
is useful, for some species of Luidiaster are superficially like Cheiraster.

Variations.—The chief variations have been considered in the foregoing descrip-
tion. These concern the relative length and number of dorsal spines and their
accompanying spinules, the superomarginal and adambulacral spines, and the pedi-
cellarie. Both actinal intermediate and abactinal pedicellarie are always present,
but variable in number. Some specimens have more numerous auxiliary spinules
surrounding the dorsal spines than others, and the accompanying miliary spinelets
are subject to considerable variation in length and number. The armature of the
marginal plates is essentially of the same character as that of the large paxille or
dorsal plates. Large specimens from station 3320 and station 3223 have broader
rays than examples from deeper water, as for instance, station 3339. This variation

is probably not correlated with depth, as examples from station 4792 have rather
narrow rays.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER,. 131

Young.—The smallest specimen has R=12.5 mm.; R=3 r. There are no
papulx, and one would refer it without hesitation to Cheiraster. Large abactinal
spines undeveloped, and madreporic body adjacent to superomarginals; unlike the
adult, the superomarginals encroach conspicuously on abactinal area; one supero-
marginal and two inferomarginal spines, the auxiliary spinules being nearly as long
as spines. Furrow spines, three to five; actinal adambulacral spines, two. Two
interradii have each a pedicellaria;* no abactinal pedicellariz (station 4784).

Another larger example, same locality, has R=25 mm., R=5 r. Some of
the dorsal spines have appeared on disk; only two or three papule; madreporic body
still near margin; marginal spines relatively longer; pedicellaria in each interradius.
A specimen from station 3223, with R=40 mm., has acquired abactinal pedicellariz ;
and the dorsal spines, though fewer than in adult, extend all along the ray. The
papulz are confined to base of ray, are few in number, but area extends all the way
across ray. Dorsal muscles well developed. An example with R=56 mm., R=5 r,
has the same appearance as the very large specimens, except that the papular area
is not so extensive nor the dorsal spines so numerous, and the madreporic body is
nearer to margin.

Type.—tIn the Yale University museum.

Type-locality— Off Queen Charlotte Islands, 111 fathoms.

Distribution.—Bering Sea (Pribilof, Aleutian, and Commander Islands), south
along the coast of Alaska to Queen Charlotte Islands, British Columbia, in 56 to
159 fathoms.

Specimens examined.—F orty-two, from the following stations:

Specimens of Luidiaster dawsoni examined.

Station. Locality. Depth. Character of bottom. Number. Collection.
| Fathoms.
2842... Near Unalaska, Alaska............-.- ee pebbles seascteer ese 5- eso te 1 U.S. Nat. Mus.
wie South of Kadiak, Alaska... ..| 159 | gray sand... Do.
--| 56 black pebbles. aes 11 Do.
| 85 | black sand.............- 1 Do.

59 eC Ocendied soe eon aoe 1 Do.
aos0soo= East of Shumagin Islands, Alaska . .. 138 mud gravel.........-.- 2 Do.
$500...... Near St. George Island, Bering Sea...) 121 fine gray sand, gravel 1 Do.

BZ BAL Wel 8 Near Attu Island, Aleutians........- 135 Gonrse pepbles! --. 2 cousin secs omnes 3 Albatross, 1906.
OTD orem Near Bering Island, Commander | 64 DEDDlGk eee saves sae Cums vecbyace 2 Do.
Islands. | |
A7OICe esl ke OSes neta ee eee 716 | rocky serene 5 Do
Miare, el ee DO aoe eee ee aotcee en enen aint 72 PODDIEN eee ste es 2. coeee ues 1 OM Do.
| |

Remarks.—Ludwig (1910) has placed my Cheiraster horridus from the Hawaiian
Islands under the present species. It is, however, distinct, and if compared with
an equal sized specimen of dawsoni presents the following differences. Compare
also pl. 17, fig. 3, Starfishes of the Hawaiian Islands, with pl. 119, fig. 2, this report.

a The young of Acantharchaster was stated by me (Bull. U.S. Fish Comm. for 1903, pt. 3, p. 1043) to
be without pedicellariz. The specimens I had then were the young of Nearchaster variabilis.

132

Horridus.
Abactinal paxille of disk frequently with acces-
sory spinules one-half as long as spine.
Madreporic plate situated its own diameter dis-

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Dawsoni.4
Abactinal paxille without elongated accessory
spinules.
Madreporic plate nearly touching superomarginals

(in young specimens only).

A large pectinate pedicellaria on most or all of the
actinal interradial areas.

Mouth plates with about four very prominent ac-
tinal spines (as long as the proximal subambu-
lacral spines) parallel to furrow series.

tant from superomarginals.
No pedicellariz.

Mouth plates with three slightly enlarged actinal
spinules.

Luidiaster oxyacanthus (Sladen) differs from dawsoni in the following particu-
lars: in dawsoni the paxille bearing special central spines extend to end of ray,
only onto base in oryacanthus; dawsoni frequently has two superomarginal spines,
rather than only one; furrow spinelets longer, slenderer, and the series with a
longer base line than in oryacanthus; the subambulacral spines longer and fre-
quently three (nearly as long as the inferomarginal spines) ; actinal intermediate
pedicellarie larger, each comb having eight to twelve spinelets (about five in
oryacanthus); mouth plates with more abruptly enlarged inner teeth and seven to
ten spines on the actinal surface (about five in oryacanthus). The disk of oxya-
canthus is smaller than in dawsoni and the interbrachial angles more abrupt, less
rounded.

Genus NEARCHASTER Fisher.

Acantharchaster FisHer (part), Zool. Anz., vol. 35, March 29, 1910, p. 549.
Nearchaster Fisher, Ann. and Mag. Nat. Hist., ser. 8, vol. 7. Jan., 1911, p. 91.
chaster aciculosus Fisher.

Type, Acanthar-

Benthopectinide with the odd interradial marginal plate lacking in from
one to four interradii, though sometimes present in all; not always present in both
series of the same interradius; abactinal plates large and small intermingled, the
large primary plates bearing on a low tabulum a long sharp spine surrounded by
a circle of accessory slender spinules, variable as to number and leneth—sometimes
over half as long as the primary spine; secondary plates with a group of shorter
spinelets; primary plates of papular areas strongly stellate; papule not in circum-
scribed areas, but distributed all over disk (except sometimes the very center)
and along ray from one-third to three-fifths its length, being confined to either
side of the paxillar area distally; rays very long and slender; marginals extremely
spiny—two or three long, sharp superomarginal spines and two to five inferomar-
ginals; subambulacral spines one to three, usually two, long; furrow spines one to
seven; mouth plates large, with numerous marginal spines, those of each pair of
plates disposed in three independent series: the median teeth which are the largest
and fewest, and on either side a series of smaller spines subequal or graduated in
length toward the mesial members of the group; pedicellarize when present large,
pectinate, on any or all of the following plates: abactinals, inferomarginals, actinal
intermediates; dorsal muscle bands not attached to an ambulacral ossicle, rather
weak.

# Taken from immature specimen; a large specimen would vary
horridus is an immature form, in all probability.

» Erroneously stated by me to be lacking in young dawsoni. The specimen examined then turns out
to be young Nearchaster variabilis. The attachment of the muscle bands is unknown in horridus.

slightly from this. The type of

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 133

Remarks.—This genus agrees with Benthopecten in having long rays and odd
interradial marginal plates. It differs in possessing usually more extensive papular
areas, in having much more spiny marginals, and abactinal plates, which instead
of being flat and often immersed, are tabulate and paxilliform, bearing numerous
spines of conspicuous length.

Superficially the most striking feature of Nearchasier is this dense armament
of abactinal spines, springing from low tabulate plates which in their essentials are
like those of Lwidiaster dawsoni. The species show more outward resemblance
to Luidiaster than to Benthopecten in spite of the presence of one or more odd
interradial marginal plates.

KEY TO THE SPECIES OF NEARCHASTER.

a, Abactinal accessory spinules long and slender, forming with the primary spines a dense armament;
secondary plates with long spinules; furrow spinules one to three; papule extending to middle
ObTAVs eas ene Wet Ne ian ae ST eos see oe SIs emaln sisiee oe ae cinte ra omnes ess aciculosus, p. 133.

a’. Abactinal accessory dpiiules not greatly developed in length or number; if developed at all only
two or three around each spine; secondary plates with short spinelets; abactinal integument
firmer and plates closer together; furrow spinules four to seven.

b'. Numerous abactinal pectinate pedicellariz together with inferomarginal and actinal intermediate
pedicellarie. Disk larger. Twenty adambulacral plates to ten inferomarginals; papule
extending as far as twenty-eighth to thirty-fifth superomarginal............- pedicellaris, p. 138.

b?. Very few abactinal pectinate pedicellarie and only exceptionally actinal or inferomarginal ones.
Disk smaller. Twenty-two adambulacral plates to ten inferomarginals; papule extending as
far as ninth to twenty-third superomarginal .....................-.----.---- variabilis, p. 137.

NEARCHASTER ACICULOSUS (Fisher).
Pl. 24, fig. 1; pl. 26, figs. 1-3; pl. 55, figs. 1, la, 1b; pl. 56, fig. 3; pl. 118, fig. 3.

Acantharchaster aciculosus FisHer, Zool. Anz., vol. 35, March 29, 1910, p. 550.

Nearchaster aciculosus Fisner, Ann. and Mag. Nat. Hist., ser. 8, vol. 7, Jan., 1911, p. 92, figs. 1, 3, 5.

Diagnosis.—Rays five (very rarely six), R=146 mm.; r=16 mm.; R=9r;
width of ray at base, 19 mm. Cotype: R=217 mm.; r=16 mm.; R=13.5 r;
width of ray at base, 19 mm. ‘This is the longest-rayed specimen, station 4381.
Rays very long, slender, flexible, and tapering gradually from narrow base to
attenuate extremity. Abactinal surface covered with stellate plates of two general
sizes, large primary plates bearing a raised tabulum with a central, long, slender,
movable needle-like spine surrounded by a circle of eight to fifteen divergent, very
slender, seta-like spimules one-third to over one-half the length of spine; inter-
spersed among these, smaller plates with spinules only, or perhaps an incipient cen-
tral spine; beyond middle of ray, plates with a uniform covering of minute spinelets,
three or four to a plate; a few abactinal pectinate pedicellariz. Papule numerous,
large, distributed all over disk and on ray to about two-thirds R from center.
Superomarginals with two or three long, slender, movable spines, surrounded by
seven to twelve auxiliary spinules; inferomarginals similarly armed; about three
or four (as few as one, or sometimes all five) interradii with an odd marginal plate,
bearing one or two spines, in both series, but not always corresponding; an odd
superomarginal sometimes opposite paired interradial inferomarginals. Adambula-
eral plates with two to four furrow spinules, and two or three long, slender actinal
ones in oblique series. Mouth plates broadly spade-shaped with nine to fifteen
marginal spinules, subequal except for the inner one or two, which are enlarged to

134 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
form teeth. Typically no actinal pedicellariz ; exceptionally these may be present
in northern examples.

Description.—Abactinal surface weak, paved with fairly large, strongly stellate,
rather loosely joined plates, interspersed irregularly with fewer smaller-lobed or
entire, irrecular secondary plates. Beyond about the basal third of ray the plates
lose the lobes, become rapidly smaller and roundish or irregular, and, where there
are no papuls, closer together. The large plates of the disk and basal part of ray
have five to nine unequal lobes, the secondary plates three to five or none. Papule
numerous and large, distributed all over disk (except a very narrow interradial
area near margin) and on ray to about two-thirds of R. Beyond about one-third
R the papule are confined to the sides of the abactinal area, the medioradial region
being free from them.

Each primary abactinal plate is raised into a low circular tabulum, bearing
on its center a very long, slender, sharp, movable spine, the longest being 8 to 11 mm.
in length (or about one-half to two-thirds width of ray at base). Each large spine
is surrounded, on edge of tabulum, by a circle of eight to fifteen straight, very slender,
unequal, divergent, seta-like spinules, about one-third to over one-half the length of
spine. The small plates bear a group of divergent spinules only. In typical speci-
mens the whole surface of disk and proximal third of ray is a mass of extraordinarily
slender spinules, the primary spines rising above the more delicate armament.
Beyond the region of abundant papule (basal third). on ray the primary spines
rapidly become fewer and smaller, and the secondary spinules also diminish in
number and size, while the secondary plates have a few short, very delicate spinelets.
Beyond, about the middle of ray, all plates are armed with three or four of these
short delicate spinelets, forming a uniform nap. A few pectinate pedicellarie,
with two opposing combs of four to ten slender sharp spinelets, are scattered here
and there over the papular area.

Marginal plates of two series alternate on proximal part of ray, but farther
along they are less evidently so. Plates of both series rather thin, longer than
wide, and with central spiniferous prominence. Superomarginals numerous (upward
to eighty-five), confined to side of ray on proximal part, but encroaching more
farther along. Typically each plate bears two or three long, slender, movable,
bristling, pointed spines in a vertical series, surrounded by a circle of seven to twelve
auxiliary spinules of very variable length and robustness. In addition there are a
few minute scattered thornlets. There are from two to five odd interradial supero-
marginals, specimens from the same station varying in this respect. One of the
first superomarginals of an interradius grows at the expense of its neighbor, which
is crowded out of place. Sometimes it is difficult to determine whether such a plate
is prominent enough to be considered “odd.” Usually about three or four inter-
radii have the odd plate (bearing usually one large spine, a smaller companion,
and several spinelets). The interradial plates, whether single or double, project
above the tops of the others. The superomarginal spines are longer and stouter
than the abactinal ones.

Inferomarginals larger than superomarginals, and like them very convex, the
prominence bearing a bristling group of two or three long spines (similar to but
slightly shorter than those of superomarginals) surrounded by about ten unequal

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 135

spinules, a few of which are often considerably longer than the rest, especially near
the base of ray. There may be no odd interradial inferomarginals, or from one to
four. In the cotype there are two, and one in the process of becoming ‘‘odd;”
in the type only one. An odd interradial superomarginal may have paired infero-
marginals, corresponding to it, and vice versa. The odd or interradial unpaired
marginal is a very variable character in this species, specimens from the same station
differing widely. Age seems to have something to do with its presence, and northern
specimens are more constant in having the maximum number. Typical examples
of this species lack marginal pedicellarie.

Adambulacral plates wider than long, with slightly angular furrow margin.
Armature a palmate furrow series of two, three, or four slender, unequal, spinules,
of variable length, and on the actinal surface of plate near the inner edge, an oblique
series of two, varying to three, long, slender spines. The figures will give a better
idea than description. The inner actinal spine (especially when three are present)
sometimes stands on the furrow margin and seems to form part of the furrow series,
the spine being then very disproportionate in size. In northern examples (Wash-
ington coast) only one furrow spinule is frequently present. A few minute thornlets
stand here and there about the large spines; otherwise the surface of the plates is bare.

Combined mouth plates broadly spade-shaped with a semicircular inner margin,
and a convex actinal surface. Each plate bears a furrow comb of nine to thirteen,
rarely fifteen, untapered, blunt spinelets of nearly equal length, except the inner-
most one or two, which are considerably enlarged. On the exposed surface of each
plate three to six spaced, long, slender, sometimes flat-tipped, spines form an angu-
lar series.

Actinal intermediate plates few, confined to interradial region, and armed
with one or two slender spines (like those of adambulacral plates), with a few
accompanying spinules, or with only a very few spinules, or with minute spinelets
simply. Northern specimens may occasionally have a small pectinate pedicellaria
in one or two of the interradii. This is typically absent.

Madreporic body convex, subcircular, situated less than half its own diameter
from margin. Ridges irregularly radiating, branched; strie interrupted.

Color in alcohol, whitish.

Anatomical notes.—No superambulacral plates. Gonads confined to inter-
radial regions and opening close to first superomarginal plate. Anus small; intes-
tinal cecum similar to that of Lwidiaster dawsoni, a sac with four radiating
obtuse lobes, opening into stomach, inconspicuous, though fair-sized; no distine-
tion between dorsal and ventral stomachs; hepatic caeca extending into rays about
two or three r. Double ampullxe; tube feet large, in two rows, with sucking disk;
one Polian vesicle in each interradius except that of madreporic canal. No calca-
reous particles in tube feet, ampulle, walls of stomach, or gonad. Dorsolateral
muscles weak, forming two inconspicuous bands on outer part of ray.

Variations.—The principal variations have already been noted in the foregoing
description. The rays increase considerably in length with age. Generally speak-
ing, the spinulation is more developed in southern than in northern specimens,
especially the secondary abactinal spinules. Northern specimens have, as a rule,
more of the interradii, with an odd upper and under marginal plate, the full com-
plement being common; southern examples have usually two to four, seldom five.

136 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Fully half the specimens from station 3347 have one or two pectinate pedicel-
lari in one or more of the actinal interradial areas and between several of the prox-
imal inferomarginals. The auxiliary spinules surrounding the abactinal primary
spines are fairly long, but are poorly developed on the secondary intermediate plates.

Y oung.—Very young specimens (station 4407) have short rays, a large termi-
nal plate, fewer spines than adult, and no papule. The smallest example has
R=6.5 mm., r=2+mm.; six superomarginals; two interradii, with an odd super-
omarginal; one superomarginal spine, no abactinal spines, a group of three to five
spinelets to each plate, one or two inferomarginal spines, one furrow spinelet, one
or two actinal adambulacral spinules, six marginal mouth spinelets, three on sur-
face of plate; no actinal intermediate plates. In general, the specimen resembles
a young Cheiraster with two odd marginals. An example as small as R=38 mm. is
sufficiently like the adult to make identification easy. The characteristic abactinal
plating and spinulation is very evident, and the actinal armature is essentially like
that of adult; but the rays are short and the papulz restricted to disk and proxi-
mal fourth of R.

Type.—Cat. No. 27799, U.S.N.M.

Type-locality.—Albatross station 4402, between San Diego and San Clemente
Island, California, 542 fathoms, green mud.

Distribution.—From south of the Alaska Peninsula to northern Lower Califor-
nia, in about 300 to 800 fathoms, usually in about 500 to 700 fathoms; principally
on green mud.

Specimens examined.—Ninety-two, from the following stations:

Specimens of Nearchaster aciculosus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.

Milacaens Off Gray’s Harbor, Washington .... 559 DLOWNiO0ZES. o -cese eee eee eee Z U.S. Nat. Mus.
2923. ....- Off San Clemente Island, California. - 822! \Vereenvmuds... ------ ----e eae eee=s 1 Do.
2980.....- Off Anacapa Island, California....... GOd laos se On Seno te ose ee eee 5 Do.
B00 cenaae Off Sea Lion Rock, Washington..... 686) Tile oae GOs poco a csetla ceeencceeeeenae 4 Do.
DU78 aeoe as |iweam C0 cesedvunessaseeee cee c een ee sees AUT | sea ee GOS. Sa ncee eae ce eene sceeeeeee 1 Do.
3338...... Vicinity Shumagin Islands, Alaska. . 625 greenimud; sand): .o2-s-ceseseseeae 3 Do.
Off Destruction Island, Washington. 516 preenimud:: <. 2.25 .550-0-c-ee os orene 24 Do.

-| Off Cascade Head, Oregon.........- 345 MU os cae a cee ee coos eRe eee 14 var. Do.
3670...... Off Monterey Bay, California ....... 581 greenimud,'sand 2 -.- -S-ccssc ce. ee 1 Do.
4335.....: Southwest of San Diego, California...| 500-524 | green mud, fine gray sand.......... 1 Albatross, 1904.
oY RES Vicinity of San Diego................ 549-585 | green mud 1 Do.

4353. .....)...26 WO Riese wenn soar nese ece eee 628-640 |..... GO RRG=5 7 Do.

4354.2... eee WO sn nce ieee see wa ce tatene oS eeees 642-650 |..... COL es enes ster eaen teas eee eee 3 Do.

Aa oS | Vicinity of Los Coronados Islands, | 618-692 |..... GO ss Seh so sana cee eee 1 Do.
| southwest of San Diego.

4398......| Between San Diego and San Clemente 620 greenimud,TOCKS:.=.----22.<.seecee 1 Do.
| Island, California.

4400.22... ae WO acc eect eet c eee: 500-507; |])'greem muds 2-2 sicece ences ose soeee 5 Do.

NE ia sd fons oe RD See deol ste aw ne eae te 542 12 Do.

4405...... Between San Clemente and Santa | 654-704 |_.... COE cats coon. cos oe eee cece eee 2 Do.
Catalina Islands, California

AO SS Seas eee sorte ae RO ee 334-600 | rocks, shells, fine gray sand........ 3 Do.

4421 ...... | Between Santa Barbara and San 241-298 | green mud, rocks................-.. 1 Do.
| Nicholas Island, California.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 137

NEARCHASTER VARIABILIS (Fisher).
Pl. 24, fig. 2; pl. 27, fig. 4; pl. 55, figs. 2, 2a; pl. 56, fig. 4; pl. 119, fig. 1.
Acantharchaster variabilis Fisher, Zool. Anz., vol. 35, March 29, 1910, p. 550.

Diagnosis.—Resembles N. aciculosus, but differs in having a firmer abactinal
integument, with closer-fitting plates; in the reduction of the spinules surrounding
abactinal primary spines, these being either very short, or, when lengthened,
comparativey few and and not forming a dense chevaux-de-frise; in having short
accessory marginal spinules; in having, as a rule, proximally four to six (instead of
one to three) furrow spinules; and in the variable extent of papular area on ray
(one-third to three-fifths length of ray). As this species is a close relative of
aciculosus, the description will be more intelligible if in the form of a comparison.

Description.—The spines are usually stouter than in typical aciculosus. The
accessory abactinal spinules, which form such a characteristic armament in acicu-
losus, are here reduced greatly in length, in extreme cases forming a ring of insig-
nificant spinelets about the bases of the primary spines and a small paxilliform
group on the secondary plates. Sometimes a few, or rarely nearly all, are length-
ened on part of the primary plates, but never on all, nor on the secondary plates
except very unequally and rarely. The papule are distributed all over disk except
for a narrow median interradial line, and extend along the ray a variable distance,
ranging from the ninth to the twenty-third superomarginal, or for about one-third
to three-fifths the total length measured along the side. Distally the papule are
confined to a rather narrow band on either side of the median radial area, and near
the tip of these bands the papule are often few and rather widely scattered. Abac-
tinal pectinate pedicellariza very few or absent. The armature of the marginal
plates follows the reduction seen on the abactinal plates. The accessory spinules
are small and weak and clustered about the base of the three primary spines. Some
of the latter on inferomarginal plates may have flattened tips. Odd interradial
superomarginals three to five, more often four or five than three. Interradial
inferomarginals usually fall one behind the superomarginals.

Adambulacral furrow spines proximally three to six, distally three to five,
slender, blunt, the two, three, or four central spines of the series the longest, and
proximally longer than width of plate. In the type there are five or six furrow
spines proximally, of which the three or four centralmost are the longest. © In
specimens from station 3330 there may be three, four, or five furrow spines, some-
what shorter than in type. Subambulacral spines, two, sometimes three, or very
rarely four on the first few plates, long, slender, blunt, in a diagonal transverse
series, the longer, blunter outer spine being on the adoral side of the plate. The
outer spine is about three times as long as width of plate, or proximally two and
one-third to two and one-half times as long as longest furrow spine; the inner slightly
shorter. An aberrant specimen from station 3330 has only one subambulacral on
most of the distal adambulacral plates. Along adoral border of plate are a few very
small spinelets. There are eleven to thirteen marginal spines on mouth plates, of
which the two inner are abruptly longer; five to seven long spines stand on the
actinal surface, forming a triangular group.

138 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Madreporic body very convex with irregular strie and situated, as a rule,
about its own diameter distant from margin. Measurements of type: R=175
mm.:r=15 mm.; R=11.6 r. Breadth of ray at base 19 mm.

Variations.—The principal variations have already been stated in the fore-
going description. These are so striking in certain particulars that it has been
difficult to decide whether the specimens belong to one or two species. In the
type the papule extend along the ray as far as the thirteenth to sixteenth super-
omarginal plates, while in two specimens from station 3330 they reach as far as the
twenty-third, in one other to the eighteenth and in another to the fourteenth, forming
a pretty regular transition. In asmall specimen from station 4238 the papule reach
to the eighth superomarginal, and in a full grown example (no locality) as far as the
ninth to twelfth plate. In the large examples, therefore, the variation is between
nine and twenty-three, the type being between the fwo extremes. The specimens
from station 3330 are otherwise aberrant in having usually only four, or three and
four (but in one case five), furrow spinules proximally, and one specimen has only
one subambulacral spine distally. This last example has two to six inferomarginal
pedicellaria proximally, and rather more than the normal number abactinally.
The two varieties intergrade so perfectly and the differences in spinulation alluded
to are so inconstant that I have regarded the specimens as belonging to a single
species.

Type.—Cat. No. 27800, U.S.N.M.

Type-locality— Albatross station 3331, north of Unalaska, 350 fathoms, mud.

Distribution.—Southern Bering Sea to southeastern Alaska, 108 to 351 fathoms;
in shallower water than N. aciculosus.

Specimens examined.—Ten; Albatross station 3330, same as type-locality, 351
fathoms, black sand, mud, four specimens, U.S.N.M.; station 3331, type-locality, 350
fathoms, one specimen, U.S.N.M.; station 3608, Bering Sea between Unalaska and
St. George Island, 276 fathoms, gray sand, U.S.N.M. one specimen; one specimen
without locality, probably Bering Sea, U.S.N.M.; station 4230, vicinity of Naha
Bay, Alaska, 240 to 108 fathoms, rocky, one specimen, Albatross, 1903; station 4238,
vicinity of Yes Bay, 229 to 231 fathoms, mud, rocky, two specimens, Albatross, 1903.

Remarks.—The specimens ranged under this species differ from aciculosus in
having much less conspicuous abactinal accessory or auxiliary spinules and a papu-
lar area of variable extent on the rays. The absence of long accessory abactinal
spinules gives variabilis a very different facies from the preceding form.

From pedicellaris, which I first believed to be a subspecies, variabilis differs
in lacking the numerous inferomarginal and abactinal pedicellarize, and in the
variable and less extensive papular areas. The disk of pedicellaris is larger and the
accessory spinules mote numerous and longer, giving the abactinal surface an
appearance only slightly less bristling than that of aciculosus.

NEARCHASTER PEDICELLARIS (Fisher).
Pl. 23, fig. 4; pl. 25, fig. 1; pl. 27, fig. 3; pl. 118, fig. 2.

Acantharchaster variabilis pedicellaris, FisHer, Zool. Anz., vol. 35, March 29, 1910, p. 550.

Diagnosis.—Resembling A. variabilis but differing in having a large number
of abactinal pectinate pedicellarix which are scattered all along ray; in the pres-
ence of numerous inferomarginal and usually one or two actinal interradial pedi-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 139

cellarize; in the constantly extensive papular areas; in the typically larger disk.
R=168 mm.; r=22 mm.; R=7.6 r.; breadth of ray at base, 26 mm.

Description —The dorsal pectinate pedicellariz are scattered thickly over
the abactinal area, extending far along ray, especially near the marginal plates.
The perfect pedicellariz of disk and proximal portion of ray consist of two combs,
each with about twelve tapering sharp, convergent spinelets. Smaller plates have
often only three spinelets to a comb, and sometimes there is one comb without an
opposing companion. The accessory or secondary spinules (five to eight) are very
unequally developed on primary plates. Only two or three may be developed to
any size, these occasionally attaining the length of the primary spine, while the
rest are successively shorter; or two, three, or four accessory spinules are developed
to one-half or more of the length of the primary, and no other spinelets are present
on plate; or, the spinules are short and form a ring around the base of the spine.
Many of the largest plates have a number of small spinelets in addition to the
accessory spinules. The secondary plates have a rudimentary central spine, with
a few small spinules, or else the Jatter only.

Papule very numerous, all over disk, and extending along ray as far as the
twenty-eighth to the thirty-fifth superomarginal, or beyond middle of ray. Beyond
about the fourteenth or fifteenth superomarginal they are confined to either side
of median radial area.

Three specimens have five odd interradial superomarginals, one has four, and
one has two. ‘The spines of the marginals are similar to those of variabilis. Some
of the proximal inferomarginals have four or five subequal slender, tapering, blunt
spines in addition to six or more shorter accessory spinules; over most of the ray
there are three primary spines, and distally two. The accompanying accessory
spinules are conspicuous, four or five, or sometimes more; occasionally two or
three are nearly as long as the primaries. In addition, small-spaced spinelets are
present, especially on actinal surface of plate. The first few superomarginals (as
well as inferomarginals) have a vertical single or double series of six or eight acces-
sory spinules, in addition to two or three stout, long, sharp primary spines (near
end of ray only one). Farther along ray the accessory spinules are slenderer and
shorter, seven or eight in number and grouped around the longer spines. In addi-
tion to these are several small scattered spinelets in a ring, outside the spinules.
On the vertical suture between many of the inferomarginals, but of irregular occur-
rence toward end of ray, is a conspicuous pectinate pedicellaria with two opposing
combs of about twelve slender spinelets. The actinal interradial areas have usually
one or two large pedicellariz and a few rather long spinules.

The adambulacral plates have a palmate furrow series of about six or seven
slender, untapered, round-tipped spinules, the two or three central ones longer
than width of plate and the others graduated on either side, the lateralmost
spinelet being very short. Not quite typical examples have five furrow spinules.
There are usually two (proximally three) slender, long, truncate spines in a
diagonal transverse series on the actinal surface of each adambulacral, the outer
being slightly the longer, and situated nearer the adoral margin of plate. These
spines are proximally about two and one-half times as long as width of plate.

140 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Mouth plates of the same shape as in variabilis, with eleven or twelve to sixteen
furrow spinules, the innermost two being abruptly enlarged to form teeth; the
others form an independent comb and are subequal or the mesial a trifle the longest.
Five to seven spines stand on the surface of the plate, four or five parallel with
the suture margin and the rest between them and the free margin, forming thus
a triangular group.

Type.—Cat. No. 27801, U.S.N.M.

Type-locality.— Albatross station 3337, south of Unimak Island, Alaska, 280
fathoms, green mud, rocks.

Distribution.—Known only from type-locality.

Specimens examined.—Four from the type-locality and two without labels.*

Remarks.—Typical examples of this species have the secondary or accessory
abactinal spinules about intermediate in point of development between aciculosus
and variabilis. The species differs from aciculosus in the presence of numerous
pedicellarie, more numerous furrow spines, and stouter spines generally, as well
as larger disk. The madreporic body is situated more than its own diameter from
the margin in pedicellaris; less than its diameter in aciculosus. The more extensive
and constant papular areas, the numerous abactinal pedicellarie, and slightly
more numerous furrow spines will serve to distinguish pedicellaris from variabilis.
In pedicellaris there are twenty adambulacral plates to the first ten superomarginals
(not counting the odd interradial); in variabilis there are twenty-two; in aciculosus,
nineteen or twenty.

Genus MYONOTUS Fisher.

Acantharchaster FisHEer (part), Zool. Anz., vol. 35, March 29, 1910, p. 551.

Myonotus Fisuer, Ann. and Mag. Nat. Hist., ser. 8, vol. 7, Jan., 1911, p. 91. Type, Acanthar-

chaster intermedius Fisher.

Diagnosis.—Benthopectinidex similar to Nearchaster but differing in having the
papule confined to the disk and proximal seventh of ray; in the comparatively
few marginal mouth spines, regularly graduated in length toward the inner teeth,
and in the very stout dorsal muscle bands; marginal plates very spiny (two or
three superomarginal and at least three inferomarginal spines); primary abactinal
plates of papular areas strongly lobed and in the form of tabulate paxille, bearing
a large central spine, and a circle of small spinelets on the tabulum; odd interradial
marginals present in all interradii, the unpaired superomarginal not conspicuously
larger than neighboring plates, and its spines not larger than succeeding supero-
marginal spines; adambulacral furrow spines one to three, long; subambulacral
spines, two.

Remarks.—This genus resembles Nearchaster in the following points: very
spiny marginals; general structure of the lobed, tabulate, spiny, primary abactinal
plates; the rather inconspicuous odd interradial marginals, these not being promi-
nent as in Benthopecten, nor having more conspicuous spines than the other supero-
marginals. It differs from Nearchaster in having the papule confined to the disk
and proximal seventh of ray, in having much stouter dorsal muscle bands, and

mouth plates with few marginal spines regularly graduated in length toward the
inner teeth.

@ Bottle broken by earthquake, 1906.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 141

Myonotus differs from Benthopecten in having prominent paxille bearing on
the tabulum a central movable spine and a circle of accessory spinelets, in not
having the odd interradial marginal plates and spines more prominent than the
other superomarginals, and in having much more numerous and conspicuous
marginal spines.

MYONOTUS INTERMEDIUS (Fisher).
Pl. 24, fig. 3; pl. 27, fig. 1; pl. 118, figs. la, 1-1/.

Acantharchaster intermedius FisHer, Zool. Anz., vol. 35, March 29, 1910, p. 551.
Myonotus intermedius Fisher, Ann. and Mag. Nat. Hist., ser. 8, vol. 7, Jan., 1911, p. 92, figs. 2, 4.

Diagnosis.—Resembling in general appearance a long-rayed Luidiaster with
odd interradial marginals. Characterized by the very spiny marginals (two or
three primary superomarginal spines, and proximally four or five, distally two or
three, inferomarginal spines, both sets with accessory spinules); primary abac-
tinal plates with a prominent denticulate spine, surrounded by a single series of
much smaller spinelets; papular areas slightly two-lobed on ray, and extending
to seventh superomarginal (one-seventh length of ray); furrow spines proximally
two, long; distally three, shorter; subambulacral spines two, unequal, much
longer than furrow spines; mouth plates with five or six marginal spines graduated
in length from long inner tooth to a short outer spinelet; on actinal surface of
mouth plates, five rather long spines in an angular series; dorsal muscle bands
stout. Rays five; R=158 mm.; r=17 mm.; R=8.3 r. Breadth of ray at base,
21 mm.

Description.—Abactinal plates stellate on papular area; lobes of primary
plates five to eight; secondary plates with two or three lobes or none; beyond
papular area plates are roundish or irregular, closely placed. Abactinal surface
has much the appearance of that of Luidiaster dawsoni, but the primary spines
are rather more numerous. These spines have strongly denticulate sides and each
is mounted on a low tabulum and surrounded by a single circle of ten to fifteen
short, sharp spinelets; secondary plates with one to four or five very inconspicuous
spinelets. The primary spines are much shorter than in Nearchaster aciculosus
(average 4 or 5 mm.). Beyond the middle of ray, spines are lacking and plates
are covered with a uniform nap of minute granuliform spinelets, three or four to a
plate. Abactinal pedicellariz are very few and small, with four or five spinelets.
Papule distributed all over disk and to seventh superomarginal, or about one-half
length of ray; area distally slightly two-lobed.

Marginal plates not, so evidently alternate as in Nearchaster aciculosus, but
they are irregular in this respect. On some parts of ray they appear to be nearly
opposite, while near by they alternate. Armature similar to that of Nearchaster
aciculosus. Superomarginal spines two or three, surrounded by five to eight very
slender and shorter accessory spinules. Numerous well-spaced, small, sharp spine-
lets are scattered over plate, mostly near the larger spines. Inferomarginals with
proximally four or five spines; farther along ray, two or three. Accessory spinules
prominent. There are five odd interradial marginals in both series, scarcely if any
larger than adjacent plates although placed on a slightly higher level. Each bears
two unequal or subequal spines, not quite so long as the adjacent superomarginal

142 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

spines. A very few pectinate pedicellariz, with ten or twelve short, blunt, slender
papilla, occur irregularly on three rays between the first and second or second
and third inferomarginals, and three of the actinal interradial areas have one or
two similar pedicellarie. The actinal intermediate plates are very few (five to
seven, and fifth adambulacral abuts against second inferomarginal), and are naked
except for one to three prominent spines or spinules on each plate.

Adambulacral plates, with proximally one, two, and, distally, three unequal
tapering furrow spinules, and on the actinal surface, two much longer, slender
blunt spines about two and one-third or two and one-half times width of plate.
That nearest margin is slenderer and slightly shorter and often stands at the aboral
inner corner of plate, fairly on margin.

Mouth plates with a less extensive furrow margin than species of Nearchaster;
furrow spinules five or six graduated from a long inner tooth to a very short outer
spinelet. About five tapering sharp spines stand on the surface in an angular series.

Madreporie body fairly large convex situated its own diameter from margin
and surrounded by four prominent spines. Striz fine, radiating from center.

Color in life, above pinkish buff, grayish at base of ray on papular area; actinal
surface, Naples yellow; tube feet, pinkish.

Dorsal muscle bands much stouter and broader than in Nearchaster, and more
as in L. dawsoni, but not so thick as in that form. After the proximal ninth of
ray the muscle becomes evident and runs parallel with the margin, being about 1.5
mm. wide. On the outer third of ray it joins its fellow and the entire width of the
narrow dorsal integument is covered with muscle. In N. aciculosus the bands are
thin enough to be translucent. This muscle is not attached to a proximal ambu-
lacral ossicle, as in L. dawsoni.

Type.—Cat. No. 27802, U.S.N.M.

Type-locality.— Albatross station 4530, Monterey Bay, California, 958 to 755
fathoms, very soft mud, one specimen, Albatross, 1904.

Distribution.—Known only from type-locality.

Genus BENTHOPECTEN Verrill.

Benthopecten Verrtu1, Amer. Journ, Sci. and Arts, vol. 28, September, 1884, p. 218 (footnote).
Type, B. spinosus Verrill.

Pararchaster SuADEN, Narr. Challenger Exp., vol. 1, 1885, p. 610, fig. 204. Type, P. pedicifer
Sladen; figure.

Diagnosis.—Benthopectinide with odd interradial marginal plates and rather
flat immersed abactinal plates without definite arrangement and bearing single
spines or groups; usually enlarged spines on disk; no paxille; papule confined to
disk and base of ray, exceptionally extending to middle of ray. Rays very long,
tapering, and flexible; disk small. Marginal plates suboval or triangular, elongate in
the direction of ray, and confined strictly to margin, the two series not exactly oppo-
site; each plate with a prominent boss. Superomarginal plates with usually one
long cylindrical tapering spine, and sometimes a companion; inferomarginal plates
with one or more similar spines (in one species clavate); general surface of plates of
both series naked, or only with minute isolated spiniferous thornlets. Odd supero-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 143

marginal usually with a very prominent spine. Actinal interradial areas with very
few intermediate (ventral) plates. Armature of adambulacral plates consisting of a
semicircular furrow series of small uniform spines, radiating fan-like, and one or
more larger actinal spines. The furrow series are typically separated from each
other by a considerable interval. Pectinate pedicellaria may be present on any or all
of the following regions: Abactinal plates, inferomarginal plates, actinal inter-
radial areas. Superambulacral plates always absent; tube feet with sucking disks.
Intestinal ececum saccular, four-parted.

Remarks.—Ludwig in his recent paper on the Notomyota recognizes Parar-
chaster (type, P. pedicifer) for those species which have several spinelets on each
abactinal plate, reserving for Benthopecten the forms which have one spine or spinelet
toa plate. His genus Pararchaster corresponds to the second section [B] of Sladen’s
key (Challenger Asteroidea, p. 5), less armatus.

Such a division of Benthopecten into two genera, or even subgenera, would be
useful, but in practice is difficult, in the absence of other correlative characters. At
the start we find that Benthopecten in the restricted sense has three species, incertus,
cognatus, and simplex with certain plates bearing two or three spinelets.

Among the following species claviger would undoubtedly fall in Pararchaster.
Mutabilis has generally single spinelets on the rays, but the large spines of the disk
are frequently surrounded by a circle of upward to ten or eleven short spinelets,
while other plates have one to three or four short stubby spinelets, in no wise differ-
ent from those present on typical ‘‘ Pararchaster.” But mutabilis, while presenting
several points of difference, greatly resembles Benthopecten spinosus, the type of the
genus. One would not consider placing the two species in separate genera. Ben-
thopecten acanthonotus is difficuit to classify, if two groups are recognized. It would
seem to belong with Benthopecten restricted, but in the type-specimen I find
numerous plates with two and three spinelets. Another’ specimen, doubtfully
referred to this species, and which, if not acanthonotus, is very closely related, has
five or six spinelets surrounding the enlarged spines of the disk. The type also
has two or three small spinelets next to some of the enlarged spines. The type of
Pararchaster is not so well differentiated from Benthopecten as some other species
included in the genus by Ludwig.

Thave maintained Benthopecten in the extended sense in which Verrill and I have
already used it, and coextensive with Pararchaster Sladen, Perrier, Keehler, and Lud-
wig previous to 1910. Ido this because I find great practical difficulties in using the
two generic names. I believe that it will be possible to split up Benthopecten, but I
doubt if Pararchaster can be used as a name, because its type is too closely allied to
B. spinosus.

Some hint of a possible division was afforded by the distribution of papule in
B. acanthonotus. Here the papule extend far along the ray ar are confined to the
basal portion, depending apparently upon the age of the individual. The character
is, however, too variable to be of use as a generic feature in this group, and I think
the same is true of the number of spines on the abactinal plates, unless accompanied
by some more conservative structural feature as yet undetermined.

144 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

KEY TO THE SPECIES OF BENTHOPECTEN HEREIN DESCRIBED.

a'. Inferomarginal plates with pectinate pedicellariz at base of ray only; or if occurring beyond middle
of ray, actinal spinelets clavate (b2). Odd interradial marginal plates normally present in all
interradii. Actinal interradial areas small. i ‘

b'. Abactinal plates of ray with only one slender central spinule; abactinal pedicellarize numerous
on ray extending nearly to tip; actinal and inferomarginal spines of proximal part of ray not
clavate; papulz often extending beyond eighth superomarginal . verte -acanthonotus, p. 144.

b?, Abactinal plates of ray proximally with one spinule and several minute spinelets, beyond papular
area with one to several minute spinelets; proximally the actinal and inferomarginal spines
clavate: abactinal pedicellarie usually confined to disk and base of rays; papulze extending to
fifth superomarginal Bg De Ae i ace erate tot faa) otetetere oterecee Sees -claviger, p- 150.

a?, A majority of interradii usually without odd interradials, and inferomarginals with pedicellariz
nearly to tip of ray. A few prominent spines in center of disk, the others short and stubby, robust.
Interradial areas largers-....-- ~~ 22 -e = sec cw ewe = sem enele oe mein minim minnie eielaem sie mutabilis, p. 147.

BENTHOPECTEN ACANTHONOTUS Fisher.
Pl. 22, fies. 2, 3; pl. 23, figs. 1, 2; pl. 28, fig. 3; pl. 54, fig. 5; pl. 56, fig. 2.

Benthopecten acanthonotus Fisuer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 299;
Zool. Anz., vol. 35, March 29, 1910, p. 547.

Diagnosis.—Rays five. R=164 mm., r=14 mm., R=11.7 r. Breadth of
ray at base, 16 mm. (from station 2919). Rays narrow, depressed, long, and very
gradually tapering to the attenuate extremity. Abactinal surface with lobed
plates (on papular areas) bearing a single spinule, and also one to several very
small spinelets. Pectinate pedicellarie numerous on abactinal surface, and extending
far along ray. Papular areas extensive; papule extending as far as seventh to
twentieth superomarginal. Superomarginals with one spine; inferomarginals with
two; adambulacrals with four to seven furrow spinelets and two much longer acti-
nals; mouth plates with four or five furrow and four or five actinal spines. Actinal
interradial areas small with one to three pectinate pedicellarie. Inferomarginal
plates with pectinate pedicellarie at base of ray only. Superambulacral plates
absent, but each ambulacral with a fairly high thin dorsal ridge or carina, joined
tightly to marginal plate at its outer end.

Description.—Abactinal surface beset with rather widely spaced very slender
rugose or thorn-covered spinules; on ray one to a plate, but on disk usually accom-
panied by from one to six very small spinelets, sometimes forming a circle when
numerous. Spinules are largest on disk, being there unequal, however, and ranging
from about 1.5 mm. to 6 mm. in length; on rays they are about 1.5 mm. long.
Primary abactinal plates, particularly the basals, have the longest spines. Abac-
tinal pedicellariz numerous, composed of two or three opposing series or combs
of three to six converging sharp spinelets. The apparatus is usually circular in
form and from 0.75 to 1.5 mm. in diameter, there being a total of nine to fifteen
spinelets, except far*along ray where the pedicellarie are beginning to form, and
have only four or five spinelets. Pedicellariz rather numerous on disk and more
so on ray, where they are found usually on either side of the median radial area, and
extend nearly to tip of arm. (See figure.)

Abactinal plates when examined from inner surface are seen to be irregularly
lobed. In the type the plates are more delicate than in the other specimen. Plates

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 145

may touch by the lobes or may be entirely separated. Beyond papular area plates
lose the lobes, are irregular to quadrate elliptical in form, and on the outer part of
ray are quite widely spaced. Papule numerous, one to a pore, and distributed all
over disk (except narrow interradial line), and basal part of ray, but from fourth to
fifth superomarginal to far along ray (varying from eighth or ninth to twentieth
superomarginal) the papule are confined to either side of the medioradial area.
Naturally the papule dwindle as they recede from disk. Thus the extent of the
papular area on rays is variable. The papule extend farther along ray than in
any other species of the genus.

Marginal plates thin, confined to side wall of ray, elongate elliptical in form,
the inferior margin of the superomarginal being slightly oblique to long axis of ray.
The plates of the two series are nearly opposite. Superomarginals fifty-four, each
bearing on a prominence of the center a long slender tapering spine, that of
third and fourth plate being longest (8 mm.) except for the odd interradial spine.
Two or three capillary spinelets stand on the adoral half of the plate and one or
two near the spine. Inferomarginals each with usually two spines in a vertical
row, the lower, half or two-thirds the length of the upper; the latter about the
length of superomarginal spines. A few scattered capillary spinelets stand on the
general surface. Dorsal interradial marginal bears a 10 to 12 mm. spine and
several spinelets. A large and two small specimens from station 2919 have no add
plate in either series in one interradius. The odd inferomarginal is armed with one
or two shorter spines. Between the odd inferomarginal and the first paired plate,
between the first and second, and sometimes the second and third or third and
fourth also, is a prominent pedicellaria with six to ten tapering spines in each comb.
The longest of these spines equals one-third or slightly more of the length of the comb.
The apparatus does not usually occupy the whole height of the plate. One or
two additional pedicellariz occasionally occur in the neighborhood of the seventh
to tenth plates, but beyond that inferomarginal pedicellariz are absent.

Adambulacral plates with a semicircular prominence into furrow. Armature
consists of (1) a furrow comb of four or five slightly curved, slender, blunt spinelets,
three of which are about equal, the extra ones (laterals) being shorter. One speci-
men (station 2919) has regularly six or seven furrow spinelets. There is a wide
interval between successive combs. (2) On actinal surface of plate are two slender
slightly tapering blunt spinules in a transverse series, sometimes the inner, some-
times the outer being the longer, the longest equaling one and one-half to two
plates in length.

Mouth plates large, convex actinally. Armature (1) a furrow series of four
or five spines which increase rapidly in size toward the inner angle, where the
innermost of each plate form two prominent teeth, much heavier than the others.
(2) On the actinal surface is a linear series of four or five rather long, slender,
blunt spines, decreasing in length toward outer end of plate; the inner two stand
in a series parallel with furrow, the rest parallel to suture.

Actinal interradial areas very small. In the type the second adambulacral
plate is in contact with the first paired inferomarginal; this is, however, extreme.
The two other large specimens have either the fourth or fifth adambulacral touching

57444°—Bull. 76—11——10

146 BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

the second paired inferomarginal. There may be as many as twenty intermediate
plates in each angle, although the type has only two or three. One to four pecti-
nate pedicellarie with five to seven spinelets in each comb occur on each inter-
radial area of the example from station 2919. The type has only one pedicellaria
to an area. The plates have besides one or two spinules.

Madreporic body convex, prominent, circular, with coarse irregular.centrifugal
striations. It is situated one-half to more than its own diameter from the odd
superomarginal plate.

Variations.—Only three large specimens are available, the type which lacks
the distal portion of all the rays, and two examples from station 2919. The
latter (one with one entire ray) have a considerably larger disk and larger actinal
interradial areas than the type as well as more extensive papular areas on the
ray. The rays are also wider at base. But a third specimen from the same
station has a disk relatively the same size as type, and papular areas which extend
only to the fifth superomarginal. The type is possibly a more immature indi-
vidual, which would account for the difference in development of disk, actinal
intermediate plates, and papular areas (to eighth superomarginal). The absence
in one interradius of the interradial unpaired marginal (of both series) is of interest.
One superomarginal in this interradius is slightly larger than its neighbor and
corresponds to the odd plate. In Benthopecten mutabilis the odd plate is frequently
absent; and indeed in the type of the genus, B. spinosus it is also sometimes absent,
as a full-grown specimen from station 2711, off Cape May, has paired plates in
two interradii. Variations in adambulacral spines, and in occurrence of pedi-
cellarie have already been alluded to in the description. A young specimen
with R about 50 mm. lacks pedicellarie entirely. There are six prominent disk
spines, of which the central is largest. The central slender spinule of the other
abactinal plates is characteristic, differmg markedly from the minute spinelets
of claviger or the stubby granuliform tubercles of mutabilis. In these young the
rays have a hirsute appearance. Another specimen, but a trifle larger, has
inferomarginal but no abactinal pedicellariz. It is interesting to note that a
very small example (R=11 mm.) has two inferomarginal pedicellariz, with twice
three or twice four spinelets, in each interradius; but there are no prominent
abactinal spines and no abactinal pedicellariz. There are five prominent inter-
radial superomarginals and spines. There is but one furrow spine. An example
with R=70 mm. has abundant abactinal and the usual number of inferomarginal
pedicellariz.

Type.—Cat. No. 22330, U.S.N.M.

Type-locality.— Off San Diego, California (longitude of Point Conception),
1,059 fathoms, mud.

Distribution—Southern California; 984 fathoms to 1,059 fathoms.

Specimens examined.—Seven, the type, and six from station 2919, off Cortes
Bank (west of San Diego), in same general locality as that of type, 984 fathoms,
gray mud.

Remarks.—This is a variable species, especially in the extent of the papular
areas, size of disk, and number of abactinal accessory spinelets. The type is probably

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 147

not a fully grown example, and therefore has less extended papular areas on rays,
smaller disk, and fewer accessory abactinal spinelets than two examples from station
2919. A third and smaller specimen from station 2919, has a disk relatively as small
as that of type and even less extensive papular areas. The slight differences in
abactinal spinulation are not sufficient to warrant specific separation.

In his key, Ludwig (1910) places B. pectinifer in the second section and acan-
thonotus in the first. Acanthonotus belongs in the second section, however, as
there is no sharp differentiation between disk and ray in the dorsal spinulation,
except with respect to the central, and primary basal spines. In this second section,
acanthonotus differs from huddlestoni in having two inferomarginal spines, four or
five ambulacral furrow spines, four or five oral spines, two subambulacral spines,
and in lacking inferomarginal pedicellariz throughout the ray. From pectinifer,
acanthonotus differs in the rather stouter abactinal spines, somewhat smaller mar-
ginal pedicellariz, and single inferomarginal spine; pectinifer differs slightly in hay-
ing six to eight oral spines, and five or six furrow spines (instead of four or five)
and two or three subambulacral spines (instead of two). The two species are very
close; it is possible they are varieties of one wide ranging form, since acanthonotus
shows considerable individual variation.

BENTHOPECTEN MUTABILIS Fisher.
Pl. 22, fig. 1; pl. 23, fig. 3; pl. 54, figs. 6, 6a.
Benthopecten mutabilis FisHer, Zool. Anz., vol. 35, March 29, 1910, p. 548.

Diagnosis.—In general appearance similar to B. spinosus Verrill, but usually
lacking some of the odd interradial marginal plates (always in one or the other
series, as a rule in both); abactinal, actinal intermediate, and inferomarginal
pedicellariz, the latter series extending to tip of ray, the first extending far
along ray, as a rule; (pedicellariz lacking in B. spinosus). Abactinal surface
with several large spines near center of disk, each usually with several small
spinelets surrounding; other plates of disk with one to three or four very short
stubby spinelets, and plates or ray with typically one such spinelet. Supero-
marginal with one, inferomarginal with two unequal spines. Adambulacrals
with five or six subequal furrow spinelets and two actinal spinelets. R=100
mm.; r=15 mm.; R=6.6 r. Breadth of ray at base, 17 mm. Interbrachial
arcs wide and rounded and beyond papular region the rays slope very gradually
to an attenuate extremity; from interradial line to about fifth superomarginal
they slope abruptly making the wide interbrachial angles.

Description —There are no prominent abactinal spines except near center
of disk within the radius of the madreporic body. These spines are about 6 or 7
mm. Jong, tapering and acicular, and are borne on the primary plates: central,
radials, basals, and probably infrabasals also in some cases. The other plates
of disk bear one to three or four, less commonly five to six or seven, very short
stubby spinelets and several usually occur around the base of the central spines.
On the rays each plate has usually one central spinelet, but in one specimen there
are regularly two, three, or even four, one being larger than the others. This
is aberrant, however. These spinelets are all under 1 mm. in length, and most

145 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

of them 0.5 mm. or less. The larger ones occur on the papular areas and are
rudimentary central spinules. There is great variation in their number, but they
are never numerous and never extend very far upon ray, being about 1 mm. long.

In some specimens abactinal pectinate pedicellariz: are about as numerous on
disk and ravs as in acanthonotus, but on one example they are comparatively
searce. Each comb consists of three to six short robust, blunt, spinelets, typically
much stouter and broader than in acanthonotus. While most of the pedicellariz
consist of two combs there are some with three, in which case each comb has few
spinelets. Far along ray the spinelets are very short and inconspicuous.

Abactinal plates of papular areas much less distinctly lobed than in acantho-
notus. The larger plates are roundish with a few irregular slight projections.
The intermediate small plates are roundish on disk, but on ray more irregular
with two or three lobes, or none. On papular areas the plates generally touch,
but on ray they become quite widely separated, especially on the median radial area.
Papule confined to base of ray and disk, but are absent from the center of disk
and a narrow interradial line. They do not extend as far upon ray as its width at
base.

Marginal plates thin, confined to side wall of ray, each superomarginal with
one stout tapering sharp spine, the inferomarginals on proximal part of ray with
two, on distal with one such spine. Of the two the lower is much smaller than the
other, which is about the same size as the corresponding superomarginal spine.
The general surface of plates is naked, but one to several additional small spinelets
occur on the plates of both series. Superomarginals, in type, forty-one. In the
type there is but one odd interradial superomarginal, the other four interradii
having paired superomarginals. The same specimen has three odd inferomarginals.
This character is very variable. One specimen has no odd superomarginal, and
two odd inferomarginals; another (otherwise not very typical) has five odd supero-
marginals and three or four odd inferomarginals; another has three odd supero-
marginals; another, one. The interradial superomarginal (paired or unpaired) usu-
ally bears a stout spine and is raised higher than the rest, but when paired the
spine is frequently weak.

Inferomarginal plates with a pedicellaria on the vertical suture, these occurring
all along ray with now and then a suture lacking the apparatus. Proximally
each comb has eight or nine short, stubby, basally webbed papille, which soon
become five or six, and on the outer part of ray about three.

Adambulacral plates with five or six blunt furrow spinelets and two much
larger, tapering actinal spines in a transverse series. The furrow spinelets are short
and the central only slightly longer than the laterals.

Mouth plates with five or six marginal spinelets, of which all but the much
enlarged inner tooth are similar to though a trifle longer than the corresponding
adambulacral spinelet. On the actinal surface is a linear series of four or five
longer spinules, not markedly different from those of B. spinosus.

Actinal interradial areas fairly extensive for this genus, with one to four, or
without pectinate pedicellariz. Sometimes they are incomplete (with one comb).
Each comb has three to seven blunt basally webbed papillae or spinelets. The

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 149

number and arrangement of pedicellariw is very irregular. Intermediate plates
extend to about the fourth inferomarginal and each has one to three unequal
stumpy spinules.

Madreporic body circular, situated remote from margin, so that outer edge is
about middle of minor radius, or even adcentral to the middle point. In B. spinosus
the madreporic body is nearer margin, so that its inner or adcentral edge is at the
middle of r, or outside of middle.

Anatomical notes —Superambulacral plates absent, but their place is taken
by a thin dorsal carina of the ambulacral ossicles. The outer end of this is fastened
to the superomarginal plate, while the main axis of the ossicle abuts against the
inferomarginal. Gonad confined to disk, opening close to margin and interradial
line; interradial septum membranous. Intestinal ceecum with four obtuse divi-
sions; no real distinction between dorsal and ventral stomachs. Polian vesicle in
each interradius except that of madreporie body. Two ampulle; tube feet with
small sucking disks.

Variations.—Individual variations have already been noted in the description.
The most important of these is the variable number of odd interradial marginal
plates.

Young—A young specimen, R=30 mm., r=5 mm., lacks pedicellarie and
bears a striking resemblance to young B. spinosus except that there are only two
unpaired interradial superomarginals. The characteristic abactinal spines of the
disk are present and very conspicuous, while the other abactinal plates bear a
single tiny spinelet each. A similar-sized individual of B. spinosus has only one
or two disk spines and very prominent unpaired interradials.

Type.—Cat. No. 27803, U.S.N.M.

Type-locality— Albatross station 2859, off Prince of Wales Island, British
Columbia, 1,569 fathoms, gray ooze; eight specimens.

Distribution.—Known only from type-locality.

Remarks.—As already indicated, this species in general appearance most
nearly resembles B. spinosus Verrill, but differs in having an abundance of abactinal
and inferomarginal pedicellariz#, only one large inferomarginal spine with a shorter
companion below it, shorter rays, as well as a broader disk, and in lacking one or
more odd interradial superomarginals. From B. pedicifer (Sladen) it differs in
having much shorter rays, larger disk, less numerous furrow spinelets, one less
actinal adambulacral spine, more prominent abactinal disk spines, and also in respect
to the variable odd interradials. B. pedicifer was taken by the Challenger south
of the Cape of Good Hope and west of the Crozet Islands. Sladen notes the same
variation in abactinal spinelets in pedicifer that occurs in mutabilis. From B. acan-
thonotus, mutabilis differs in respect to the abactinal spinulation, the former having
a delicate spinule on each plate of ray, not one or more very insignificant spinelets;
in the restriction in acanthonotus of the inferomarginal pedicellarie to the base of
ray. All pedicellarie in mutabilis have shorter and thicker often webbed spinelets,
and the actinal interradial regions are larger and fuller. The difference in inter-
radial plates also holds, and the madreporie body is nearer margin in acanthonotus.
From B. antarcticus Sladen, mutabilis differs in having only one superomarginal

150 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

spine, and two subambulacral spines, and from semisquamatus in having a much
broader disk, shorter and proximally broader rays, shorter spinelets on rays, numer-
ous inferomarginal pedicellariz, and a variable number of odd interradial plates.

BENTHOPECTEN CLAVIGER Fisher.
Pl. 22, fig. 4; pl. 54, fig. 4; pl. 56, figs. 1, 1a-b.
Benthopecten claviger Fisuer, Zool. Anz., vol. 35, March 29, 1910, p. 548.

Diagnosis—Rays five. R=110 mm.; r=12 mm.; R=9.1 r. Breadth of ray
at base, 14 mm. Disk small, rays long and evenly tapering; interradial angle
abrupt. Abactinal surface of disk with numerous conspicuous thorny-surfaced
spines which extend upon rays for a short distance, these surrounded at base by
eight to fifteen or even more very small thorny spinelets; numerous intermediate
plates with one to five tiny spinelets; on ray plates with one to five similar spine-
lets forming a fine nap. Interradial odd superomarginal very prominent. Abac-
tinal pectinate pedicellariz present but usually variable in number. Inferomarginal
and actinal adambulacral spines of proximal part of ray club-shaped, the former
two, the latter two (sometimes three) in number. Furrow spines four to six,
slightly curved. Pectinate pedicellarize between proximal inferomarginals, rarely
farther along ray. The most characteristic features of the species are the clavate
actinal and inferomarginal spines, the prominent thorny disk spines and the
extremely delicate abactinal spinelets which often are exceedingly small. Tube
feet with small sucking disks.

Description.—Disk and basal part of rays (papular areas) covered with numer-
ous spaced, tapering, blunt, superficially thorny spines, which decrease in size
from center of disk outward, ranging from 5.5 mm. to 1.25 mm. in length. These
stand on an eminence of plate which is surrounded by eight to fifteen or even
more very small, delicate, microscopically thorny spinules or spinelets, ranging
from about 0.2 to 0.4 mm. in length. Numerous small intermediate plates bear
one to five very tiny spinelets (about 0.2 mm. in length) which cannot be seen
without a glass. Plates of ray bear one to five minute sharp, slender spinelets
about 0.3 mm. long at base of ray and 0.1 mm. toward tip. These groups of
spinelets (often standing on plate in a series) form a sort of fine uniform nap on
the abactinal surface of ray. Some of the major spines are slightly club-shaped.
Abactinal pectinate pedicellarise are variable in number. Typically they are
absent or very few on rays, and few on disk. One specimen from the type-locality
has numerous pedicellarize along rays, as well as scattered over the disk. Each
apparatus is subcircular or elliptical and has two to four robust, but short, papille
in each comb. These blunt comb-spinelets are many times thicker, but only
slightly longer than the spinelets of abactinal plates. Each apparatus is from 0.5
to 1 mm. in diameter.

Papule extend along ray to about fifth or sixth superomarginal. There is a
small area in center of disk and along each interradial line from which they are
absent. On the papular areas the abactinal plates vary greatly in shape with age.
In adults they are strongly lobed, the larger with six or seven lobes. The smaller

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 151

secondary plates have two or three lobes and all plates touch and usually join by
these. Outside the papular areas the abactinal plates are roundish, and far along
ray are spaced in the median radial areas. The ‘primary apical” plates are
prominent in small and medium-sized specimens.

Marginal plates in form not different from the two preceding species. Plates
of the two series are obliquely opposite, but not alternate, the lower plate being
slightly farther distad than the upper. The intervening suture is oblique. Supero-
marginals, about forty in number, with a prominent boss on the upper rounded
margin bearing a single, tapering, bluntly pointed, thorny (almost hispid) surfaced
spine, proximally 5 or 6 mm. long; exceptionally two are present on one or two
plates; seven or eight small thorny spinelets are irregularly distributed over the
surface of the otherwise naked plates. The odd interradial superomarginal is
prominent, is elevated half its height above the level of the other superomarginals,
and bears a single upright spine 7 mm. long.

Inferomarginals bear two spines (or sometimes only one) similar, and the
longer subequal, to those of upper series, except at base of ray where both spines
are typically club-shaped, with a broader tip than base. Farther along ray,
however, the spines taper. The lower of the two is the shorter. Ten or twelve
spinelets are scattered on surface of plate. The unpaired inferomarginal bears
one or two short club-shaped spines with sometimes one or two accessory club-
shaped spinules. Typically the proximal first, second, and sometimes third, or
first and third, or first only, vertical inferomarginal sutures have a pectinate pedi-
cellaria with three to six robust sharp converging spinelets to each comb. The
aberrant specimen noted above as having numerous pedicellarie on the abactinal
surface of rays has the inter-inferomarginal pedicellaria extending far along rays.

Adambulacral plates wider than long; the furrow series of consecutive plates
well spaced. Armature (1) a furrow series of four to six subequal slightly curved
spinelets, the median two or three about as long as plate, or the laterals markedly
shorter. (2) On the actinal surface are two strongly club-shaped spines in a
transverse series, with sometimes proximally a third much smaller one at outer
end of the line. Far along ray the spines are not so obviously club-shaped, or
they may taper a trifle, but are blunt and stout at tip. One or two additional
spinelets stand on the adoral side of the plate. The form of the proximal actinal
adambulacral spines seems to be characteristic and holds in specimens from widely
separated localities.

Mouth plates with marginal graduated series of five or six spines, the inner
and sometimes the next much enlarged; parallel to median suture is a linear series
of about six actinal club-shaped spines, which become slenderer as they proceed
outward.

Actinal interradial areas small; much smaller than in the B. mutabilis; six
to eight plates in each area, these extending to the second inferomarginal. Each
plate has a few unequal clavate spinelets, one usually larger then the rest. In the
aberrant specimen already referred to, each area has one to five pectinate pedicel-
lariee with three or four pointed spinelets in each comb. Another specimen has

152 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

one or two rudimentary pedicellariz (the spinelets being very short) in each area.
The type has none.

Madreporic body within its own diameter from the interradial plate, and sur-
rounded by about five prominent spiniferous plates. The adcentral of these is
the primary basal.
ral plates; the ambulacral ossicles pro-
duced dorsally into a thin carina as in other species of genus (except at base of ray).

Young.—From station 3601 are a number of young and middle-sized speci-
mens. These exhibit most of the characteristics of the adult, especially the club-
shaped inferomarginal and actinal adambulacral spines, pedicellariz, and minute
abactinal spinelets. The spine-bearing plates of disk are very prominent, but the
spines themselves are not so numerous nor so prominent as in the adult. The
thorny surface of the spines is quite apparent. Far along ray the abactinal plates
have only one or two minute spinelets. In such a specimen (R=43 mm.) there is
a considerable area at center of disk in which no papule occur and the actinal
interradial areas are relatively as large as the types.

Type.—Cat. No. 27804, U.S.N.M.

Type-locality.— Albatross station 3788, off Cape Blanco, Oregon, 1,064 fathoms,
green mud, three specimens.

Distribution.—Southern Bering Sea to Oregon and from 987 to 1,064 fathoms,
green mud or ooze.

Specimens examined.—Besides three from type-locality, thirteen from three
stations as follows: 3307, southern Bering Sea (lat. 53° 55’ N.; long. 170° 50’ W.),
1,033 fathoms, green ooze, two specimens; 3601, Bering Sea, midway between the
above and St. George Island, 1,044 fathoms, green mud, fine sand, ten specimens,
medium sized and small; 3607, north of Unalaska, 987 fathoms, green mud, one
specimen.

Remarks.—This species can be distinguished from the two others here de-
scribed by the characteristic clavate inferomarginal and actinal adambulacral
spines of the proximal part of the ray, and by the prominent disk spines in connec-
tion with very delicate spinelets—several to a plate. The only species described
by Ludwig from the Panama- Galapagos region with which this needs comparison is
B. spinuliger, which differs in lacking the prominent spines of disk, in having
numerous (thirteen to sixteen) nal mouth spinelets, and in ince the
characteristically formed actinal spines.

In Ludwig’s key to Pararchaster (restricted) this species falls in the first section,
next to pedicifer. It differs from this in the clavate actinal and marginal spines, as
well as in the following features: more prominent and more numerous abactinal
spines, in having three subambulacrals on proximal plates only, nineteen proximal
adambulacrals to ten inferomarginals, as well as in proportion and minor details of
armature. The two forms are not at all nearly related.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 133

Family ODONTASTERID® Verrill, 1899.

Gnathasterine (part) Perrier, Expéd. sci. du Travailleur et du Talisman, 1894, pp. 244, 251.
Odontasteride VerriLi, Trans. Conn. Acad., vol. 10, 1899, p. 201.—Fisuer, Bull. Bur. Fisher-
ies for 1904, vol. 24, June 10,1905. p. 302.

Form similar to that of the more paxillose Goniasteridx, being either pentagonal
or broadly stellate. A single recurved, more or less hyaline median spine to each
pair of mouth plates, or two such spines, side by side; in the latter case one of
these spines arises from near the apex of each mouth plate. An odd interradial
marginal plate in both series. No superambulacral ossicles. Tube feet with sucking
disks as in Goniasteride. Marginal plates well developed, either decreasing regu-
larly in size distally, or at first increasing (as in Pentagonaster Gray) so that the
rays are dilated. Abactinal, actinal intermediate and adambulacral plates as in
Odontaster.

This family includes three well marked genera, two of which are extra limital.

KEY TO THE KNOWN GENERA OF ODONTASTERID®.

a’, A recurved hyaline spine on each mouth plate, two side by side at each mouth angle.
b'. Marginals decreasing regularly in size from base to extremity of rays. ----- Asterodon % Perrier.
b?. Marginals at first increasing in size; rays dilated...............--.---- Diplodontias 6 Fisher.
a’. A recurved hyaline spine common to the two mouth plates, therefore one at each mouth angle.
Odontaster Verrill, p. 154.

Remarks.—Sladen placed his genus (@nathaster in the Pentagonasteride next
to Nymphaster, while Perrier relegated Asterodon, which when described included
both Asterodon and Gnathaster (or Odontaster as now known) in the Archasteride.
Bell in 1893, in his revision of Odontaster, followed Perrier. In 1894 Perrier placed
Asterodon, Goniodon, and Gnathaster in the subfamily Gnathasterine of the Archas-
teride. In 1899, when Verrill dismembered the Archasteride, he created the family
Odontasteridee and placed the new group next to the Goniasteride.

The positive characters of the genera included in this family fully warrant, in
my opinion, the step taken by Verrill, and JT venture to suggest that the family is
nearer the Goniasteride than to any group formerly included in the Archasteride
(which is now restricted to Archaster). The fact that all the species described
previous to Sladen’s report were originally placed in goniasterid genera*® further
bears out the unanimity of opinion concerning the structure of the forms. Never-
theless Perrier and Bell believed the group to be nearer the Archasteride, while
freely admitting its intermediate character. The goniasterid similarities are to be

a Echinodermes de la mission scientifique du Cap Horn. I. Stellérides. Mission scientifique Cap
Horn. Zoologie, vol. 6, Paris, 1891, p. 129.

b Goniodon Perrier (Expéd. sci. du Travailleur et du Talisman, 1894, p. 244); type, Pentagonaster dila-
tatus Perrier. As this name is antedated, and therefore invalidated, by Goniodon C. L. Herrick, Denison
Univ. Scientific Laboratories, Bull. 3 (April), 1888, p. 4 (type, G@. ohioensis, a mollusk), I renamed the
group Diplodontias in 1908 (Smiths. Misc. Coll., vol. 52, p. 89).

¢ These species were:

Goniodiscus singularis (now Asterodon singularis).
Pentagonaster dilatatus (now Diplodontias dilatatus).
Astrogonium miliaris (now Odontaster miliaris).
Astrogonium meridionale (now Odontaster meridionalis).
Calliderma grayi (now Odontaster gray).

154 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

found in the structure of the abactinal, marginal, and actinal intermediate plates,
the form of the adambulacral plates, the absence of superambulacral ossicles, the
form of the tube feet with their well developed sucking soles or disks. The adam-
bulacral armature is not unlike that of Gephyreaster (described in this report), which
is an annectant genus near Pseudarchaster and Mimaster. The abactinal para-
paxilla:, which are the sole ‘“‘Archasterid” reminders, are not true paxille and are
similar in structure to those of Mediaster, Nymphaster, and allied genera, which no
one would think of placing in the old Archasteride. Professor Verrill’s course has
in a measure cut the Gordian knot, and the genera are far better in a separate
family than in the Goniasteride.

Genus ODONTASTER Verrill.

Odontaster Verritt, Amer. Journ. Sci., vol. 20, 1880, p. 402. Type O. hispidus Verrill; Proc. U.S.
Nat. Mus., vol. 17, 1894, p. 262; Amer. Journ. Sci., vol. 49, 1895, p. 136; Trans. Conn. Acad.,
vol. 10, 1899, p. 205 (restricted).—Betu (part), Proc. Zool. Soc. Lond., 1893, pp. 259-262.—
Lupwia, Zeitschr. Wiss. Zool., 1905, p. 42.

Gnathaster SLADEN (part), Challenger Asteroidea, 1889, p. 285.—PrrrieR, Expéd. sci. du
Travailleur et du Talisman, 1894, p. 244.—VerrRrRiLL, Trans. Conn. Acad., vol. 10, 1899,
p. 204 (restricted to meridionalis, penicillatus, and grayi).

Acodontaster VERRILL, Trans. Conn. Acad., vol. 10, 1899, p. 204 (for Gnathaster elongatus Sladen

and Astrogonium miliare Gray).

Diagnosis.—One odd, median, hyaline, more or less movable recurved spine
on each pair of prominent mouth plates, which may be close together or separated
by an open fusiform space covered with membrane. An odd triangular interradial
plate in both marginal series. Marginal plates usually prominent, decreasing in
size distally, and covered with granules or short spinelets. Abactinal plates more
or less paxilliform, often stellate, arranged in series parallel with the radial, also in
oblique transverse rows on rays; papule single, in the angles around radial paxille,
and sometimes in center of disk. Adambulacral plates small, wider than long,
with straight furrow margin as in Goniasteride. Furrow spinelets two to five,
and several longitudinal rows of two or three on the surface of plate. Actinal
intermediate areas broad, the plates in regular series parallel with furrow, and
covered with granuliform spinelets or well-developed spinelets. Simple pedicel-
lari sometimes present. Superambulacral plates absent. Tube feet with well-
developed sucking disk.

Remarks.—The genus as here defined includes the restricted genera Odontaster,
Gnathaster, and Acodontaster recognized by Verrill in 1899. It does not include
Asterodon Perrier, nor Diplodontias Fisher (Goniodon Perrier).

ODONTASTER CRASSUS Fisher.
Pl. 29, figs. 14; pl. 56, fig. 6.
Odontaster crassus Fisner, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 302.

Diagnosis.—Nearly related to O. grayi Bell. R=21mm.; r=13mm.; R=1.5r.
Breadth of ray at base 15mm. Form stellato-pentagonal with distinct rays, these
short and blunt; disk large. Marginal plates massive, the superomarginals forming
a raised broad border to abactinal area; eight to a ray, and covered with granuliform
spinelets; odd interradial plate large; inferomarginals less tumid than superomar-
ginals and granulation coarser. Abactinal plates paxilliform on papular areas and
lobed, with fifteen to twenty-three spinelets on radial series. Papule restricted to

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 155

a central and five radial areas. Adambulacral plates with eight or nine spinelets
terete, tapering and bluntly pointed, arranged with two on furrow margin, and on
actinal surface in about three longitudinal series, two spinelets to each. Mouth
plates with a movable, recurved, conspicuous lanceolate spine common to the pair,
and a marginal series of five to eight spinelets graduated from center of series toward
either end. Actinal interradial areas large, with squarish plates in regular series
parallel to furrow, each plate bearing a group of five to twelve radiating stout, short,
pointed spinelets. Superambulacral plates absent. Tube feet with well-developed
sucking disks. No pedicellarix.

Description.—Abactinal plates paxilliform on papular areas (q. v.); elsewhere
simply strongly convex or only slightly elevated and with flattish top. Plates are
arranged in regular series parallel with the median radial which contains largest
plates with the exception of the conspicuous primary basals and central plate.
Plates are also in regular rows parallel to interradial line, but inside of the pentagon
defined by the five basals the plates are not regularly arranged. Opposite first two
marginals are five rows of paxille. Tabulum of paxilla is so formed that summit
is narrower than base, the side sloping upward gradually. Tabulum on papular
areas, surmounted by a radiating group of fifteen to twenty-three (on radial row)
slender tapering spinelets which are longer than height of tabulum. The paxille
decrease in size toward margin of area. Right in center of disk is a small isolated
papular area with paxilliform plates. The central plate has twenty-seven spinelets.
About two rows of plates along interradial line and one or two rows connecting the
basal plates (thus forming a pentagon inclosing the central papular area) have
shorter spinelets than the rest. The primary basal has thirty-five to forty spinelets
or elongate granules.

There are five radial papular areas, separated by two or three rows of plates
along each interradius, and a central area isolated by or inclosed within a pentagon
the corners of which are the five basal plates. The sides of this pentagon are about
three plates thick, although there is no regular arrangement of plates. Papule, six
about each plate, extend nearly to tip of ray, and are absent from margin opposite
first two or three marginals. Plates of papular areas slightly lobed, the lobes of
adjacent plates fitting together end to end, but overlapping slightly upon radial
series, and always in central papular area where the plates are smaller and more
crowded. Elsewhere the plates are subcircular or very irregular, three to five
sided. Here and there adjacent to papular areas are very small plates, probably
large plates in course of development.

Marginal plates massive, the superomarginals forming a broad, raised, arched
border to abactinal area; last two or three plates in contact medially, though the
last pair of plates is usually separated by wedge-shaped terminal plate. In type,
seventeen plates to side, or eight to ray; interradial slightly larger than others,
subtriangular superomarginals transversely oblong, wider than high, tumid, covered
with spaced granuliform spinelets, which are thimble-shaped on rounded margin of
ray, and decrease in size and thickness toward inner edge of plate. A narrow bare
channel between consecutive plates, and a longitudinal channel on side of ray,
separating the two series.

Inferomarginals correspond in number and position to superomarginals, but
are not so tumid. Granuliform spinulation is coarser than on superomarginals,

156 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

and increases similarly in size toward outer edge where the granules are thimble-
shaped or truncate cylindrical. The unpaired plate is not quite so large as that of
dorsal series. Rather deep channels between plates. Terminal plate, cordate, con-
vex, broad end distally ; covered with short spinelets, two being thicker than the rest.

Adambulacral plates wider than long. They are short witha narrow transverse
suture, the plates having a crowded appearance. Furrow margin straight, and
plates generally oblong like some species of Tosia and Pentagonaster. Armature
consists of two terete, tapering, bluntly pointed spinelets on the furrow margin, one
slightly smaller than the other, the longest slightly shorter than width of plate. On
actinal surface are six or seven similar but sometimes sharper spinelets which
diminish in size as they recede from furrow, and are usually disposed in three
more or less regular longitudinal series (or oblique longitudinal), two spinelets
to each, and the spinelets unequal as in furrow series; spinelets of first and second
series often slightly longer than furrow series, the outermost shorter, about the
size of adjacent actinal intermediate spinelets. The spinelets form a fairly thick
chevaux-de-frise all along the border of the very narrow ambulacral groove, resem-
bling that of Gephyreaster swifti.

Mouth plates with a movable, recurved, conspicuous, lanceolate spine, which
has an obtuse inner end or base and a sharp outer glassy tip; one to each pair of
plates; this spine is rather broad actinally, not much compressed, and nearly as long
as interradial dimension of plates. Each plate is a low isosceles triangle in shape,
the base being toward interradial suture. Along the straight furrow margin are
five to eight spinelets similar to but often smaller than adambulacral spinelets,
and graduated in length from center toward either end of series, the inner being
shorter than outer. On surface of plates five to eight spinelets form an irregular
line at side of median spine.

This recurved median spine or ‘tooth,’ which forms the principal character, in
addition to the unpaired marginal plates, of the family Odontasteride, is nothing
more than the very much developed median or odd marginal spine such as is found
in most species of Pseudarchaster, and in Gephyreaster. In growing it is bent back-
ward over the surface of the mouth plates, and comes to lie wholly on the exposed
surface of the pair, the marginal spinelets encroaching upon the inner angle of the
combined plates.

Actinal intermediate areas large. The plates are square and arranged in
regular series parallel to furrow, along which they extend three-fourths the length
of R (to fifth inferomarginal). Plates bear a group of five to twelve radiating,
stout, short, tapering pointed spinelets (on some specimens much longer on middle
of plate), occasionally forming a fairly ornate rosette.

Madreporic body rather small, situated midway between center and inner edge
of interradial plate; shape very broadly oval; striations coarse, irregular.

Color in life: marginal plates orange; paxillar area lemon yellow. (Young
specimen.)

Anatomical notes.—Anal aperture present. Intestinal caecum very large, con-
sisting of a small subcircular central portion and four main radiating divisions
the tip of each being smaller and bifid. Intestinal coecum with its branches more
capacious than dorsal stomach (less hepatic ceca). Aperture from intestinal cecum
into dorsal stomach wide. Dorsal stomach sharply differentiated from ventral.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 157

Hepatic ceeca large, extending halfway to tip of ray. Each consists of a smooth-
walled ventral half, like a trough, the upper half being wider and divided with
dorso-ventrally plicate walls. Ventral stomach small with stout retractor muscles.
Gonads in a single thick tuft on either side of the membranous interradial septum.
Polian vesicles large; apparently one on each interradius. No superambulacral
plates. Ambulacral ossicles small and crowded. First pair greatly enlarged, each
with a high rounded crest. Between these crests run transversely a thick muscle
band. The enlargement of the first pair of ambulacrals with the accompanying
development of the dorsal transverse muscles is undoubtedly connected with the
movement of the combined mouth plates. These, being angular, almost completely
close the actinostome, and the small marginal spines completely close the opening.
Owing to the small size of the actinostome the stomach is probably protruded, as
in Asterias, an assumption to which weight is given by the strong retractor muscles.
The enlargement of the first pair of ambulacrals is abrupt, and so noticeable as to
form a good generic character. The same enlargement occurs, but in a less degree,
in Asterodon singularis (Miller and Troschel). Tube feet small, with sucking disks.
No deposits in walls.

Variations.— Variation in this species is comparatively slight. The sides are
a trifle more arcute in some individuals than in others. The number of marginal
plates varies, with size, from five to a ray (eleven to side of body) in a specimen
with R=7.5 mm. to eight in the type with R=21 mm. The granulation is a little
coarser on marginal plates in one or two specimens than in type, and the actinal
spinulation is longer and closer (stations 4312, 2907). Small specimens have fewer
spinelets to actinal and abactinal plates and the papular areas are much smaller.
In the smallest specimen, mentioned above, there are about five adambulacral
spinelets, two of which stand on the furrow margin, the marginal plates are extremely
massive, and the largest radial paxillee have ten to fifteen spinelets.

Type.—Cat. No. 22333, U.S.N.M.

Type-locality.— Albatross station 4313, vicinity of San Diego, 92 to 243 fathoms,
gray sand, broken shells.

Distribution.—Monterey Bay to San Diego, California, 43 to 284 fathoms.

Specimens ecamined.—Eleven, from the following stations:

Specimens of Odontaster crassus examined.

Station. | Locality. Depth. Nature of bottom. | Number. | Collection.
| |
] | Fathoms. |
2802...... | Santa Barbara Channel, Cali- 284 Wellownnu oss sccenseneesenanes | 1 | U.S. Nat. Mus.
| fornia. |
ayecee— = Off Point Conception, Cali- 44 DNS PTS VARI sao ene ences 1 Do.
fornia.
208)2<.... Off San Nicholas Island, Call- | 45 coarse gray sand, broken shells. . 1 Do.
fornia.
4311...... Off San Diego, California..... 110-143 | green mud, fine sand............ | 1 | Albatross, 1904
aBY2 0 522 erates Os. ctoecan ben pe ore a eee ae 135-95 | fine greenish gray sand, rocks. . J 1 Do.
4313...... ty ey Rees sencwen eee ns 92-243 | gray sand, shells, rocks.......... | 1 i Do.
45S. cn Monterey Bay, California....] 118-53 | rocky.................-2--2s00-- Wea g Do.
: EES 4 | ROGRBoe ee nakasenne esd eed | Do.
66-69 green mud, rocks..............-. 1 Do.
43 beach worn pebbles and stones, 2 University of California.
broken shells.

158 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Remarks.—Odontaster crassus is the North Pacific representative of O. grayi
(Bell), to which it is very closely related, and with which it may ultimately have to
be united, if the differences which are believed to exist are not constant. O. grayi
is found in the region of the Strait of Magellan and Falkland Islands.

In grayi the paxillar spinelets number about the same as in crassus, but the
spinelets are shorter than or subequal to tabulum, while in crassus they are slightly
longer. In a specimen of grayi which I examined at the Museum of Comparative
Zoology, in Cambridge, the primary basal plate is only a trifle larger than the others;
in erassus it is much larger. This difference may be due entirely to age. In grayi
there are two furrow spinelets and four or five, rarely six, actinal adambulacral
spinelets, in about two rows; in crassus there are six or seven spinelets in three rows
(or without order). In grayi the recurved tooth is slightly shorter, and thicker on
the outer part (less attenuate distally), and the suboral spinelets are only two or
three to each plate; in crassus the tooth is distally attenuate and quite sharp, and
there are five to eight suboral spines to each plate. In grayi the innermost mar-
ginal mouth spinelets are larger than the middle ones; in crassus the marginal spine-
lets diminish markedly in length from the middle, so that the innermost spinelets are
much smaller than the middle ones, and relative to the length of the tooth much
smaller than the inner spinelets of grayi, which are at least half as long as the tooth
(those of crassus being about one-fourth as long as tooth).

In this report Verrill’s groups Acodontaster, Gnathaster (Sladen, restricted), and
Odontaster are united to form a single genus, which takes the oldest name, Odoniaster.

Family GONIASTERID Forbes, 1841 (part).

Goniasteride Forses, A History of British Starfishes, etc., 1841, p. 77 (Asterina, Palmipes,
Goniaster).—VeErriLL, Trans. Conn. Acad., vol. 1, 1867, p. 343.—PErRRIER, Révision des Stel-
lérides, 1875, p. 185 (includes also Oreasteridz, Asteropide, Porania).—VicurER, Squelette
des Stellérides, 1879, p. 166 (includes also Linckiide [part], Oreasteride, Asteropide).—
VERRILL, 1899, p. 145.—FisHeEr, 1906, p. 1045.

Pentagonasteridx PrrRieR, Mém. sur les Etoiles de Mer, etc., 1884, p. 231.—SLapEN, Challenger
Asteroidea, 1889, p. 260.—PrrriErR, 1894, p. 336.—Lupwie, 1897, p. 157; 1905, p. 106.

Pentacerotide Gray (part), Ann. and Mag. Nat. Hist., vol. 6, Dec., 1840, p. 275 (included
Oreasteride, Goniasteride, Linckiidee, Echinasteride, Uniophora, Asterope, Nepanthia).

Antheneide Perrier, 1884, and authors.

Phanerozonia with thick and massive marginal plates, large disk, usually con-
spicuous primary apical plates, and tessellate abactinal and actinal intermediate
plates; abactinal plates polygonal, circular, or stellate, sometimes united by internal
radiating ossicles, or forming a close mesh with numerous secondary intermediate
plates; the plates may bear a central tabulum, paxilliform in structure, or may be
simply flat, naked, or covered with granules, or bear an enlarged spine. Papulz
usually confined to radial areas. The plates may be obscured by a tough skin, of
varying thickness, which is superficially smooth, or covered with granules, or
granules and pedicellarie. Tube feet with large sucking disks. Superambulacral
reece es rudimentary, or absent. Pedicellariz foraminate, excavate, or

vivalved.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 159

Remarks.—The vicissitudes undergone by this group are indicated in a general
way by the synonymy. When Forbes instituted the family in 1841, in his notable
work A History of British Starfishes, he included only Asterina, Palmipes, and
Goniaster. The first two genera belong to a different order, and the group was
obviously quite unnatural. About the same time Gray published A Synopsis of
the Genera and Species of the Class Hypostoma (Asterias Linneus), and included
such genera as were then known in a large and unnatural family, Pentacerotide,
which comprised the Oreasteride, Goniasteride, Linckiide, Echinasteride, Asterope,
Uniophora, and Nepanthia. This family was evidently built around Culcita and
Oreaster as a nucleus. In 1867 Verrill adopted the name Goniasteride, including
the following genera: Oreaster, Nidorellia, Paulia, Asterope, Parasterina (Patiria],
Asterina [Asteriscus]. The family in that sense was not equivalent to the present
Goniasteride, but Verrill evidently had other genera as well in mind. In 1875,
Perrier, in the Révision des Stellérides, excluded Ferdina (Linckiide), and the various
forms of Asterinide adopted by Verrill, his family being equivalent to the Gonias-
teridz as here used, with the addition of the Oreasteridx, Asteropide, and Porania.

Viguier (1879) in his Squelette des Stellérides modified Perrier’s classification
including Fromia, Metrodira, Ferdina, but excluding Porania and all the Asteropide
except Asterope [Gymnasteria]. Porania and the balance of the Asteropide were
placed in the Asterinide. In 1884, Perrier, in the Mémoire sur les Etoiles de Mer,
relinquished the name Goniasteride, dividing the family as constituted in 1875 into
four, namely, Pentagonasteride, Antheneide, Pentacerotide, and Gymnasteride.
Sladen (1889) adopted this classification, while expressing his doubts as to the
validity of the Antheneide (pp. 260, 338). Sladen’s Pentagonasteride included
one genus, @nathaster (=Odontaster), which is now placed in a separate family, and
it lacked the Pseudarchasterine, Antheneine, and Amphiaster. Perrier, in 1894,
(Expéditions Scientifiques du Travailleur et du Talisman) rearranged somewhat
the genera of the Pentagonasteride.* He excluded Gnathaster, as also Leptogonaster
and Mimaster, and its allies, relegating them to the Archasteride. While there
are reasons in favor of so classifying Mimaster, none can be adduced to support such
a treatment of Leptogonaster. Verrill,in 1899 (Revision of Certain Genera and
Species of Starfishes), established two new subfamilies, the Mediasterine for Medi-
aster, Nymphaster, and Nereidaster, and the Hippasteriine for Hippasteria and
Cladaster. He added to the Pseudarchasterine Paragonaster and Rosaster, and
separated Odontaster and its allies in a distinct family, the Odontasteride. He
recognized the Antheneide with Anthenea and Pseudoreaster, and placed the Mimas-
terine in his new family Plutonasteride. He also definitely reinstated Gonias-
teride as a name, and renamed the “Goniaster’’ obtusangulus of Sladen,? Pseudo-
reaster.

alt may be noted parenthetically that Goniasteride was dropped because the name Goniaster
was without reason fastened upon Pseudoreaster obtusangulus (Lamarck), a species not even mentioned
by Agassiz in the original description. Thus the placing of ‘‘Goniaster” in the Antheneide necessitated
the new name Pentagonasteride.

+ Most writers have overlooked the fact that Agassiz definitely cited Asterias tessellata Lamarck as
the type of Goniaster. (Mém. soc. sci. nat. Neuchatel, vol. 1, 1835, p. 145.)

160 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

In 1906 (Starfishes of the Hawaiian Islands) I accepted Verrill’s classification,
but separated Pseudarchaster and its allies as a separate family, hoping to reduce
the Goniasteride to a more homogeneous assemblage. The objection to this course
is the presence of a pretty complete series of intermediate forms which render the
adequate definition of families extremely difficult.

It would be very convenient to recognize the following families: the Mimas-
teride (Mimaster, Gephyreaster); the Pseudarchasteride with subfamilies Pseud-
archasterine (Pseudarchaster, Aphroditaster, Paragonaster) and Mediasterine
( Mediaster, Ceramaster, Nereidaster, Nymphaster, Rosaster), and possibly Nectriine;
the Goniasteride with subfamilies Goniasterine, Chitonasterine, Leptogonasterine,
and Hippasterine; the Antheneide (Anthenea, Pseudoreaster). To do this would
also necessitate the admission that there are no definite family boundaries. For
instance, Pseudarchaster has been found to possess incipient bivalved pedicellariz
(Ps. dissonus). Pseudarchaster also has definite superambulacral ossicles and
paxilliform abactinal plates. Its near relative Paragonaster has (in P. formosus)
rudimentary superambulacrals and less elevated tabula to the abactinal plates.
The structure of the mouth plates of Gephyreaster allies that form to Pseudarchaster,
while the adambulacral armature is not that of the Pseudarchasterine, but of the
Mimasterine; the genus is nearly intermediate. Among Mediaster and its allies,
Rosaster (with rudimentary superambulacrals) approaches pretty closely the struc-
ture of Paragonaster (with the exception of the pedicellariz), while Mediaster itself
(with rudimentary superambulacrals) is almost indistinguishable superficially from
some forms of Ceramaster. The last group has usually been called either Pentago-
naster or Tosia, and is really intermediate between the Mediasterine and Goniasterine.
Its species have either high or low abactinal tabula, and no trace of superam-
bulacrals. They are closely allied to perfectly typical Goniasterine, such as
Tosia (restricted), Plinthaster, and Pentagonaster. The line between a restricted
family Goniasterid and Pseudarchasteride would be difficult to draw, although
the extremes are very different. In all genera the actinal surface is much more
uniform than the abactinal, and the structure of the actinal plates is practically
the same.

The fact that Hippasteria is usually classed in the Antheneide (although accord-
ing to Viguier, Verrill, and the writer it is closer to typical Goniasteridx) indicates
in a measure that no hard and fast line can be drawn between a restricted Gonias-
teridee and Antheneidx. Cladaster Verrill has the abactinal surface exactly like
that of the Goniasterine, while the actinal is more nearly like Hippasteria. Clad-
aster, in fact, is intermediate and may be classed in either group. The only feature
in which the Antheneidee might be said to differ from the Goniasteride is in the
possession of a thick abactinal skin (which is, however, evidently homologous
with the thinner skin of Anthenoides, Leptogonaster, or Stellaster). For the stellato-
reticulate structure of the abactinal skeleton is similar to that of Hippasteria (but
Cryptopeltaster, classed by Ludwig as Hippasteria, does not have the secondary
abactinal plates in the form of connecting ossicles, and Cladaster Verrill does not
have even secondary abactinal plates). The pedicellarie are similar in the two
groups, but other genera not belonging in either have them (Anthenoides, Gilbert-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 161

aster). If one were to use as a family character the abactinal membrane (to which
one is practically reduced by elimination), what about Leptogonaster and its allies,
or Chitonaster? These would be equally deserving of family rank. It seems that
the characters of the Antheneide are really not characteristic in the sense of exclu-
sive and that the family is itself an artificial one.

Consequently the limits of the Goniasteride have in this paper been somewhat
extended to include the Antheneidx of Perrier and authors, and Pseudarchaster and
its close allies. Thus the group corresponds to the Pentagonasteridx of Sladen (1889)
and in addition Pseudarchaster, Aphroditaster, Anthenoides, and Amphiaster. The
last was classified by Sladen in the Oreasteride (a family very close to Goniaster-
ide), but its affinities are rather with Goniaster. If it were left in the Oreasteride,
there would be some trouble including it within the family diagnosis.

Within this family the range of outward form is great, between such genera as
Pseudarchaster or Nymphaster on the one hand, and Hippasteria, Anthenoides, and
Anthenea on the other. The genera agree in having well developed marginals, large
actinal interradial areas, with the intermediate plates flat, tessellated, and fairly
regular; block-like adambulacrals with a straight or not very prominent furrow
edge; and abactinal plates either flat or tabulate, in mosaic, or variously joined by
intermediate plates, but never in the form of isolated paxille. The tube feet
always have a well developed sucking disk.

The scheme of classification adopted is as follows. The asterisk indicates
genera of which I have examined species.

Family GONIASTERID.® Forbes, emended.
Subfamily MIMASTERIN AS Sladen.

Genus MIMASTER Sladen.
Genus *GEPHYREASTER? Fisher.

Subfamily PSBEUDARCHASTERINZ® Sladen.

Genus *PSEUDARCHASTER Sladen.
Genus APHRODITASTER Sladen.
Genus *PARAGONASTER Sladen.

Subfamily NECTRIIN Perrier.

Genus *NECTRIA Gray.}

@ Remarks have already been made concerning the intermediate character of thisgenus. Theabac-
tinal plates and mouth plates recall Pseudarchaster, while the marginal and adambulacral plates, with
their characteristic armature, are similar to those of Mimaster. Mimaster itself is the least goniasterid
of any genus in the family.

b Nectria is placed in a separate subfamily on the basis of the intermarginal papule. Sladen places
the genus near Mediaster, but Perrier makes a special subfamily Nectriine (1894, p. 333). In 1875 he
placed Nectria with the Goniasteride and in 1878 in the Linckiidew, from which Sladen removed it in
1889. Perrier and Sladen do not mention the important character of the intermarginal papulw. Pedi-
cellarie and superambulacral plates are also present. For further details, see Nectria, below.

57444°—Bull. 76—11——11

162 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Subfamily GONIASTERIN Verrill.

Genus *ROSASTER Perrier.

Genus *NYMPHASTER Sladen.
Genus *NEREIDASTER Verrill.
Genus *MEDIASTER? Stimpson.
Genus *CERAMASTER (Verrill) including PHILONASTER Keehler.
Genus *PLINTHASTER (Verrill) including *PYRENASTER Verrill.
Genus *LITONOTASTER Verrill.
Genus EUGONIASTER) Verrill.
Genus *PELTASTERSE Verrill.
Genus *TOSIA#¢ Gray.

Genus *PENTAGONASTER¢ Gray.
Genus *SPHARIODISCUS Fisher.f
Genus *GONIASTER Agassiz.
Genus *AMPHIASTER Verrill.
Genus *GONIODISCASTERg Clark.
Genus *ICONASTER Sladen.

Genus *ASTROCERAMUS Fisher.
Genus JOHANNASTER Kehler.
Genus *CALLIDERMA Gray.

Genus *CALLIASTER Gray.

Genus MILTELIPHASTER Alcock.
Genus *GILBERTASTER Fisher.
Genus CIRCEASTER Kehler.
Genus LYDIASTER Keehler.
Genus *CLADASTER? Verrill.

@ Mediaster is placed by Perrier in the Astrogoniinz, that is, the Pseudarchasterine. There is much
to be said in favor of this classification. Verrill places the genus, along with Nymphaster and Nereidaster,
in a special subfamily Mediasterine. There is no subfamily difference between Mediaster and Cera-
master, and to include the latter in the Mediasterine would be to annex species which some writers even
now call Tosia and Pentagonaster. Ceramaster forms a perfect connecting link between Mediaster and
its allies and Pentagonaster and its allies. All the genera near Mediaster (Rosaster, Nymphaster and
Nereidaster) are near the typical Goniasterine in the structure of the actinal intermediate, adambulacral,
and mouth plates, and in having always true bivalved or foraminate pedicellarie. These are always
lacking in the Pseudarchasterine, where the pedicellarie, if present, are spiniform-fasciolar, or are incip-
ient bivalved, and derived from the former. I have merged the Mediasterine with the Goniasterine.

+ Type, Pentagonaster investigatoris Alcock. ‘‘ Eugoniaster is related to Peltaster but differs in having
the abactinal plates all small and similar, and also-naked centrally, and in having the marginal plates
mostly naked, except around the margin. The large bivalved pedicellarie are similar in the two
genera.”’ (Verrill.) The genus differs from Plinthaster in the character of the pedicellari, and in the
adambulacral armature, which is graded from the furrow to the actinal surface. There are nosecondary
plates, and the abactinal plates are flat and rounded. The actinal plates are granulated.

¢ Peltaster Verrill, 1899, p. 168; type, P. hebes Verrill =Goniaster nidarosiensis Storm, 1881 (Grieg,
Bergens Museums Aarbog, 1905, No. 3, p. 3). I have seen the type in the Yale Museum.

4 Tosia Gray, 1840 (Dec.), p. 281; type, 7. australis Gray; monotypic. There are numerous speci-
mens of 7. australis in the Museum of Comparative Zodlogy.

¢See notes on Pentagonaster, beyond.

J For diagnosis, see notes beyond.

9 Notes on some Australian and Indo-Pacific Echinoderms. Bull. Mus. Comp. Zodl., vol. 52, No. 7,
1909, p. 110. Type, Asterias pleyadella Lamarck: see notes below.

h Cladaster may be, with reason, placed in the Hippasteriine. The genus lacks entirely, however,
the abactinal secondary plates which are characteristic of the three genera placed in the Hippasteriine, as
well as the characteristic abactinal and marginal tubercles and bivalved pedicellarie. The abactinal
skeleton therefore does not differ materially from that of Tosia or Pentagonaster. Cladaster could be

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 163

Subfamily HIPPASTERIIN CE Verrill.

Genus *HIPPASTERIA Gray.
Genus *EVOPLOSOMA Fisher.
Genus *CRYPTOPELTASTER Fisher.

Subfamily LEPTOGONASTERINZ® Perrier.

Genus **OGMASTER von Martens.
Genus *LEPTOGONASTER Sladen, including ANTHENIASTER Ver-
rill.

Genus *ANTHENOIDES Perrier.
Genus *STELLASTER Gray.

CHITONASTERIN4E£ new subfamily.
Genus CHITONASTER Sladen.

ANTHENEIN CE new subfamily.

Genus *ANTHENEA Gray.
Genus PSEUDOREASTER)? Verrill (GONIASTER Perrier, but not
Agassiz).

NOTES ON CERTAIN GENERA OF THE GONIASTERIDZ.

PARAGONASTER.—Paragonaster is classified by Sladen in the Pentagonasteride
while Pseudarchaster and Aphroditaster are placed in the Archasteride. Perrier
classes all three in the same subfamily (Astrogoniine) of the Pentagonasteride.
Paragonaster is evidently close to Pseudarchaster, as the adambulacral armature,
mouth plates, actinal intermediate plates, and inferomarginals are very similar in
the two groups. The unpaired prominent tooth at the inner angle of the combined
mouth plates is a good correlative character to indicate relationship, and is one used
by Perrier (1894, p. 337). Verrill (1899, p. 196) also places Paragonaster next to
Pseudarchaster and Aphroditaster.

Necrria.—Through the kindness of Dr. H. L. Clark, I have received for examina-
tion two specimens of Nectria, one from “‘ Australia,’ No. 474, Museum of Compara-
tive Zoélogy, and the other, No. 1932, from ‘“‘Westernport, Victoria.”

The former specimen seems referable to N. ocellifera Lamarck, but the latter with
R about 80 mm. is possibly a new species, as it differs in having the tabula of the
abactinal plates very flaring at the top, hiding the papular spaces beneath, and in
possessing numerous abactinal and fewer actinal slender jawed foraminate pedi-
cellarie. Perrier and Viguier state that the genus lacks pedicellaria.

The following notes on the anatomy were made from the larger specimen:

The abactinal plates are in the form of relatively very large tabulate parapaxille,
the crown of which flares greatly and is covered with close set granules. The base
is not so large as the crown and is five- or six-lobed. The lobes of adjacent plates

classed in the Hippasteriinz on the basis of the actinal skeleton and the few and heavy adambulacral
spines. The pedicellaria# are more characteristic of the Goniasterinew although of a form not unknown
in the Hippasteriins (as, forexample H. californica). The two aberrant genera Gilbertaster and Cladaster
have been here classed at the end of the Goniasteriine, rather than within the Hippasteriine. If other

species are discovered which disprove the value of the abactinal features referred to, it will be well to
replace Cladaster, at least, in the Hippasteriine.
@The unusual characters of Chitonaster fully warrant its segregation in a separate subfamily.

b Pseudoreaster Verrill, 1899, p. 148; type, Asterias obtusangulus Lamarck.

164 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

do not touch except on outer half of ray but are joined by internal radiating ossicles
as in Mediaster, there being four to six radiating from each plate. The ossicles are
fairly large, except on outer part of ray. In the large triangular meshes left by the
ossicles papule emerge, but it is not possible to determine the number. On the side
of many of the tabula, usually obscured by the overhanging rim, is a curious pedi-
cellaria with three to six slender curved jaws, about as long as the height of the
marginal granules. On the actinal surface, somewhat similar pedicellariz occur
on a few intermediate and adambulacral plates near the mouth plates. These
have two, three, or four very slender, slightly tapering jaws.

A highly important feature of the anatomy of Nectria is the presence of papulx
between the marginal plates. Near the base of the ray there appears to be more
than one to an area, but on the outer part there is only one. The series ceases at
about two-thirds the length of ray.

The adambulacral armature and mouth plates are essentially goniasterid. In
the Victorian specimen there are three longitudinal series of robust truncate subpris-
matic or quadrate spinelets to each plate and about three in each series. The char-
acter of the armature is not unlike that of Mediaster and some species of Ceramaster.

Superambulacral plates are present but not very well developed. They abut
against the lower end of the ambulacral ossicles and against a single row of spaced
supplementary internal intermediate plates, apparently distinct from the regular
actinal intermediate plates. Proximally some of the superambulacrals abut directly
against the actinal intermediates.

Interbrachial septa made up of large flat imbricated plates. First ambulacral
ossicle enlarged. Tube feet with large sucking disks and double ampulle. Actinal
intermediate plates imbricated internally, arranged in cheyrons.

Nectria agrees with typical Goniasteride in the tabulate structure of the abac-
tinal plates, the form of the pedicellariz, and in the character of the adambulacral,
mouth, and actinal intermediate plates, but differs from the Goniasterine in having
a definite series of papule between the two series of marginal plates. The genus
seems to be most nearly allied to Mediaster, but on account of the presence of inter-
marginal papule is placed in a separate subfamily, the Nectriine, Perrier 1894.
Perrier, however, failed to designate the really important character which distin-
guishes Nectria from other Goniasteride.

RosastEer.—I have examined Perrier’s types in the Museum of Comparative
Zoblogy. Verrill places the genus near Paragonaster, but I think it is much nearer
Nymphaster, and well within Verrill’s Mediasterine. Perrier states that no pedi-
cellarie occur, but on the actinal surface of one of his types (No. 450, Coll. Mus.
Comp. Zo6l.) in two interradii there is a good-sized spatulate foraminate pedicellaria
whose expanded jaws reach nearly across the plate. A smaller specimen (R=16
mm.) in the Stanford Collection (received from the Museum of Comparative Zodlogy)
also has two conspicuous ventral pedicellarie. Rudimentary superambulacral
plates (few in number and of very small size) are present.

As compared with Nymphaster, the radial paxilliform plates are higher and more
spaced, and the interradial plates are slightly tabulate, not flat, as in Nymphaster.
Marginal plates of Rosaster have deep grooves between them, but no spinelets in the

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 165

grooves. In Nymphaster these plates have no abrupt furrow, at least in the dorsal
series, being like Mediaster. The adambulacral armature differs from that of
Nymphaster in having on the surface of the plate spaced from furrow series a row of
two or three stouter and longer tapering spines, one of which is larger than the other
two. The furrow spinelets are long, slender, compressed, and four or five in number.
The general facies of this species, as well as the character of the pedicellariz, ally it
to Mediaster rather than Pseudarchaster.

Printuaster Verrill and Prrenaster Verrill.—The type of Pyrenaster Verrill
(1899, p. 166) is Pentagonaster dentatus Perrier (1884, p. 242). According to Verrill
Pyrenaster ‘‘is distinguished from Tosia and Pentagonaster especially by the exis-
tence of smaller secondary rounded plates between the primary abactinal plates,
and also by the greater specialization of the furrow series of adambulacral spines,
for these do not grade into the actinal granulation, as they do in the genera referred
to. In this respect the genus is more nearly like Mediaster, etc.”” I have examined
the type, P. dentatus (No. 408, Coll. Mus. Comp. Zoél). The only secondary
abactinal intermediate plates, on the radial areas, are one or two proximally.
There are a few small plates here and there on the center of disk. But Pentagonas-
ter perriert (No. 418, Coll. Mus. Comp. Zo6l.) the type of Verrill’s Plinthaster (which
I consider of generic rank) has secondary plates proximally on the radial areas and
also a few on interradial areas. There are more secondary plates than in the type of
dentatus. The adambulacral armature is of the same character in both species. A
difference between the two species (but hardly of generic importance) is to be found
in the shape of the pedicellarie. In dentatus the pedicellariz are few, small, forami-
nate, with narrow spatulate slightly tapering chisel-like jaws, while in perrieri they
are much larger with broad jaws (nearly as wide as high) which on the actinal sur-
face fit into slight depressions of the plate. The abactinal pedicellarie are of the
same form (similar to those of Tosia tuberculata Gray) but smaller. In both species
a few distal superomarginals are in contact, but this is subject to individual variation.

I think there is not sufficient difference to make two genera. The general
appearance of the two groups is strikingly similar, and added to this the abactinal,
actinal, and adambulacral plates are the same. Plinthaster has page priority and
should be adopted as the name for the enlarged genus. The following are the

species:
Plinthaster perrieri (Sladen). Plinthaster nitidus (Verrill).
Plinthaster comptus (Verrill). Plinthaster ceramoideus (Fisher).

Plinthaster dentatus (Perrier).

Verrill refers Pentagonaster affinis Perrier to Pyrenaster, but as he did not see
the type, his specimens were not correctly identified. I have examined Perrier’s
type (No. 403, Mus. Comp. Zodl.) and it is not a Pyrenaster or a Plinthaster. The
abactinal plates are wholly granulated and the radial ones are subtabulate. Second-
ary abactinal plates are evident in a small example about one inch in diameter, but
in the type they are not markedly different in size from the primary plates. The
species may be referable to Ceramaster, although it is by no means typical.

Lrronoraster Verrill, 1899, p. 171; type, Pentagonaster intermedius Perrier,
1884, p. 243, pl. 5, figs. 5 and 6.—I have seen specimens of the type-species. The

166 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

genus is remarkable for the thin dorsal integument, few papular pores, and numerous
(seven or eight) small furrow spinelets. The marginal plates are unusually small.
“Abactinal plates polygonal, flat, thin, closely united, finely granulated, with two
or more rows of granules around the edges, but with a small, central, round, naked
area, in the type. Papular pores rudimentary, few, small, obscure, not visible
except when the plates are denuded; they occur only between the three central rows
of plates, in a very circumscribed basal radial area. Actinal plates granulated,
rather large, angular, of various forms, not forming regular rows. A small elongated
pedicellaria, with two, three, or four spatulate blades, occurs on the center of many
of the adambulacral plates and on some of the actinal plates” (Verrill).

Tosra Gray and PENTAGONASTER Gray.—The type of Tosia is T. australis Gray;
that of Pentagonaster, P. pulchellus; both montotypic when first characterized.
Stephanaster Ayres and Astrogonium Sladen are synonymous with Pentagonaster.
There are four specimens of the type species of Pentagonaster in the Museum of
Comparative Zoélogy. Stephanaster bourgeti is excluded from the genus and made
the type of a new genus, Sphzriodiscus.

It seems necessary to segregate the Australian species in these two genera, as
they do not agree with European or American species commonly called Pentagonas-
ter and Tosia. The latter are now classed in several quite different groups—Ceram-
aster, Plinthaster, Peltaster, Litonotaster, and Spheriodiscus.

Pentagonaster differs from Goniaster in lacking abactinal tubercles; in having
the last marginal plate of both series enlarged; in lacking abactinal secondary
ossicles, and numerous papular pores separated by intermediate granules between
the dorsal plates of papular areas; in having abactinal and actinal intermediate
plates free of granules except for a marginal series of granules (applies to adults).

Pentagonaster differs from Spheriodiscus in having the last marginal plate of
both series (not the penultimate) largest; in having the marginals entirely naked; in
lacking the flat granulated abactinal and actinal plates (these being more or less
convex and naked in Pentagonaster).

Pentagonaster is distinguished from Tosia by the enlarged distal marginal of
both series (if a marginal is enlarged in Tosia it is the last superomarginal only); in
the character of the pedicellariz, which have narrow spatulate jaws (as in Goniaster) ;
in Tosia they are low, bivalved and wider than high, or are absent altogether.

In the type of Tosia, there are very few short bivalved pedicellarixe on the
abactinal plates; the actinal surface lacks pedicellariw, but the plates are naked
centrally, there being usually two or even three or four rows of granules around
each plate. The adambulacral plates are very like those of Pentagonaster—wider
than long, with two furrow spinelets, thick and stubby, behind these two more, and
then two rows of about three granules each. The abactinal plates are strongly
stellate not only on the radial but interradial areas as well. Single papule emerge
between the lobes of plates, and are distributed all over the disk from center to
angles.

The group to which 7: tuberculata (Gray) belongs is somewhat different—that is, if
the specimens in the Museum of Comparative Zodlogy are really Gray’s species. The
abactinal plates are arranged in radial and parallel rows, but next to radial series in
a small specimen, and over most of the petaloid radial area in a larger specimen are

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 167

wholly granulated, small secondary plates wedged between the primaries. There
is a fairly distinet row on either side of the median radial series, and laterally from
this for three or four rows the secondaries are numerous but not regular in occurrence.
From the inner side the primary plates are slightly stellate with blunt lobes; papule
are single and absent from interradial areas. Marginal plates centrally bare; actinal
intermediate plates mostly subquadrate, covered with good-sized spaced quadrate
granules; adambulacral plates slightly wider than long with three or four subquadrate
furrow spines and behind these three rows of spines and granules, the armature
being graduated into actinal granulation. Many adambulacral plates bear a short
rather high pedicellaria with two oblong jaws. The abactinal plates have the
central part naked and convex, though some are wholly granulated.

Verrill classifies this species in his restricted Yosia. The presence of the
abactinal secondary plates as well as of pedicellariz higher than those of typical
Tosia suggests the alliance of the form with the less spiny Goniasters. More speci-
mens will be necessary to clear up the limits of Tosia and Pentagonaster. If any
uniting of genera takes place it should be Goniaster, Pentagonaster, and Tosia, not
Tosia with Plinthaster or with Ceramaster. The two genera under discussion differ
quite widely from either Plinthaster or Ceramaster in the character of the abactinal
and actinal plates which in the Australian-New Zealand forms are naked and con-
vex (but not tabulate) and in the marginal plates which tend to have either one or
both series with enlarged distals.

SpHzriopiscus Fisher.~—Type, Stephanaster [= Pentagonaster] bourgeti Perrier.
Marginal plates thick and wide, few, the penultimate or antepenultimate plate of
both series larger than the rest; marginal plates with small spherical granules about
the borders, and over general surface are scattered spherical granules sunk in special
shallow pits; last two superomarginals in contact. The granules about borders of
plates are rather numerous, and in several series on the outer or lateral face of plates
of both dorsal and ventral series. Abactinal plates flat-topped, only very slightly
tabulate on radial areas, covered with granules; papule absent from center of disk
and a narrow interradial strip. Actinal intermediate plates wholly granulated.
Adambulacral plates wider than long, with four or five furrow spinelets; behind these
two or three thicker, shorter spinelets, and then five to eight granules. Pedicellarie
with two slender spatulate blades which rest in inconspicuous depressions. They
may occur on all the plates. Contains two species, Spheriodiscus bourgeti (Perrier),
Cape Verde Islands, and Sphzriodiscus ammophilus (Fisher), Hawaiian Islands.

This genus differs from Pentagonaster in having the upper and lower plates
wholly granulated; in having not the ultimate but the penultimate or antepenulti-
mate marginals enlarged; in having marginals with spaced granules. The abactinal
plates are flat-topped and slightly tabulate on radial areas; the actinals are flat,
never convex.

GontasTeR Agassiz..—Type, A. tessellata Lamarck, by designation. Verrill
(1899, p. 150) has revised this genus. He gives the type as (. euspidatus Gray, to
which species the Asterias tessellata of Lamarck may be restricted, as it is necessary

a Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 171.
b Mém. soc. sci. nat. Neuchatel, vol. 1, 1835, p. 191.
¢ Idem, p. 145.

168 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

to dispose of the name. Lamarck’s A. tessellata included also” the previously
described A. granularis Retzius, 1783. Gray seems to have had a right to consider
the name invalidated by granularis, so that cuspidatus is perhaps tenable. There is
no doubt, however, that Agassiz had cuspidatus (and not granularis) in mind when
he placed tessellatus as the type of Goniaster, for his description of the genus clearly
indicates it. By most writers tessellatus will be preferred to cuspidatus on the
ground of priority. Since ¢essellatus was a composite species originally, the elimina-
tion of granularis would seem to necessitate the restriction of the name to the only
available group—that later named cuspidatus by Gray.

Amputaster Verrill.—Type, A. insignis Verrill.—Most of the abactinal plates
bear an upright blunt tubercular spine nearly as thick as the diameter of the plates,
which are roundish or substellate internally, and connected by about six radiating
ossicles or secondary plates. The iaterspaces between the primary plates, the
surface of the secondary ossicles, and the base of the tubercles are covered with
rather coarse granules. Several papule emerge in each of the six areas surrounding
a primary plate (these papular areas are covered with granules as noted above).
The first superomarginal is largest, smooth, and convex, the second has a heavy
tubercle, the third is like the first, the fourth like the second, and the next three
very strongly convex. The inferomarginals are similarly disposed. The edges of the
plates, which are subcircular, as well as the small interpolated intermarginals are granu-
lated. Nearly all the granulated actinal intermediate plates have prominent tuber-
cles smaller than those of dorsum. Each adambulacral has a prominent actinal blunt
tubercular spine (smaller than those of actinal intermediate plates), and a furrow
series of three or four blunt, straight spinelets, as long as the actinal but slenderer.

This genus is more nearly related to Goniaster than to any in the family
Oreasteride.

GoNIODISCASTER.—This genus was named by H. L. Clark® for ‘‘Goniodiscus”’
pleyadella (Lamarck). The old name Goniodiscus Miller and Troschel being unten-
able,° was changed by me to Gontodiscides, with G. sebe as type. Unfortunately
(7. sebe and G. studeri are young Culcita,¢ so that Goniodiscides along with Randasia
Gray becomes synonymous with Culcita. But “‘Goniodiscus”’ pleyadella (including
Pentagonaster validus Bell) is not a young Oreaster, as Perrier * suggested. It seems
to be related to Goniaster. There is a close granulation all over the test, but judging
from the dried specimen examined, no skin, such as characterizes Anthenoides.
Although the papule are in areas on the disk, the same holds true in adult Goniaster.
The pedicellarie are minute pincers and small bivalves. In adults the third or
fourth marginals from apex of ray are swollen slightly. Abactinal plates strongly
stellate, but there are no intermediate secondary plates, as in Goniaster, and no
marginal tubercles, only a single short tubercle on each of the five primary (?)
radials. A plate on either side of the interradial line is enlarged. Adambulacral
plates frequently have a good-sized pincer-shaped pedicellaria with tapering jaws
as high as the actinal spinelets.

@ Trans. Conn. Acad., vol. 1, pt. 2, April, 1868, p. 372.
> Bull. Mus. Comp. Zodl., vol. 52, No. 7, March, 1909, p. 110.

¢ See Fisher, Bull. U. S. Fish Comm. for 1903, pt. 3, 1906, p. 1070.
4Clark, Bull. Mus. Comp. Zodl., vol. 51, April, 1908, p. 281.

¢ Révision des Stellérides, p. 232.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 169

LeproconasTER and ANTHENIASTER.—About the only difference between Lep-
togonaster Sladen and Antheniaster Verrill is the wider distribution in the latter of sec-
ondary abactinal plates. Antheniaster epixanthus Fisher varies considerably, how-
ever, in the number of abactinal secondary plates. Young specimens have no more
than the adradial row, described by Sladen (1889, p. 330), in the type of Leptogon-
aster crispatus. The character of the abactinal plating, the granulation, actinal
interradial areas, and especially the adambulacral armature and abactinal membrane
are essentially alike in both species, or not different enough to warrant two genera.
Anthenoides is the nearest relative. Its differences are brought out in the key. As
Professor Verrill classed my epizanthus in Antheniaster, I have felt justified in using
that species for comparison.. I now think it is a Leptogonaster, intermediate in
respect to the abactinal plating between crispatus and sarissa (Alcock) and differing
specifically in details of abactinal and marginal granulation, in the adambulacral
armature, and distribution of pedicellariz.

OemasteR.—It is rather curious that this genus should have been classed next
to ‘‘Goniodiscus,” as its appearance is so unlike any species of that group. In the
specimen in the Museum of Comparative Zoélogy (467) the abactinal membrane
must be very thin, for it is scarcely apparent in the dried specimen. There are no
secondary abactinal plates, and only the plates of the papular areas have slight
lobes. There are no granules on any plates but the adambulacrals and mouth
plates. This agrees with Gray’s figure? and description” of ‘‘ Dorigona Reevesii,”
which is the same as Ogmaster capella. Sladen’s comparison of the genus with
Stellaster in his key (1889, p. 264), stating that the test is overlaid by a granulous
membrane, is not only misleading, but is contradicted in the very next line. There
is no partjcular similarity between Stellaster and Ogmaster.

KEY TO THE KNOWN GENERA OF GONIASTERID®.

I. Abactinal plates tabulate on the radial areas and often paxilliform, never flat nor covered with a
continuous membrane.
a', Superomarginal plates not contiguous on the rays, one or more series of abactinal plates reaching
the terminal plate (occasionally a few distal plates in contact).
b!. Abactinal plates stellate with paxillz; actinal intermediate areas with imbricating plates in
transverse series bearing paxilliform groups of spines.............-.------- Mimaster Sladen.
b?. Abactinal plates with tabulate paxilliform plates; actinal intermediate areas with imbricating
plates bearing codrdinate granules or spinelets.
cl, An unpaired (but not recurved) median tooth common to each pair of mouth plates; pedi-
cellarize when present fasciolar, rarely valvular; superambulacral plates present, though
frequently very small.

d', Adambulacral armature: furrow series radiating, subpalmate; actinal spines in a coérdinate
group, frequently one slightly enlarged, appressed; inferomarginals usually with a few
appressed enlarged spinelets.

e'. Several series of abactinal plates extending far along ray and usually more than one
attaining the terminal plate; rays fairly broad; superambulacral plates fairly well
developed; pedicellarie pectinate, fasciolar, or valvular. .Pseudarchaster Sladen, p. 179.

e?. Superomarginal plates of ray separated by single series of roundish or quadrate granulated
(not paxilliform) plates; superambulacral plates rudimentary.¢ No pedicellariz.

Paragonaster Sladen.
d?. Furrow series of spines not radiating or subpalmate.

@ Synopsis, pl. 7, fig. 3, 3a.
b Idem, p. 7.
¢ Superambulacral plates rudimentary in Paragonaster formosus (Verrill).

170 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

e!. Adambulacral armature in longitudinal series parallel to furrow; furrow spines numerous;
actinal interradial areas small, and postadambulacral plates with fascioles; marginal
plates without conspicuous channels between. ...------------ -Aphroditaster Sladen.

e. Adambulacral armature a compact group of similar spines in more or less evident trans-
verse series; furrow spines two or three; actinal interradial areas large with numerous
imbricating plates covered with codrdinated compact groups of numerous short spine-
lets, leaving narrow interradial channels; marginal plates with conspicuous inter-
vening channels. .....-----+-+eeeeeeeee eee e ence eeneeeees Gephyreaster Fisher, p. 175.

©. No unpaired median tooth to each pair of mouth plates; pedicellariz when present low
bivalved, or higher with two or more spatulate jaws; superambulacral plates rudimentary
or absent.
d', Abactinal plates connected by independent internal slender regular radiating ossicles over
the papular area; rudimentary superambulacral plates present.

e!, Abactinal plates medium-sized or small, roundish, not developed into very large cylindrical
tabula; marginal plates conspicuous; bivalved or two-jawed spatulate pedicellariz;
adambulacral spinelets more or less prismatic in longitudinal series, numerous; no
intermarginal papulz; no internal supplementary actinal intermediate plates.

Mediaster Stimpson, p. 196.

e?. Abactinal plates lobed, very large with large cylindrical tabula bearing coérdinated gran-
ules; marginal plates comparatively small; pedicellarize when present with two to
six slender upright jaws; supplementary internal actinal intermediate plates in neigh-
borhood of the ambulacrals; intermarginal papule...............----- Nectria Gray.

d?. Abactinal plates without internal slender radiating connecting ossicles; the plates of the
papular areas slightly to conspicuously lobed. No superambulacral ossicles.

e!. General form stellate frequently with long rays, a single series of abactinal plates reaching
the tip of ray, or the few distal superomarginals in contact medially; pedicellariz
highandspatulate is. o- tases sce sn scee saeee eee aceeee ee eee Nereidaster Verrill.

e?. General form pentagonal, never with long rays; two or three superomarginals sometimes
in contact medially; pedicellariz, if present, small bivalved, or with two or three
spatulate jaws, but not high and spatulate...........-...-- Ceramaster Verrill, p. 204.

a?. Superomarginal plates contiguous throughout most of ray, the more or less tabulate abactinal
plates being confined to disk; slender spatulate pedicellariz.

b'. Abactinal plates only slightly tabulate, granulate. No enlarged actinal adambulacral spines;
furrow spinelets not very regular nor spaced from actinal; pedicellariz all small; no rudi-
mentary superambulacral plates’. a2\-22 (a2 s2m- se teenies teeeeee eee Nymphaster Sladen.

b?. Abactinal plates of radial areas strongly tabulate, covered with short spinelets rather than
granules; furrow spinelets long, regular, and spaced from actinal; one or two enlarged actinal
adambulacral spines; a few comparatively large slender spatulate actinal pedicellariz;
rudimentary superambulacral plates ...-. Socket eset con eee eee ..-Rosaster Perrier.

IL. Abactinal plates only exceptionally tabulate (Ceramaster); usually flat or convex, roundish, poly-
gonal, or stellate; smooth, granulous or spiny; in mosaic, with or without secondary intermediate
plates which do not form, with the primaries, a close reticulated skeleton.

a‘. Body not covered by a smooth or granulous membrane, nearly or quite obscuring the abactinal
plates.

b'. Abactinal plates with upright spines or tubercles of conspicuous size, but no low bivalved pedi-
cellarise; pedicellarie, when present, spatulate or like miniature sugar tongs; more than a
pentagon of five tubercles on disk.

c'. No large sharp or blunt actinal intermediate or adambulacral spines; armature of adambula-

cral plates graduated into actinal granulation...............0..--.02-0-+ Goniaster Agassiz.

ec’, Adambulacral plates with one or more conspicuous sharp or blunt movable or rigid spines on
their actinal surface; adambulacral armature not graduated.

d'. Small granulated plates and granules between primary abactinal plates, each of which

bears a large smooth conical tubercular spine; actinal intermediate plates each with a con-

spicuous upright blunt spine; scattered intermarginal plates......... Amphiaster Verrill.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 171

@. No secondary abactinal plates; no intermarginal plates.
e!. Actinal intermediate plates without spines, or with only low tubercles; plates at least
centrally bare; the smooth marginals with one or more rigid spines or tubercles.
j'. Disk flat but fairly thick; no abactinal pedicellarie; actinal spines without swollen

MIT AG OM DING PONS! oto one Ses ose eee ee eeene te ccc cnan ccs Calliaster Gray.
J?. Disk very thin, rigid, flat; actinal spines with swollen multifid or bifid points; spatulate
pedicellarise on both surfaces of body........-.-..------------ Milteliphaster Alcock.

e?, Actinal intermediate and inferomarginal plates with appressed movable sharp spines;
all plates granulated; no tubercular marginal spines; radial abactinal spines present in
ONG EPECION= retest ena eee hie see tee ns eh ete ce oni gars oe Calliderma Gray.

b?. Only a pentagon of five short tubercles on disk, or more often none; never any abactinal spines
on rays; no superomarginal spines.
c'. Abactinal plates of rays not abruptly and conspicuously larger than those of disk.
d'. No inferomarginal and actinal intermediate spines.

e'. Marginal plates of both series not regularly decreasing in size from interradius to apex of
ray, but increasing either slightly or markedly. Sometimes the last marginal is en-
larged; sometimes the second, third, or fourth plate from the end is largest, and the
marginals are graduated in breadth on both sides of this, abruptly toward the apex and
very gradually toward a point between the enlarged plate and base of ray; the en-
larged plate is always either appreciably longer or wider (or both) than the preceding
or succeeding plates.

f'. Pedicellarize never bivalved; no tubercles on disk; papuls not in areas, single; outlines
of plates distinct; abactinal interradial plates not conspicuously enlarged.

g'. Last marginals of both series enlarged; abactinal, marginal, and actinal intermedi-
ate plates smooth, slightly convex (never tabulate) bordered by a single series of
Pranlesase cee sae na ae te ance see eee eee scecae cr Pentagonaster Gray.

g?. Not the ultimate but the penultimate or antepenultimate marginal of both series en-
larged; abactinal and actinal plates flat and closely granulated; marginals with
scattered) cranules' inpltsss-2--2e-see oe oe once ee ceeees reece s Sphzriodiscus Fisher.

f?. Abactinal interradial plates (four in each interradius) enlarged; five primary radial tu-
bercles on disk, forming a pentagon; bivalved pedicellariz of small size on abactinal,
marginal, and actinal intermediate plates, largest on latter area; abactinal plates with
minute pincers also; all plates closely granulated; abactinals stellate (no secondary
intermediate plates); papular areas with three to five papular pores (except near end
of ray where only one or two), granulated, and bearing one or two minute pincers;
actinal interradial areas large; adambulacral plates with five to eight slender graduated
marginal spinelets and a row of two or three clavate or oblong rounded granuliform
spinelets just behind the furrow series; rays slightly swollen near tip or with parallel
sides; third or fourth marginal from apex the largest; outlines of abactinal and actinal
intermediate plates not distinet Pee eer ena eae reta Goniodiscaster Clark.

é. The marginal plates evenly graduated in size, or if the ultimate or penultimate is en-

larged, only in upper series.
f'. Numerous distal marginals not united to form long slender rays.
g'. Secondary plates of small size interpolated between the primary abactinal plates.
i}. Abactinal plates flat, either smooth or granulated, but not tabulate or paxilliform.
i. Abactinal plates small, closely granulated; furrow spinelets not spaced from
actinal ones, nor sharply differentiated; bivalved pedicellariz.
Peltaster Verrill.
#. Abactinal plates naked centrally, but surrounded by a single series of granules
on radial areas (over interradial areas granules may be only scattered around
edge of plate, or absent); adambulacral plates with sharply differentiated
furrow series; no bivalved pedicellariw, but a few small chisel-bladed ones

on adambulacral plates. ..............-.-. Plinthaster (incl. Pyrenaster) Verrill.
#?. Abactinal plates tabulate on radial areas, suggesting low paxille. Secondary
plates present only in adult specimens ..........-.- Ceramaster@ Verrill, p. 204.

g?. No secondary plates on radial areas; sometimes a few on center of disk.

aC. leptoceramus Fisher.

172 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

h'. Abactinal granules, if present, of usual form.
i!, Abactinal plates of normal thickness, not very thin; papule normal, not few or
absent; furrow spinelets not unusually numerous.

j!. Abactinal and marginal plates with large spaced granules; adambulacral spines
relatively heavy, two in the furrow series, with an enlarged actinal spine;
actinal pedicellarie, with broadly spatulate blades; actinal surface resem-
bling: Hip pasteria. «..-- 52 -c1cjain olele wlnlel=wi=)nis = sieee= els Cladaster Verrill, p. 221.

j?. Adambulacral spines not thick and heavy; no conspicuously enlarged actinal
adambulacral spine.

i+. Abactinal plates smooth, at least centrally; pedicellariz when present,
bivalved.
Ll’. Abactinal plates strongly stellate, naked centrally, the smooth part con-
vex and bordered by one or two rows of bead-like granules; distal
superomarginal sometimes enlarged (but never the inferomarginal).

Tosia Gray.
2. Abactinal plates round, small, with a single series of skin-covered granules;
true bivalved pedicellarie.........:.-...--.------- Bugoniaster Verrill.

3. Abactinal plates flat, smooth, save for some small encircling granules on
radial areas; actinal plates granulated; furrow spinelets fairly nu-
merous, differentiated, and spaced from actinal adambulacral granules;
pedicellarie modified bivalved with thin jaws......- Plinthaster Verrill.

2, Abactinal plates, as well as actinals, closely granulated all over.

Ll. Abactinal interradial plates not larger than the medioradials; abactinal
plates slightly to conspicuously tabulate; general form pentagonal;
papule single, or if more than one to area, then no granules between
abactinal plates; actinal row of adambulacral spinelets not conspicu-
ously thicker than the marginal series; outlines of plates all distinct.

Ceramaster Verrill, p. 204.

. Abactinal interradial plates larger than medioradials (carinals); abac-
tinal plates strongly stellate, not at all tabulate; whole animal closely
granulate, obscuring outlines of plates; papule one to five to an area;
papular areas granulate, bearing sometimes pincer pedicellariz; general
form stellate; first actinal row of adambulacral spinelets thicker and
more granuliform than the rather slender furrow series; numerous
minute bivalved pedicellarize on both surfaces and on marginals.

Goniodiscaster (young) Clark.
7. Abactinal plates very thin, flat, finely granulated but with small central naked
area; abactinal integument quite flexible; papule rudimentary. Many fur-

row Spinelete®ais.s..jc stis. Sos Seceiamerye obs eeeeee a eae Litonotaster Verrill.
h?, Abactinal granules large, irregular, flat, plate-like; large bivalved pedicellarie
on both surfaces; no prominent spines.........-.-.------- Gilbertaster Fisher.

J?. Marginal plates numerous, and superomarginals contiguous, forming long slender rays;
abactinal plates smooth.

g'. Abactinal, marginal, and actinal intermediate plates smooth, bordered by four-
sided granules; abactinal granules flush with surface of plates, elongated, and
on radial areas occurring only on longitudinal sutures between the plates; on
center of disk and interradial areas they completely surround plates; adambu-
lacral plates with about four narrow spines, and behind these, four series graduated
to granules; adambulacral armature of the type of Peltaster and Pentagonaster;
pedicellarize on ventral surface small, with narrow spatulate blades.

Iconaster Sladen.

g’. Abactinal plates surrounded by small flush granules; marginal and actinal plates
with large spaced tubercular granules; actinal plates bordered by bead-like gran-
ules; adambulacral plates with large actinal spinules and divergent furrow series;
spatulate, excavate, serrate, ‘‘sugar-tongs” pedicellarie.....Astroceramus Fisher

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. LS

@. Appressed, distinct, sharp actinal intermediate spinules, in addition to the close granulation.
e'. Distal superomarginals contiguous; appressed inferomarginal spines; conspicuous enlarged
actinal adambulacral spines; marginal plates forming square edge to disk.

Calliderma Gray.
e?. Distal superomarginals not contiguous; no inferomarginal spines; no enlarged adambu-
lacral spines; marginal plates forming rounded edge to disk...._. Johannaster Koehler.

e. Abactinal plates of rays polygonal, flat, conspicuously larger than those of disk; marginal
and actinal plates granular, armed with pedicellarie, but without spines; no enlarged
adambulacral spines; disk large, rays long; superomarginals at first increasing, then grad-
ually decreasing in size.

d'. With bivalved pedicellarie (type of Hippasteria, Cladaster, Gilbertaster); abactinal plates
irregularly polygonal, bordered by a single series of granules, with or without a few scat-
tered granules on surface; plates or rays abruptly larger than those of disk.

Circeaster Koehler.

@. With spatulate excavate pedicellarie; abactinal plates of rays larger than those of disk
but transition not abrupt; abactinal plates flat, bordered by single row of granules, and
with a few granules and a pedicellaria on surface; otherwise as in Circeaster.

Lydiaster Koehler.

a®. Body covered by a membrane, which usually obscures the outlines of underlying plates and
which may have a smooth surface (the granules when present being beneath the skin) or may
be closely beset with small granules (as Stellaster). :

6'. Actinal intermediate areas small; abactinal plates hexagonal or rounded, each with a short
obtuse spinelet or tubercle; adambulacral armature a transverse series of three large obtuse
spinelets; pedicellarie if present bivalved..............-.-----.------ Chitonaster Sladen.

6?. Actinal intermediate areas large; abactinal plates polygonal, sometimes somewhat stellate; no
spines on abactinal surface, rarely a few short tubercles; adambulacral armature a furrow
comb, and on the actinal surface either granules or an enlarged flattened spine; slit-like

bivalved pedicellariz or two-jawed pincer-shaped pedicellariz.

c!. Membrane superficially smooth; granules, if present, beneath the membrane.

d'. Abactinal membrane very thin, without granules; no secondary abactinal plates in the
radial areas; no marginal granules or spinelets, nor actinal intermediate granules

Ogmaster v. Martens.

d, Abactinal membrane fairly thick and obscuring (unless dried) the outlines of underlying
plates; secondary small abactinal intermediate plates on either side of the median
radial series, sometimes more extensively distributed over radial regions.¢

e!, Superomarginals without conspicuous granules, spinelets, or pedicellarie. Actinal inter-
radial areas with slit-like bivalved pedicellariz in addition to hemispherical granules.
Anthenoides Perrier.
e?. Superomarginals with granules or a few short spinelets or only a few small papilliform
pedicellarie; no slit-like bivalved pedicellarie, but actinal intermediate area with

or without papilliform pincer-shaped pedicellariz.
Leptogonaster Sladen (incl. Antheniaster Verrill).

c?. Membrane closely beset with very small crowded granules. Abactinal plates polygonal or
substellate; inferomarginal plates with a flattened mobile spine; bivalved pedicellarie
flush with surface, frequently several papule to an area.............. Stellaster Gray.

III. Abactinal plates circular to substellate, separated by numerous secondary plates which are either
independent or else join the primary plates forming a close reticulated skeleton; in the latter
case, when viewed internally, the secondary plates appear as short rather irregular ossicles
passing between the large primary plates, so as to form a sort of heavy reticulum; but super-
ficially they appear as ordinary plates of smaller size; abactinal skeleton not overlaid by a thick
skin; a rather thin pulpy membrane sometimes present, partially obscuring the plates; pedi-
cellaris typically large, bivalved or with spatulate wide or narrow jaws. Adambulacral spines
of both series heavy, comparatively few, those of actinal surface of plate always few and heavy;

Usually robuseMArpying) apines Presenters e oso 57 cceeeccccecueeccusesccns<- Hippasteriinz.

@ Often lacking in very young specimens.

174 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

a, Abactinal plates entirely covered by large flat circular, quadrate, oval, elliptical, triangular, and
polygonal granules, whose edges are more or less free; secondary abactinal plates not forming
connecting pieces between primary plates; some of the adambulacral plates with furrow series
replaced by a large bivalved pedicellaria............--..----- Cryptopeltaster Fisher, p. 237.

a®, Abactinal plates not entirely covered by large flat granules whose edges are more or less free;
secondary plates forming connecting pieces between the primary plates; no large bivalved
pedicellariz on furrow margins.

b!. Granules of body sheathed in a thick pulpy membrane and furrow spines three to six in number
with remarkably thin, compressed, expanded tips.......-..----+-------- Evoplosoma Fisher.
}?, Granules not sheathed and hidden by a thick pulpy membrane; granules typically forming a
single series around the plate; furrow spines one to three in number, not flaring or chisel-
Inke atitip sss oc sos jesse cle cele oisicleieleiaia)-iee apie ainiami= nie oleic aie Hippasteria Gray, p. 223.
IV. Abactinal skeleton stellate reticulate, overlaid by a thick skin obscuring the outlines of the plates;
marginal plates heavy, beset with large granules, the inferomarginals with large bivalved pedicel-
lari also; actinal plates with coarse granules and large pedicellarie; adambulacral armature,
a few (about five) relatively short, slender, furrow spines, and one or more short heavy actinals;
all pedicellarise bivalved eisai c1= 5 ors ereisiefecteeiiesaei- ee seer wi -ie reeeee eeeeierie e eeeeee Antheneine.
a. Abactinal membrane beset with bivalved pedicellarize and scattered granules; superomarginals,
also, with large pedicellarie; abactinal plates stellate, the secondary plates connecting the lobes
to forma regular reticulated skeleton << 25 =~ <jajrtn ican inimeieisintoiots =e 'eiefoleterei-inie inte Anthenea Gray.
a, Abactinal surface without pedicellarise, with a few hemispherical tubercles along radial line,
and large hemispherical granules near marginal plates; abactinal plates relatively small, irreg-

ular, reticulate; two actinal adambulacral spines; actinal plates very distinct.
Pseudoreaster Verrill.

NOTE ON PEDICELLARIXE.

Several sorts of pedicellarize have been mentioned in the foregoing key. To
avoid confusion, these will be explained briefly. All the pedicellariz occurring in
the Phanerozonia are of the sessile type—that is, the jaws are not attached to a
special basal piece, but spring directly from the plate or from a specialized depression
or foramen in the plate.

A. Incipient and pseudo-pedicellarie. The jaws do not spring from a specialized depression or
foramen.

1. The spinelets stand on separate plates and meet over the suture between the plates.

la. Spinelets form simple fascioles; fasciolar.

1b. Spinelets are more specialized and in definite areas, or are more definitely coordi-
nated into a special apparatus, as two opposed combs; pectinate.

lc. The spinelets of a series fuse or widen, and thus form valves; valvular.

2. The spinelets stand usually on one plate, forming a more or less conical group; not found
in the Goniasteridx; fascicular or fasciculate.

B. Alveolar or tong-shaped pedicellarie. The jaws spring from a specialized pit or foramen, and
vary in size and proportions.

1. The jaws have no specialized depressions into which they fit when opened; the jaws
are higher than wide and often spatulate; foraminate, also spoken of as two-jawed
spatulate, or as spatulate.

2. The jaws are low and wide, with no specialized depression into which they fit when
opened. The pedicellarie are mere slits, or resemble a miniature bivalved shell. This
form intergrades with the simple foraminate type of various shapes. It is found in a
typical form in Hippasteria, Cryptopeltaster, Gilbertaster, Anthenea, and other genera;
bivalved. E

3. The jaws when opened fit into a specialized depression. This type is best developed in
the Linckiide, but occurs in Calliaster, Astroceramus, Lydiaster; excavate or sugar-tongs.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 175

Subfamily MIMASTERINUJZ® Sladen, 1889.

Mimasterine SuapEN, Challenger Asteroidea, 1889, p. 331.—Prrrier, 1894, p. 252.—VERRILL,
1899, p. 200.

Genus GEPHYREASTER Fisher.

Gephyreaster FisHer, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p.171. Type, Mimaster

swifti Fisher.

Diagnosis.—Related to Mimaster Sladen, but differs in having stout tabulate
paxille with strongly stellate bases by which plates overlap; in character of actinal
intermediate plates which, instead of bearing small tufts of spines forming spaced
paxille, are densely covered with spinelets; in the armature of the mouth plates,
which have a peculiar angular marginal series situated between peristome and
superficies of plate on the inner end; superambulacral plates present. Marginal
plates, adambulacral plates, and their armature, similar to those of Mimaster, the
first without enlarged spinules.

Remarks.—This genus is erected for the species which I called Mimaster swifti
in 1905. Since then I have received two additional specimens, and these have
convinced me that I was in error in placing the species in that genus. The char-
acter of the low tabulate paxille resembles Pseudarchaster, from which, however,
the new genus differs in the form of the adambulacral armature and marginal
plates, and to a less extent in the character of the actinal intermediate plates.
The large sucking disks of the tube feet also ally Gephyreaster swifti to Mimaster
and Pseudarchaster rather than to any genus near Plutonaster.. It is probable that
Mimaster has superambulacral plates and that-Sladen overlooked them, as is very
easy to do in forms where they are feebly developed. The median tooth pointed over
the actinostome recalls some species of Pseudarchaster. My own opinion is that
the present genus is intermediate in many respects between Mimaster and Pseud-
archaster, and I would therefore associate the three genera more or less closely in
the system. The resemblance of true Mimaster to Leptychaster is also great, but
the difference in the tube feet is a more important barrier than is commonly
supposed. In passing, it might be well to call attention to the discrepancy in
this character between MM. tizardi and M. cognatus. The latter has conical tube
feet with a little button (as in Astropectinide) at the tip. MJ. cognatus is really
not a Mimaster but in all probability a Leptychaster. The present genus is also
more remotely allied to the Odontasteridz.

GEPHYREASTER SWIFTI (Fisher).

Pl. 30, figs. 1-4; pl. 57, figs. 2, 2a-c; pl. 89, figs. 5, 5a.
Mimaster swifti Fisher, Bull. Bur. Fisheries for 1904, vol. 24, 1908, p. 301.

Diagnosis.—Rays five. R=114 mm.; r=43 mm.; R=2:65 r. Breadth of
ray at base between first and second superomarginals 50 mm. A large specimen.
R=140 mm.; r=58 mm.; R=2.5 r. General form large, robust, flattened; rays
broad at base, tapering evenly or else slightly arcuately, to a blunt extremity;
interbrachial ares wide, rounded; abactinal surface slightly inflated, sunken in
actinal interradial areas. Marginal plates conspicuous, without specialized spines,

176 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

but resembling very large paxille; furrows between plates usually conspicuous.
Abactinal paxille with stellate imbricating plates arranged in longitudinal series
on rays, bearing low stout tabula crowned with numerous granuliform spinelets.
Papule one, two, or three to each area; six areas about a plate; absent from end
of ray. Adambulacral plates with a dense armature of ten to twelve robust
subeylindrical spines or spinules, decreasing in size away from furrow; two or three
on furrow margin, and the rest in subparallel rows of two or three. Adambulacral and
mouth spines are the largest ones on animal. Mouth plates prominent with angular
marginal series situated on side of plate between true margin and peristome and
with numerous stout coarse spines on the actinal surface. Usually a single large
tooth at inner angle of combined plates, with several slightly smaller accessory
spines. Actinal interradial areas large, but narrow on rays; plates arranged in
interradial or transverse rows, densely covered with spinelets. Superambulacral
plates present.

Description.—Abactinal surface varies in the degree of inflation, usually
arched on rays and disk. It is covered with fairly large and compact, rather
closely placed, tabulate, low paxille, which are arranged in rather indefinite trans-
verse oblique rows at the sides on the proximal portion of rays, but are without
order on distal half and on central portion of disk. The paxille also form not very
definite longitudinal rows on basal portion of rays, the mid-radial series being
largest and fairly definite as far as middle of ray. On either side of this the paxille
decrease in size toward margin. These paxille are not of the high slender sort
characteristic of the Astropectinidz, but are low, with broad flat tabula (like those
of Mediaster) crowned with robust clavate or acorn-shaped granuliform spine-
lets or elongate granules. These spinelets, which vary greatly in shape, from
roundish polygonal with a broad dome-shaped or nearly truncate tip to thimble-
shaped or even acutely ovoid, form a hexagonal, roundish, or quadrate very slightly
convex group of twenty-five to forty on large plates and fifteen to twenty-five on
the smaller lateral ones according to size of specimen (largest with R=140 mm.).
Those in center are usually shorter than granules about the edge, sometimes smaller,
sometimes slightly larger. The photographic figure will give the best idea. In
addition to the above, one or two irregular peripheral series of very much smaller,
and considerably slenderer papilliform, lanceolate spinelets are placed lower down
on the pedicel, and appressed closely to bases of outer robust spinelets, so that
they are not usually superficially visible. Superficially the papule are easily seen
between the paxille, two or three to an area in the two largest individuals, usually
only one in a specimen with R=74 mm.; six areas about each plate.

Abactinal plates viewed from inner or ceelomic side are seen to be strongly
stellate, with usually six, but sometimes only five or four lobes. By these lobes
the plates overlap one another after the manner shown by the drawing. They
are arranged in regular longitudinal series parallel with the median radial. The
manner in which either adradial series overlaps (as seen from the underside) the
radial series is shown by drawing. Notice that this is different from the overlapping
of the other parallel series. At the proximal end of the regular radial series is an
easily distinguishable primary radial plate. Between this and center of disk
the plates are smaller, less uniform in size, and irregular in disposition. On the ray

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. ee

the plates decrease very gradually toward margin and distally. Papule appear
to be all over disk, barring the narrow line of insertion of interradial septum. They
are absent from the terminal fifth of ray. Here the plates are less markedly stellate.
The arrangement of papule is described in preceding paragraph.

Marginal plates conspicuous, but partaking of the nature of large paxille,
entirely without enlarged spines; a well-defined, narrow, naked groove between
superomarginals and abactinal paxille. Both series with special raised ridges
crowned with coarse granules which increase in size toward center of ridge, where
they are quite heavy and similar to very much enlarged spinelets of the paxillx,
though more tubercular and pointed. On inferomarginals the spinelets are still
thicker and heavier, and increase in size toward actinal end of plate. Supero-
marginals, forty to forty-three in number from interradial line to extremity of ray,
are much wider than long at base of ray; less so on outer part; with curved upper
and lower margins; the raised ridges strongly tumid, and when denuded are about
as thick as the adjacent intervening furrows, which are wider in the type than in
the other two specimens. A conspicuous longitudinal channel separates the plates
of the two series These furrows are not lined with fine spinelets as is usually the
ease with typical fasciolar channels. The superomarginals encroach upon the
abactinal area, forming usually a well-defined border. The spinelets vary more
or less but are usually clavate and much larger on lower half of plate, where two
to four series on the median line have abrupt conical tips. Terminal plates small
in largest specimen, but relatively larger in smallest; about as wide as long, broadest
distally; sometimes ovoid, at others roundish; granulose; wedged between last two
or three superomarginals.

At base of ray the inferomarginal plates do not correspond always exactly
with superomarginals, but in some cases may even alternate. The plates are tumid
and form a very even rounded border to actinal area, but are narrower than the
corresponding superomarginals (beyond interbrachial arc), hence are nearly to
quite quadrate beyond middle of ray; (quadrate in largest specimen, wider than long
in other two including type). The granules as a whole are more robust than those
of superomarginals, but are similar in character and increase in size toward lower
part of plate, being much larger along the median transverse region than on periphery
of ridge. The marginal plates are, in appearance, what they are morphologically—
very much enlarged paxille.

Adambulacral plates, wider than long and rather closely placed, so that the
peculiarly characteristic armature forms a dense mass of spines along margins
of furrows. On each plate are ten to twelve robust, subcylindrical, untapered,
occasionally slightly compressed, truncate or round-tipped spines which decrease
in size as they recede from the furrow and are arranged either in five longitudinal
series of two, four series of three, or may be without regular order on outer half
of plate. Furrow series is commonly oblique and composed of three instead of
two spines; or one may stand on center of margin and two just behind it. On
outer end of plate five or six robust much smaller spinelets (even granules some-
times) form a group about the spines along the border of plate. These outermost
spinelets are sometimes hard to distinguish from those on adjacent actinal inter-

57444°—Bull. 76—11——12

178 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

mediate plates. The spinulation is so dense all along the adambulacral plates
that it is difficult to distinguish the individual plates.

Mouth plates comparatively large, rather prominent actinally, this appear-
ance being accentuated by a depression in the interradial area at their outer end.
General surface is covered with numerous, robust, prismatic, subquadrate, trun-
cate, short irregular spinules or spines, which increase in size and become more
compressed (broad side to furrow) toward the free margin of plate. The true mar-
ginal series begins near outer (aboral) end, high in the furrow, next to first ambulacral
plate as rather inconspicuous flattened lanceolate or truncate spinelets, which
rapidly increase in size toward inner angle. These form a group or an angular series
(the apex being next to peristome), one part of which curves upward (as viewed
from actinal surface) along the furrow margin of plate, while the other, closely
parallel for a way, proceeds upward to the inner or dental angle of the plate. At
this inner or dental angle is a large median spine directed over actinostome, as
in Pseudarchaster. This spine is either flattened and broad at tip (type), flattened
without being conspicuously widened (station 4784), or truncate and somewhat
pestle-shaped (station 2879). On either side of the median tooth is a similar spine,
and those on actinal surface seem to grade into’ these dental spines. The median
spine is not always present. On the outer part of the plate of the largest specimen
the small irregular but very robust spinelets have a truncate concave tip. There
is so much variation in the minor details of the armature in the three specimens
that it seems useless to describe in detail the actinal surface of these plates. The
illustration (pl. 57, fig. 2) will give a better idea.

Actinal interradial areas large and paved with superficially elliptical or oblong
plates arranged in series from adambulacrals to inferomarginals. There is one of
these series to each adambulacral (as far as the plates.extend) with occasionally an
extra series. The plates of a series imbricate, the outer end underlying the inner
end of the next exterior plate, but although the independent series are close together,
the plates do not appear to overlap along the longitudinal axis of ray. On the
rays the intermediate areas are narrow, not broad as in M. tizardi. One series
of plates (longitudinal) extends nearly to tip of ray; the second series extends
only half the length of ray measured from first marginal plate; a third series about
one-fourth or to eleventh inferomarginal. Between this point and interradial
line the number of plates increases very rapidly. Each plate is slightly convex
and is covered with coarse spinelets, which are largest in center of plate, clavate,
prismatic, quadrate, or pinched, with subtruncate or rounded tips. The peripheral
spinelets are smaller truncate, pinched or papilliform, and bend over the shallow
channels between the transverse series of plates. These are similar to the ‘‘post-
adambulacral fascioles” of Pseudarchaster, but the grooves lead all the way to the
marginals. A few of the intermediate plates on largest specimen have four central
spinelets equal and closely grouped to resemble a pedicellaria; I doubt its being one.

Madreporic body inconspicuous about midway between center of disk and
inner edge of marginal plates or a little nearer center than margin. It is small
circular convex, partly overhung by paxille, a large primary basal paxilla occurring

on the adcentral side. Striations fine, ridges coarse centrifugal, branching and more
or less sinuous,

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 179

Anatomical notes.—Superambulacral plates present though small. They are
fitted into the angle between lower end of ambulacral and actinal intermediate
plates and are very inconspicuous unless treated with caustic potash. They are
absent from the first ambulacral and at the end of the ray for a considerable number
of plates. The longest superambulacrals are less than one-third height of exposed
surface of corresponding ambulacral. Gonads in a thick much-branched tuft on
either side of interradial septum. Anus present, small and inconspicuous. Intesti-
nal cecum large, divided into two digitate halves, one with four the other with
six palmately arranged lobes. Strong retractors of stomach present. Inter-
radial septa single, thick and apparently muscular, but containing no calcareous
deposits. Tube-feet large in two rows, with broad sucking disk; no deposits.

Type.—Cat. No. 22332, U.S.N.M.

Type-locality.— Albatross station 4253, Stephens Passage, Alaska, in 131 to
188 fathoms; rock and broken shells. (1903 coll.)

Distribution.—From the end of the Aleutian Chain to Washington, 34 to 188
fathoms; on rocks.

Specimens examined.—Three; besides the type, one from station 2879, off
Washington, 34 fathoms, rocks; (U.S. Nat. Mus. coll.); one from 4784, near Attu
Island, Aleutian Chain, 135 fathoms. (Albatross, 1906 coll.)

Subfamily PSEUDARCHASTERINA<AE®L Sladen 1889 (extended).

Pseudarchasterine SuaDEN, Challenger Asteroidea, 1889, p. 109.—Verrim, Trans. Conn. Acad.,
vol. 10, 1899, p. 187 (extended).

Astrogoniine PERRIER (part), Expéd. sci. du Travailleur et du Talisman, 1894, pp. 337, 338.

Pseudarchasteride FisHer, Bull. Bur. Fisheries for 1904, vol. 24, 1905, p. 303; idem, 1903, pt. 3,
1906, p. 1036.

Genus PSEUDARCHASTER Sladen.

Pseudarchaster SuaDEN, Narr. Challenger Exp., vol. 1, 1885, p. 617; Challenger Asteroidea, 1889,
p. 109. Type, Ps. discus ¢—VerRiLL, Trans. Conn. Acad., vol. 10, 1899, p. 189.

Astrogonium (part) Perrier (not of Miiller and Troschel), Expéd. sci. du Travailleur et du
Talisman, 1894, p. 338.

Diagnosis.—Abactinal plates paxilliform, usually more or less stellate, arranged
in radial rows, the median largest; papules usually six about a plate, commonly absent
from distal half of ray. Marginal plates thick, moderately large, paired, with fas-
ciolated grooves between them. Superomarginals usually separated throughout
length of ray, one to three series of paxille attaining terminal plate. They are
covered with close granules or small, crowded appressed spinelets, the infero-
marginals sometimes with enlarged appressed spinules in a transverse series or group.
Adambulacral plates usually broad with convex or angular furrow margin, and few
furrow spinelets in palmate series. Actinal intermediate plates numerous, arranged
in rows running from adambulacral to inferomarginals often also forming parallel
longitudinal series; these plates either simply granulous, or with slender spinelets,
often with an enlarged spinule on many plates. Mouth plates large and prominent,

@ No species is mentioned in the original description. Strictly, the genus had no status till 1889;
Ps. discus is the first species.

180 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

frequently with an odd median spine at inner angle of combined plates. Pedicellari
sometimes present, when they are either pectinate (formed of two opposing combs of
spinelets over a transverse suture between actinal intermediate or adambulacral
plates) or are bivalved and evidently developed from the foregoing sort. Incipient
pedicellarize or postadambulacral fascioles are present on actinal interradial areas
and readily develop into pectinate pedicellarize. Small pincer-like pedicellarie
present rarely. Superambulacral plates present, often but feebly developed.
Tube feet with a well developed sucking disk.

KEY TO THE SPECIES OF PSEUDARCHASTER HEREIN DESCRIBED.

a‘. No trace of bivalved pedicellarie on first row of actinal intermediate plates or of pectinate pedicel-
lari with conspicuously widened spinelets; superambulacral plates not well developed; actinal
intermediate and marginal plates covered with short swollen spinelets or granules.

b!. Paxille crowded; granulation of marginal plates close, the inferomarginals with low hexagonal or
squamiform granules; abactinal paxillar area narrow on rays; actinal spinulation coarser and
more granuliform; rays longer, disk smaller; postadambulacral fascioles........ parelii, p. 180.

b2. Paxille usually well spaced with prominent papulz between; superomarginals not prominent,
with spaced granules, a few conspicuously enlarged at outer end of plate; inferomarginals with
prominent enlarged spinules and spaced papilliform spinelets; rays broad and usually short;
actinal spinulation more delicate and spaced; pectinate pedicellarize sometimes present—simply

more fully developed postadambulacral fascioles..............--.--.--+.-+--- pusillus, p. 187.
a®, Bivalved or bivalve-pectinate pedicellariz; actinal spinelets rather long and delicate on inter-
mediate plates; superambulacral plates well developed................--..---- dissonus, p. 192.

PSEUDARCHASTER PARELII (Diiben and Koren).
Pl. 31, figs. 1, 2; pl. 33, fig. 3; pl. 57, fig. 5.

Astropecten parelii DipeEN and Koren, Kongl. Svensk. Vet. Akad. Handl., 1844, 1846, p. 247,
pl. 7, figs. 14-17.—M. Sars, Nyt Mag. for Naturv., vol. 6, Heft 2, 1850, p. 161.

Archaster parelii M. Sars, Ofversigt af Norges Echinodermer, 1861, p. 35, pl. 3, figs. 1, 2—VERRILL,
Amer. Journ. Sci., vol. 7, 1874, p. 500; vol. 23, 1882, p. 140; Rep. U. S. Comm. Fish and
Fisheries, vol. 11, 1884, p. 543.

Archaster parelii, var. longobrachialis DANIELSSEN and Koren, Nyt Mag. for Naturv., vol. 23,
1876, p. 17; The Norwegian North Atlantic Exp., Asteroidea, 1884, p. 88.

Plutonaster (Tethyaster) parelii StapEN, Challenger Asteroidea, 1889, pp. 83, 102, 122, 722.—
Sturrer, Bijdr. tot de Dierk., Afl. 17, 1895, p.51.—StapeEn, Trans. Roy. Irish Acad., vol. 31,
pt. 3, 1897, p. 78.

Plutonaster parelii Bett, Catalogue of the British Echinoderms in the British Museum, 1892,
p. 63.—Norman, Ann.and Mag. Nat. Hist. (6), vol. 12, 1893, p.346.—Grina, Bergens Museums
Aarbog, 1894-95, No. 12, 1896, pp. 5, 12; Stavanger Museums Aarsberentning for 1896, 1897,
p. 37.—Lupwie, Fauna Arctica (Seesterne), vol. 1, 1900, p. 449 (full synonymy of ‘‘parelii
restricted,’’ to 1897).

Pseudarchaster intermedius Stapen, Challenger Asteroidea, 1889, p. 115, pl. 19, figs. 3, 4; pl. 42,
figs. 5, 6.—VeErru1, Proc. U. 8. Nat. Mus., vol. 17, 1894, p. 249; Amer. Journ. Sci., vol. 49,
1895, p. 195; Trans. Conn. Acad., vol. 10, 1899, p. 190, pl. 25, figs. 1, la-b.

Pseudarchaster tessellatus, var. arcticus Sturrer, Bijdr. tot de Dierk., Afl. 17, 1895, p. 51 (according
to Ludwig).

Astrogonium parelii Kauer, Bull. de l'Institut océanographique, No. 99, April, 1907, p. 31.

Diagnosis.—Rays five. R=104 mm.;* r=37mm.; R=2.8r. Breadth of ray
at base, 42 mm. Rays well developed, abruptly and arcuately tapering at base,

2An unusually large specimen for this species, station 3225.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER,. 181

then very gradually to the blunt extremity; interbrachial ares wide and rounded;
abactinal area subplane, only a trifle inflated on center of disk. Abactinal paxille
small, crowded and regular, about two opposite each marginal plate, and with
five to seven polygonal or subprismatic granules surrounded by twelve to fifteen
slenderer papilliform spinelets on periphery of tabulum. Abactinal plates with
five or six prominent lobes. Marginal plates broad with close-set, flat-topped hex-
agonal granules. Inferomarginals with a transverse row of small appressed,
squamiform, pointed spinules. Adambulacral plates with five or six strongly com-
pressed furrow spinelets, and on actinal surface one or two enlarged spinules sur-
rounded by numerous shorter granuliform spinelets to the number of fifteen to
twenty-two, all thick, heavy, and membrane-invested. Mouth plates with median
tooth. Actinal interradial areas large; plates covered with rather crowded, swollen,
polygonal, papilliform unequal granules, those in center more robust, clavate, with
flaring tips. Many plates with a central enlarged spinule. Fasciolar channels or
pectinate pedicellariz between plates adjacent to adambulacrals, these continued
toward margin but less conspicuously. Spinelets forming roof of fascioles slender.
Superambulacral plates present.

Description.—Abactinal paxille small and crowded fairly regular, largest on
proximal radial regions, very crowded and small at ends of rays where only the median
radial series attains the terminal plates but the two adradial nearly reach it; one
and one-half to two paxille correspond to each marginal plate. The larger paxille
have five to seven polygonal or subprismatic, robust, truncate granules, heavier at
tip than at base, and occupying surface of tabulum, while on the periphery are about
twelve to fifteen much slenderer papilliform or subprismatic spinelets. The spine-
lets may be very compactly placed or form a more or less open group. Along the
border of the area the paxille are compressed, the tabulum being elliptical and
crowned with two rows of eight to ten granules in all.

Abactinal plates with five or six distinct lobes by which the plates touch or
overlap; toward margin, plates often very irregularly lobed, or without lobes.
They are arranged in series parallel with median radial. Papule, one to an area,
six about each plate; but absent from terminal half of ray, where the plates are
without lobes, being irregularly hexagonal or oval.

Marginal plates broad and short, encroaching conspicuously upon both areas.
Superomarginals, fifty-eight in number from interradial line to extremity of ray,
vary in width, and are wider in interbrachial are than elsewhere (8.5 to 9 mm.
wide in large examples). They form an even bevel, more or less arched on outer
part of ray and are covered with regular, close hexagonal granules in five to seven
transverse rows which are coarser at the outer (lateral) end of the plate. The
marginal granules are smaller and form very regular series. Granulation has
appearance of being very regular, compact, and smooth. Grooves between plates
invisible from exterior, and probably not functioning as fascioles. Terminal plate
medium-sized, obovoid, covered with granules.

Inferomarginals correspond to superomarginals in position, though there is
usually one additional plate at tip of ray, and they are a trifle wider. Covered
with coarse hexagonal granulation, which increases rapidly in coarseness toward
margin of ray. In some specimens the granules are slightly squamiform. In the

182 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

interbrachial are the plates bear a median transverse series of four to six flattened
lanceolate appressed spinules, which are gradually reduced in size and number
along ray, being frequently absent from the last few plates, and only one or two
beyond the middle.

Adambulacral plates with an angular furrow margin bearing a palmate series
of five or six more or less compressed spinelets, the median (or adoral admedian)
the longest and most compressed; tips rounded; lateral spinelets often with flat
side uppermost. The furrow series is continued along adoral and aboral margins
of plate in three or four spaced, stout, much smaller, papilliform spinelets. On
actinal surface one, two, or occasionally three, enlarged very robust bluntly pointed
spinules stand in a transverse oblique, or longitudinal series (only one spinule in
medium-sized and small specimens). Between them and furrow series is a semi-
circular row of three or four shorter, blunt, stout, papilliform spinelets, or some-
times very strongly flattened spinelets; these sometimes absent; on outer part of
plate are several smaller three- or four-sided unequal clavate spinlets. Exclusive
of furrow series there are about fifteen to twenty-two spinules and spinelets to
each plate, the outermost very irregular in distribution and on distal part of ray
showing a tendency to group themselves about the two or three larger spinules.

Mouth plates prominent actinally, with a bristling armature of robust, short,
untapered blunt spines disposed in a marginal and two actinal series. Of the latter
one stands on the border of the median suture, while the second, an intermediate
shorter irregular series, is located between it and a continuation of the marginal
series on the edge adjacent to first adambulacral. These spines are subquadrate,
subterete, occasionally spatulate, and are longer on the inner than on outer end of
plates. They are also variable in number according to age and locality. The true
marginal series consists of about seven robust spines in addition to a large median
unpaired spine at the inner angle of the combined plates. The three adjacent
to inner angle are usually graduated in size, then the next four are stouter and
longer. This is the plan of the furrow series; the spine counts vary.

Actinal interradial areas large, the intermediate plates extending to the tenth
to seventeenth inferomarginal, or one-fourth to a little over one-third length of
ray measured along side from interradius; plates arranged in rows running from
adambulacrals to inferomarginals. From inner side these plates are oval and
imbricate with all the surrounding plates. Plates armed with unequal, swollen,
more or less crowded granules, those in center robust and clavate, often with slightly
flaring tips bent outward, the peripheral smaller, round-tipped, occasionally sub-
prismatic, very unequal and irregular, and radiating over narrow shallow grooves
between the plates. These channels lead in an irregular course from the infero-
marginal fascioles to those between the adambulacral plates. In some specimens
they are more conspicuous than in others, and all are more conspicuous adjacent to
adambulacrals where they are often roofed by pectinate pedicellariw. Many of
the plates bear a central enlarged pointed spinule directed toward margin. The
number of plates bearing such spinules is variable.

Madreporic body small, situated one-third distance from center to margin;
striations very irregular, the ridges between stris coarse.

Anatomical notes—Superambulacral plates present, extending along ray as
far as do the intermediate plates, and absent from first ambulacral ossicle; they

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 183

are rather small. Gonads in a single tuft on either side of interradial septum,
which is membranous. Anus present. Intestinal eoecum large with ten or eleven
radiating, slightly branched divisions. Tube feet with well-developed sucking
disks; no deposits.

Variations—The principal variations have been noticed in the above description.
To recapitulate, they concern chiefly the following characters. Rays, which vary
in breadth and length, being a little shorter in Japanese specimens, but vary also
in Alaskan examples. Paxillee: more compact in some individuals, with a slight
difference in width of paxillar area on rays, due to the variable width of supero-
marginals. Superomarginals: variable in width and number, being apparently
slightly broader in Japanese specimens (but occasional examples, stations 3225
and 3258, from Alaska, also with broad plates); in number varying from thirty-
three (Japanese) to forty-three (station 3258) in examples of same size (R, 70 mm.),
the Japanese specimens consequently having longer plates. Inferomarginals: varia-
ble in width as superomarginals; covering either very compact and polygonal or
more open and squamiform. Variations in adambulacral armature, mouth plates
and actinal intermediate plates, sufficiently treated above.

Young.—The type described by Sladen is a small specimen. The smallest
specimen in the collection from station 4792 has the following dimensions: R, 19
mm.; r,8 mm. Superomarginals, twenty. The actinal interradial areas are small
and lack enlarged spinules, as do the adambulacral plates on the proximal half of R.
Terminal plate large, subglobose; proximal superomarginals with small bare spot;
spinules of inferomarginals only a trifle enlarged beyond granules.

Type-locality— Christiansund”’ (Norway), 30 fathoms.

Distribution —Off the eastern coast of the United States and Canada (Jat.
44° 26’ to 37° 59’ N.) in 85 to 1,608 fathoms [Verrill, 1899]; off Norway (Chris-
tiania Fjord to Finmark, lat. 58° to72° N.) eastward to the Murman coast; west-
ward to the Shetland Islands, and Ireland [Ludwig, 1900]; Bering Sea (Bering
Island; Pribilof Islands) extending at least as far as Kadiak Island in the eastern
north Pacific, and south along the coast of Asia to the Sea of Japan, 70 to 351
fathoms.

Specimens examined.—Twenty-seven, from the following stations:

Specimens of Pseudarchaster parelii examined.

Station. | Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
$225 5 5-5 Near Unimak Island, Aleutians. .... 85 DISCKI BANU a onte caer ane ccehscra€ 1 U.S. Nat. Mus.
81 gray sand, gravel. | 11 Do.
70 black sand, gravel 2 Do.
3330. ..... | North of Unalaska Island, Aleutians.) 351 MUM ch eae es aN aabas teams fast te cae 1 Do.
3487......| Bering Sea, west of Pribilof Islands -- 81 1 Do.
3490. .....| West of Pribilof Islands............. TS Janne GO ewes nancwseu ns one csdanccenens 1 Do.
3548...... | North of Unimak Island ..-........- 91 DIACK SANG caeeuc caus Coca scene Race 1 Do.
3606. ..... Bering Sea, north of Unalaska Island. 87 green mud, fine sand............... 1 Do.
MITA etal mm Uyak Bay, Kadiak Island........... 66-67 GIS UNG sao Pc tuee seep rete 2 Albatross, 1903.
4201...... Shelikof Strait, Alaska.............. 65-48 | blue mud, sand, gravel 1 Do.
4202... .... anes Mosepanenavckoessntasacccmusees 109-94 | blue mud, fine sand 1 Do.
4784. ....| Near Attu Island, Aleutfans......... 135 COATSE DEDDIES. <5 eveasecueccrsceces 1 Albatross, 1906.
R1OA ws oo Near Bering Island.................. | 72 MOU Dian macccs svuccusUyteyeuen cas 1 Do.
4855..... | Matsushima, Sea of Japan..........- | 70-89 IDEN INGE oon op cuevenwcemeek eat 2 Do.

184 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Remarks.—This species is variable in the Atlantic, and the Pacific specimens
are certainly no exception to the rule. Greater differences are observable between
specimens from station 3258 than between an example from off Newport, Rhode
Island, and two from Norway, and a slightly larger one from 3257, near Unimak,
Aleutian Islands. The specimens from Kadiak Island and Shelikof Strait are frankly
not typical, but appear to be intergrades with alascensis. Japanese specimens
are variable and not quite typical. The specimens from Matsushima, in proportion
to length of ray, have fewer and hence longer marginals, which are wide. All the
granulation is low, coarse, and very compact, and the adambulacral spinelets are very
heavy in proportion to length, those on outer half of plate being granuliform. The
wide marginals are duplicated in an Alaskan specimen from station 3258, there
being also a more typical specimen from the same dredge haul. The enlarged
actinal intermediate spinelets on both Alaskan and Japanese specimens (of which I
have examined a number) are quite variable, sometimes being absent, while equal-
sized examples have them. Postadambulacral fascioles are not always evident.

It appears as if this species, spreading south along the Alaskan coast, had
changed into a form with narrower superomarginals, having less granuliform arma-
ture on the actinal surface, and less compactly placed granules on the abactinals.
Along the Asiatic side the development has been toward fewer and broader supero-
marginals, with more compact abactinal granules, and an accentuation of the
granuliform character of the actinal armature. The Japanese form has departed
less from the type than has the British Columbian.

This species is the north Pacific and Atlantic representative of Ps. discus Sladen
(from Messier Channel, between Chile and Wellington Island), to which it is closely
related. It may be that Ps. pulcher Ludwig (Galapagos to southwest of Acapulco),
founded on very small specimens, is the connecting link between the two forms.

A few words concerning the name adopted may be in order. In the Museum
of Comparative Zodlogy are three specimens of “ Astropecten parelii”’ from Norway,
presented by Professor Sars in 1852. These agree very well with the original
description and figures of Diiben and Koren. ‘There are no differences of importance
between these specimens and an example of Pseudarchaster intermedius from off
Rhode Island. Similarly, the Norwegian specimens belong to the same species as
the Alaskan. There is far more difference among the various Alaskan examples
than exists between these on the one hand and either the Norwegian or Rhode
Island specimens on the other. The width of the paxillar area on the arm varies
in this species, even in specimens from the same locality. It is not surprising,
therefore, that the area is narrower in the specimens from Norway, just mentioned,
than in the figure of Diiben and Koren.* Diiben and Koren figure also the supero-
marginal plates and paxille, the inferomarginals and adambulacral plates, and
fortunately the tube feet. These have strong sucking disks. The figure is enough
to place the animal in Pseudarchaster.

That Astropecten parelii and Pseudarchaster intermedius are one and the same
(somewhat variable) species I have not the slightest doubt. I also doubt if Pseud-
archaster fallax is anything more than a variety.

a Plate 7, fig. 14.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 185

Sladen placed parelii in Tethyaster, a subgenus of Plutonaster, while Bell,
Norman, Grieg, and Ludwig relegated it to Plutonaster. Such a course is untenable,
because Plutonaster belongs to a different family altogether, and has pointed tube
feet, never sucking disks. This character alone is enough to exclude parelii from
Plutonaster. As a matter of fact, parelii is so near the type of Pseudarchaster [that
is, discus] that one is obliged to search carefully to find trenchant differences. The
following are the considerations which lead one to rank parelii in Pseudarchaster:
tube feet with sucking disks; character of the marginal plates, especially the infero-
marginals; the characteristic adambulacral plates and armature, which is totally
unlike Plutonaster; the actinal intermediate plates with spaced granules and
incipient central spinule; the armature of the mouth plates, there being an unpaired
median tooth directed over the actinostome; the presence of postadambulacral
fascioles very characteristic of Pseudarchaster and never found in Plutonaster or
allied genera.

In a specimen of parelii from the Copenhagen Museum: R=43 mm., r=14
mm.; superomarginals twenty-three; the superomarginals are considerably arched
above the abactinal paxillar area on arms; the latter is narrower than the mar-
ginal plates, there being at the middle of ray only three longitudinal rows of paxille,
and only a single row reaches terminal plate; about four and one-half to four
paxille correspond to two superomarginals; furrow spinelets five or six; one or
two actinal spinelets enlarged slightly; first row of actinal intermediate plates has
incipient fascioles;.a few of the actinal intermediate plates with slightly enlarged
spinelets.

This specimen differs from the Alaskan chiefly in having tumid superomarginals
and a narrower paxillar area on rays. There are about five transverse rows of
of granules on the superomarginals (the same as in Alaskan specimens).

PSEUDARCHASTER PARELII ALASCENSIS Fisher.
Pl. 31, figs. 3-5; pl. 33, fig. 4; pl. 40, fig. 3; pl. 57, fig. 6.
Pseudarchaster alascensis Fisher, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 303.

Diagnosis.—Very close to Ps. parelii, and differing in having shorter and
narrower superomarginals with less compact granules, less compact paxillar area,
papilliform inferomarginal spinelets, more delicate and longer actinal intermediate,
adambulacral, and mouth spinelets. The general facies of the two forms is different
enough, but the actual differences, allowing for individual variation, are not great.
R=99 mm.; r=34 mm. R=2.9 r. Breadth of ray at base 40 mm.; at middle
15 mm.

Description.—Paxille are not markedly different from those of parelii, but are
slightly spaced instead of crowded, and a trifle smaller, the central spinelets having
convex instead of flat tips. The figures will give the best idea of this difference.

The superomarginal plates are decidedly narrower than in Japanese and most
northern specimens of parelii. In these the widest plates in interbrachial are are
9.5 mm. (R=77 to 100 mm.) and in alascensis the same plates are 5.5 to 6 mm.
(R=73 to 99 mm.). In alascensis the granules are not close-set, forming a hard,
smooth mail, but are spaced and those on outer edge of plate are enlarged, with
strongly convex or conical subtubular tips. A specimen from station 4201 has
these enlarged into small subconical spinules in the interbrachial are.

186 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

The inferomarginal armature consists of slightly spaced subpapilliform slightly
flattened spinelets with a transverse series of enlarged lanceolate appressed spinules.

Adambulacral plates with five or six tapering furrow spinelets compressed as
in parelii but longer and more delicate, as are those of actinal surface. The figure
will show the differences better.

The armature of the mouth plates is composed of terete or slightly compressed,
tapering, rather slender spinelets, much longer than in parelii. The arrangement
of spinules, however, is not materially different. be:

Spinelets of actinal intermediate plates are spaced and of a swollen papilliform
shape. A number of plates bear an enlarged spinule, but these are usually scarce.
Compared with a Japanese specimen of parelii, the spinelets are longer and less
eranuliform, and decidedly more spaced; the difference is not so obvious, however,
when Alaskan specimens are used for comparison. Postadambulacral pectinate
pedicellarie present; but sometimes scarcely more than simple fascioles.

Type.—Cat. No. 22334, U.S.N.M.

Type-locality.— Albatross station 4236, near Yes Bay, Behm Canal, Alaska, in
147 to 205 fathoms, rocks, coarse sand.

Distribution.—Oregon to Queen Charlotte Sound and north, intergrading with
Ps. parelii in the region of Kadiak Island.

Specimens examined.—Nine from the following localities:

Specimens of Pseudarchaster parelii alascensis examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
| Fathoms.
2 eeesee | Queen Charlotte Sound, off Fort Rupert, Van- | 107-68 | rocks, coarse sand...... 3 Albatross, 1903.
couver Island, British Columbia. | |

4236...... Vicinity of Yes Bay Behm Canal, Alaska -.-... 147-182 | rocks, coarse sand 1 | Do.

ADGT 22-2 Stephens Passage Alaska..........--..---------- | 198 TOCKY. soeitecr eee 1 | Do.
4258.52 2 Vicinity of Funter Bay, Lynn Canal, Alaska... -| 300-3138 |mud Saceece eee 2 ‘ Do. ,

JAD eae | Off Queen Charlotte Island, British Columbia et 876 | green mud....... 1 U.S. Nat. Mus,
3788 ..-..- Off Cape Blanco, Oregon¢.........-...-.-.-.--- | 1064 | GO saeco eee 1 Do.

@ Not typical.

Remarks.—Since describing this form in 1905 I have seen considerable material
not then accessible, including specimens of Ps. parelii from the Atlantic and the
Sea of Japan. The special characters of the present race are exhibited only by
specimens from off British Columbia and southern Alaska, but examples from
Kadiak and vicinity are intermediate, though a little nearer to parelii. It is
impossible to assign them satisfactorily to either form.

That Ps. pusillus, from California, is not a dwarfed variety of this race, or of
parelii, becomes very evident when equal-sized specimens are compared. A small
specimen (R=37 mm.) from station 4201 is nearer typical parelii than the adults;
the paxillar and marginal granulation is more compact, and the adambulacral
spinelets shorter. Compared with pusillus, this specimen has longer and narrower
rays, very wide superomarginal plates, more crowded paxille (in view of their
variation in pusillus the size is of no importance), much narrower paxillar area,

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 187

less prominent inferomarginal spinules, much more compact inferomarginal and
actinal intermediate spinulation, more numerous adambulacral spinelets, and
especially lacks the prominent marginal fascioles of pusillus.

A very young specimen from station 2860, probably of this race (although in
absence of adults from the same station it is not possible to be positive) has R=7
mm., an arcuate pentagonal form, and a large naked area on each of the seven or
eight superomarginals of the ray. The adambulacral plates have three or four
furrow spinelets and the mouth plates an odd mouth spine. There are no enlarged
inferomarginal spinules as in small specimens of pusillus.

PSEUDARCHASTER PUSILLUS Fisher.
Pl. 32, figs. 1-4; pl. 33, figs. 1, 2; pl. 57, figs. 4, 4a; pl. 59, fig. 4.

Pseudarchaster pusillus FisHer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 304.

Diagnosis.—Rays 5. R=32 mm.; r=14.5 mm.; R=2.2 r. Breadth of ray
at base, 17.5 mm. A variation with shorter rays: R=1.8 r. Rays short and
broad at base; interbrachial ares very wide and rounded; abactinal surface sub-
plane to slightly arched. Paxille spaced and small or large and more closely
placed; very variable; papule conspicuous, extending 0.7 to 0.8 length of ray;
abactinal plates strongly lobed. Marginal plates conspicuous, the inferomarginals
with a widely spaced subpapilliform granulation and a median transverse row of
four prominent sharp spinules which are reduced to two on outer half of ray.
Superomarginals with three to five transverse rows of spaced granules one to three
enlarged into subtubercular granules on outer end of plate; inferomarginal spine
on outer end of plate very prominent in young specimens. Adambulacral plates
angular, with palmate furrow series of four or five terete blunt spinelets, and on
actinal surface a semicircle of unequal ‘‘barley-corn” spinelets on border of plate
surrounding a similar but enlarged central spinule. Sometimes pectinate pedicel-
lari present on furrow between adambulacral plates, and on some of actinal
intermediate plates near mouth plates. Mouth plates with prominent median tooth
at inner angle, and an angular marginal series of unequal spinelets. Actinal inter-
radial areas extensive but plates not encroaching far on ray; plates with ‘‘barley-
corn” spinelets and frequently an enlarged spinule. Superambulacral plates
present.

Description.—The variable shape of this species is noted under “‘ Variations.”’
Paxille usually small and spaced but extremely variable in size; arranged in
regular series parallel with median radial, except over center of disk where there is
no regular arrangement. Paxille also form transverse oblique series and decrease
in size toward marginal plates and center of disk. Paxille of type (and of speci-
mens from hard bottom) have three to five robust, but small, clavate or pestle-
shaped granules, alternating with or surrounded by a regular series of six to ten,
or slightly more, slenderer spinelets or granules. Small lateral paxille with only
three to six spinelets. An extreme variant from type (station 4367) has large
rather closely placed paxille with a convex crown of upwards to fifteen central low
polygonal or roundish granules, surrounded by numerous slender spinelets in a
peripheral series. About three transverse rows of paxille, in this specimen, stand

188 BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

opposite two superomarginals (four in type). Only one series of paxillee reaches
terminal plate, but the two adradial very nearly. In typical specimens papule
are readily seen between paxille, generally six about each.

Dorsal plates or bases of paxille are distinctly and often very conspicuously
lobed on the proximal radial regions. At tip of ray beyond the papular area the
plates become hexagonal or elliptical; along margins of area they lose the lobes,
and vary greatly in shape, even becoming triangular; on center of disk the plates
are without well-defined lobes and are irregular. On the proximal radial areas
there are six lobes to a plate, the plates overlapping by them. These lobes vary
greatly in length, being longest and slenderest on typical forms with small paxille.
Papulz extend along ray nearly to tip (0.7 to 0.8 length measured on side) ; else-
where they are generally distributed except where interradial septum joins abac-
tinal surface.

Marginal plates well developed. Superomarginals, about twenty-five in num-
ber from median interradial line to extremity of ray, form a steep, arched bevel to
border, Plates wider than long, the width decreasing gradually alongray. They are
covered with spaced, low, hexagonal, or roundish, convex-tipped granules forming
two or three transverse series, with an additional series of smaller, slenderer, often
pinched granules on either transverse margin. The larger median granules increase
in size toward outer and lower margin of plate near which are one to three enlarged
subtubercular granules.

Inferomarginals correspond to superomarginals in position, and form a similar
bevel to actinal surface. They are covered with a spaced papilliform granulation,
and a median transverse row of four prominent sharp spinules, which are reduced
to two on outer half of ray. Narrow fasciolar channels run transversely between
the plates of both series. When denuded of granules the specialized ridges of
marginals are seen to be much thicker than intervening channels. Terminal plate
ovoid and prominent, the narrower end toward paxillar area. Young specimens
have two prominent tubercles at tip, on either side.

Adambulacral plates with an angular furrow margin; of this the side toward
mouth is shorter than the aboral; or the margin may appear, superficially, to be
curved. Plates wider than long. Armature consists of (1) a palmate furrow
series of four or five cylindrical or slightly compressed, slightly tapering, blunt-
tipped spinelets, sometimes swollen and subpapilliform. Rarely there are six or
as few as three. (2) On actinal surface is a semicircle of unequal, very small,
swollen papilliform spinelets on border of plate, surrounding a similar but enlarged,
often curved, central spinule. The outer plates of the series frequently have the
spinelets arranged in two rows, and there are two enlarged spinules; or these
spinelets may be arranged in a circle around the enlarged spinule, which increases
in prominence toward tip of ray.

Mouth plates prominent actinally, and the combined pair with a broadly oval
contour, the wider end toward actinostome. Armature of margin consists of a
prominent median (unpaired) spine, slightly flattened, and on each plate an angular
series rising toward peristome very slightly. This series consists of two or three
spinelets on margin toward actinostome and about four or five on the furrow margin,

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 189

each series being graduated in size toward the angle, where there is usually one very
small spinelet. This marginal angle is nearer the inner than the outer or aboral
end of free margin. On actinal surface is a superficial series of seven or eight terete
blunt spinelets bordering the suture and on the adambulacral margin four or five
continuing the furrow series, both decreasing in size toward the outer end of plate.
A spinelet of latter series sometimes becomes enlarged and stands midway between
the two series, forming an actinal spinule homologous with that of the adambulacral.

Actinal interradial areas fairly large, but plates extending only one-third
length of ray measured alongside, or one-half measured from center of disk. Plates
arranged in series extending from adambulacrals to inferomarginals, and covered
with comparatively few, spaced, swollen, small spinelets usually forming a circle
surrounding an enlarged central spinule. These spinelets have a narrow proximal
and a swollen pointed distal portion, and resemble the so-called ‘‘barleycorn”’
spinelets of many Astropectinide. When dry these spinelets become very slightly
slenderer, but retain the characteristic form.

The channels running interradially between the series of intermediate plates
form undifferentiated fascioles. A small percentage of the specimens have pectinate
pedicellarize or fascioles on the actinal intermediate area or between the adam-
bulacral plates. In thelatter case a comb of about four spinelets on each plate bend
over the transverse suture and meet spinelets of an adjacent plate. This is simply
the ‘“‘postadambulacral fasciole”’ transferred to the adambulacral plates. The
actinal intermediate pectinate pedicellarie occur mostly over the transverse sutures
between proximal plates of the series adjacent to adambulacrals. Three to five
somewhat enlarged, slightly flattened spinelets with truncate or rounded often
slightly flaring tips compose each comb. There is considerable variation in the
amount of specialization or differentiation undergone by the spinelets. Some are
considerably flattened, especially in the case of the adambulacral pedicellariex.
An interradial pedicellaria just back of mouth plates, and one on either side is of
fairly frequent occurrence. Rarely a small pedicellaria is found between these and
marginal plates.

Madreporic body fairly conspicuous, situated midway between center and
inner edge of superomarginals; edges overhung by seven to ten paxille; outline
variable; striations irregular centrifugal.

Color in life: Abactinal surface dull coral red, rather deep, varying to maroon in
small specimens, in which case the abactinal paxillar area is much darker than
marginal plates, which are bright coral red; actinal surface of inferomarginal plates
pinkish buff; actinal intermediate areas grayish, often tinged with pink; tube feet
olive buff torawsienna. A second phase is of the same general tint, but much paler.

Anatomical notes.—Superambulacral plates present but poorly developed, and
absent from first two ambulacral ossicles; rudimentary or lacking beyond middle
of ray. Gonads forming a single tuft at either side of interradial septum. Anal
opening present,small. Intestinal caecum divided into two parts, each of which has
three or four irregular branches with a few small pocket-like subdivisions. One
Polian vesicle in each interradius. Tube feet with well-developed sucking disks;
no deposits in walls.

190 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Variations.—This is a variable species both in respect to shape and details of
ornamentation, particularly in the size of paxille. The last variation seems to be
correlated with the character of the bottom.

The extremes in body form are best appreciated from the photographic figures,
and range from a short broad-rayed almost arcuate-pentagonal shape (which is
unusual) to one with smaller disk and longer, distally attenuate rays. The follow-
ing table will give an idea of the differences:

Table showing variations in Pseudarchaster pusillus.

Breadth of | Number of
Station. | R. re Ris:

|
| ray at supero-
base. marginals.
[eae ocak: ea A ——_|—

4423 | 29 16.0 AS se | 18.0 20
4308 28 14.0 One, je a70) 21
4423 32 14.5 Dh2 iyi 17.5 25
si91 | 44 18.0 2.4451 | 20.0 30

Next to the shape the size of paxille is most noticeable in variability. The
extremes have been noted in the foregoing description. Specimens with small
spaced paxille are from hard sandy bottom; those with larger and more closely
placed paxillee (as, for instance, the specimen mentioned from station 4367) are from
softmud. The large paxille evidently are a protective device for the papulz, prevent-
ing their being clogged with mud. The danger of such an event is very slight on hard
sand or shells; hence the partial atrophy of the paxillar crown in specimens from
such a substratum. Unfortunately the type is not so typical in regard to paxille as
the variety with more closely placed and slightly larger ones. The specimen from
station 4367 is extreme. (See figures.)

There is variation in the width of superomarginals, and in the degree in which
they encroach on paxillar area. This gives a slightly different appearance to
abactinal paxillar area. Small variation occurs in the number of transverse rows
of granules, as well as in the length of inferomarginal spinules. Variations in
occurrence of pedicellariae and in adambulacral armature have already been
mentioned in the description.

Y oung.—Numerous very small specimens from stations 4433 and 4533 are easily
recognizable as Pseudarchaster. R=5 mm. Paxille have one to four spinelets;
superomarginals seven; terminal plate large with two spinelets; one very prominent
inferomarginal spine, forming longitudinal series all along ray; adambulacral plates
with three furrow and two to four much shorter actinal spinelets; the median un-
paired tooth of mouth plates very prominent. The conspicuous inferomarginal
spine with several associated spinules projecting out from side of ray is very
characteristic.

Type.—Cat. No. 22335, U.S.N.M.

Type-locality.—Albatross station 4423, between Santa Barbara and San Nicholas
Island, California, 216 to 339 fathoms, shells, black pebbles, green sand.

Distribution.—North of Monterey Bay, California, to Los Coronados Islands,
Lower California, and from 54 to 382 fathoms.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 191

Specimens examined.—Four hundred and sixty-one, from the following stations:

Specimens of Pseudarchaster pusillus examined.

Station. | Locality. Depth. Nature of bottom. |N umber. Collection.
| Fathoms.

2892 -< 22. | Santa Barbara Channel, California. . 284 yellow mudi.:s.--<2ssecsssecceees 57 U.S. Nat. Mus.
29094... . | Off Point Conception, California... 205 green mud za 272 Do.
20458 5. = 5- South of Santa Cruz Island, Cali- 266 gray sand, gravel, broken shells... 2 Do.

fornia.

3105. ....- Off Half Moon Bay, California... ... 217 1 Do.
3187....-. Off Point Sur, California............ 11 Do.
3189...... South of Point Sur, California....... 218 3 Do.
Blot ceneee eee WO See ees enews renee noes eee es 211 47 Do.
3200...... Santa Barbara Channel, California..| 265 5 Do.
OM sees = Monterey Bay, California........... 382 2 Do.
3669. ..... Off Monterey Bay, California....... 278 green mud, fine sand............... 4 Do.

4339...... Vicinity of Los Coronados Islands, | 241-869 | green mud.....................--.. 1 Albatross, 1904.

Lower California.

43505 -2 =. Near San Diego, California.......... 191-98 | gray sand, green mud 1 Do.
4367....--|...-. Ge eee ee newee ane ee ae on aene ee ae| Mes alO | PECON UMN Goin noe acne aoe 1 Do.
215-240 |..... C0 Sonex nen casasa ieee cans seas 3 Do.
216-339 | shells, black pebbles, green sand. ..| 3 Do.
197-281 | black sand, pebbles. 1 Do.
270-243 | green mud.......... | Do.
293-194 | green mud,sand................... | 32 Do.
71-54 | very hard sand.................---. | 1 Do.
20008 LALO ee eect ote a eS | 1 Do.

a The majority of these specimens are from 2909. The contents of more than one bottle are represented, however. Several
_were broken together during the California earthquake of 1906, and the above label was in the remains of one bottle.

Remarks.—This species resembles Ps. verrilli Ludwig” in several ways, but
differs in lacking the independent connecting plates on the papular areas of dorsal
skeleton, as well as in having true pectinate actinal intermediate and interadam-
bulacral pedicellariz (when any pedicellariz are present), in having inferomarginal
spinules, and enlarged superomarginal granules on outer end of plate, a different
adambulacral armature and a prominent unpaired tooth at inner angle of mouth
plates. Ludwig’s Ps. pulcher® appears to be closely related to pusillus, but the
papule are not so restricted in the latter, and the absence of marginal spinules, par-
ticularly in the young, as well as the absence of an enlarged unpaired tooth are suffi-
cient to separate it. Other differences exist in details of structure.

Ps. pusillus differs from Ps. parelit in having shorter, broader rays, breader
paxillar area on rays, less confpact paxille, less compact and more hemispherical
superomarginal granules, much more prominent inferomarginal spinules, especially
in young, less densely granulated actinal intermediate plates, less numerous furrow
and actinal adambulacral spinelets. The general facies of the two species is very
different.

b Mem. Mus. Comp. Zo6l., vol. 32, 1905, p. 116. (Gulf of Panama).
e Idem, p. 110.

192 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
PSEUDARCHASTER DISSONUS Fisher.
Pl. 34, figs. 1-3; pl. 57, figs. 7, 7a.
Pseudarchaster dissonus Fisuer, Zool. Anz., vol. 35, March 29, 1910, p. 551.

Diagnosis.—Closely related to Ps. pectinifer Ludwig, but differs in having
actinal bivalved pedicellarie of a highly characteristic form, coarser granules to
abactinal paxille and superomarginal plates, proximally smaller adambulacral
plates, fewer actinal adambulacral spinelets, and pointed furrow spines. R=170
mm.; r=61 mm.; R=2.6 r.. Breadth of ray at interradial line, 72 mm.; at one-
third R from center, 50 mm.; at two-thirds R, 24mm. Interbrachial arcs wide and
rounded; rays tapering, with straight sides, pointed; abactinal area sunken on
interradii and capable of inflation. Abactinal paxille large, with fifteen to twenty-
five central, and about the same number of slender peripheral granules; abactinal
plates subcireular. Marginal plates narrow, tumid, the two series difficult to differ-
entiate; covered with thickened papilliform spinelets or elongate granules on upper
series and spinelets on lower; two to four or five enlarged spinules on inferomar-
ginals. Adambulacral plates longer than broad, with very acute furrow margin
bearing six to eight spinelets, and on actinal surface three to five slender spinelets
and one or two enlarged sharp spinules; all spinelets sheathed in membrane. Acti-
nal intermediate areas very extensive; plates covered with slender, skin-sheathed
spinelets; each plate with one, sometimes two, enlarged spinules, all directed
toward margin. On basal third of ray, valvate pedicellariz on transverse sutures
between first longitudinal row of actinal intermediate plates; or valvate pectinate.
Well-developed superambulacral ossicles, lodged in membranous septa running
from ambulacrals to inferomarginals.

Description.—Paxille large; largest on proximal radial areas, decreasing in size
toward marginal plates and radially toward extremity of ray; in giant specimen less
obviously toward center of disk. Paxillar crowns hexagonal; pedicel of paxilla
broad and low, widening at summit; this tabulum, in large paxille, covered with
fifteen to twenty-five central slightly spaced, roundish or subhexagonal flat-topped
granules, quite robust, and variable in shape, but usually low and subcylindrical,
and a peripheral series of fifteen to twenty-five slenderer peripheral granules.
Toward margin the spinelets are much fewer and form a convex more compact
group. Three longitudinal series touch terminal plate.

Abactinal plates subcireular, arranged in longitudinal series parallel with the
radial, except on center of disk and extreme margin of area, where there is no regular
arrangement. On the latter area the plates are often more or less elongate longi-
tudinally and in median radial area there are slight, indications of lobing. Papule
single, and six about each plate. They are distributed all over the abactinal area
except for about 8 mm. at tip of ray.

Superomarginal plates, fifty-six to a ray, are relatively small, and do not
encroach conspicuously on abactinal area. Each plate is distinctly tumid when
the spinelets are cleaned off. Sutural lines very indistinct, it being impossible to
distinguish the suture between the two marginal series without first cleaning plates.
(In slightly smaller specimen (station 3788) this line can be seen readily.) The two

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 193

series of plates together form a rounded margin to ray. Plates covered with thick-
ened papilliform granules or stumpy spinelets, like those on adjacent paxille, these
increasing in length toward lower edge and passing without break into spinelets of
inferomarginals. On proximal third of ray two or three spinelets near lower end of
plate are enlarged into more robust clavate spinules, which in interbrachial are form
a transverse row, or a group with several spinelets of intermediate size. All are
incased in membranous sheaths.

Inferomarginals opposite superomarginals and encroach much more on actinal
area than do superomarginals on abactinal. They are likewise tumid and are coy-
ered with rather slender papilliform spinelets (thickly cased with membrane),
which may become subsquamiform. Enlarged flattened lanceolate spinules form an
irregular transverse series of four or five, or there may be an oblique series of two or
three near inner end of plate, and a longitudinal series or group of several only
slightly enlarged spinelets at upper end of plate. The spinelets are packed together
and rather difficult to diagnose, inasmuch as they vary in length and intergrade
with the enlarged spinules. Terminal plate small, apparently granular, and ovoid
in shape.

Adambulacral plates with an acute furrow angle projecting conspicuously
between adjacent tube feet. These projections are rather widely spaced owing to
the length of the plates. Armature spiniform, not granuliform. It is difficult to
diagnose exactly the armature owing to slight variations. On the very angular
margin are five to eight fairly slender tapering spinelets, with membranous covering,
which cause them to appear flattened and blunt. The median, on the apex of angle,
is usually slightly longer than rest. On actinal surface are two to five delicate
spinelets and usually one or two enlarged sharp spinules sheathed in membrane. Of
the latter there is usually but one at base of ray, and two distally, where they are
more prominent. A peculiarity of the plates is the fact that at base of ray they are
so spaced that an intermediate plate may border the furrow for a short distance,
and two or three actinal intermediate spinelets may stand in between two consecu-
tive furrow series. Very soon, however, the plates come close enough together to
have transverse sutures.

Actinal interradial areas large, paved with rather small superficially oblong,
squarish, or even nearly round plates arranged in series running from adambulacral
to margin. Internally the plates imbricate, and are broadly irregular oval in form.
On the ray the plates also form longitudinal series, one extending about 0.8 length
of ray, a second over 0.6, a third about 0.5. On the basal half the longitudinal series
are not distinct except the two or three adjacent to furrow. The plates are armed
with slender spaced spinelets directed toward margin. These are slightly clavate
or are terete, vary in length, and are normally sheathed in membrane, which may
become thickened at end, obscuring sharp or blunt tips. One or two enlarged, more
robust spinules stand on each plate. Very narrow, shallow channels separate the
consecutive interradial or transverse series of plates.

On the basal third of ray, on the transverse suture separating consecutive
actinal intermediate plates of the series adjacent to adambulacrals is a valvate pedi-

57444°—Bull. 76—11——_13

194 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

cellaria with jaws about twice as long as high. Each jaw is on a separate plate and
they meet above the suture, being usually close to the furrow margin, and opposite
a suture between two adambulacrals. They are confined to the proximal third of
R, and are modifications of the ordinary postadambulacral fascioles. In Ps. pec-
tinifer the spinelets of these are prominent and show a tendency to become flattened.
In dissonus the flattening is probably aided by a coalescence of spinelets, and the
result is a pedicellaria essentially like that of the typical Goniasteridx, except for
the fact that the jaws stand on separate plates. The simplest postadambulacral
fascioles are thus incipient pedicellariz. The figures will, perhaps, show the exact
form better than further description.

Combined mouth plates elliptical, slightly convex, broadest at about the
middle, and of the same general appearance as those of pectinifer. Besides the large
median unpaired blunt tooth are nine or ten slightly sharpened marginal spines.
The actinal surface is armed with numerous slender spinelets and one or two enlarged
spinules. (See description below of a smaller specimen.)

Madreporic body small, situated 0.4 minor radius from center. Striations deep
and radiating. It is partly overhung by paxille.

Anatomical notes.—Superambulacral plates well developed beyond eighth to
eleventh ambulacral ossicle. The first is on the fourth or fifth. The first plate
meeting an inferomarginal is on eighth to eleventh ambulacrals. The superambu-
lacrals are lodged in a sort of laminar septum, with an upper free edge, extending
between the ambulacral ossicles and the inferomarginals. In this low septum the
plate is firmly held. It is more in the form of a lamina than a rod. In the big
specimen, and to a less extent in the smaller ones, the superambulacral ossicles appear
to be jointed, probably due to artificial fractures partly healed, or possibly permanent,
owing to the unusual length of the plate. For instance, in the type as far along the
ray as there are two longitudinal series of intermediate plates these fractures occur,
and proximally there may be two in the length of a single plate. In Ps. pectinifer
the same tendency is observable, though not so marked. Over the distal region
with one row of actinal intermediate plates the superambulacrals are quite short,
and become rudimentary, then cease a short distance beyond the last intermediate
plate. Gonads in a single tuft on either side of interradial septum. Tube feet
with a prominent sucking disk; no deposits. Anus present.

Notes on an immature specimen.—A specimen from station 3601 has been
referred to this species on the evidence of the peculiar pedicellarie#, adambulacral
armature, and superambulacral plates. Superficially it presents a number of points
of difference, which are probably due to age and the greater depth.

R=52 mm.; r=20 mm.; R=2.6 r. Rays narrow distally. Paxillze conspic-
uously smaller on center of disk and on interradial areas than on radial areas;
the latter with ten or eleven central granules and twelve to fifteen smaller com-
pressed peripheral ones. Marginal plates not encroaching much on either area,
as in type, but the covering rather more granuliform. Adambulacral plates, as in
type—narrow, with a very prominent furrow angle; about six furrow spinelets, but
as few as three on a few proximal plates. Mouth plates prominent actinally, the

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 195

pair with an ovoid form. On the broad inner truncately rounded end of combined
plates a series of seven spinelets, slightly flattened and bluntly pointed, is graduated
on either side from the median unpaired spine—there being thus three spinelets to
each plate; on edge toward mouth of furrow are three or four smaller stumpy
spinelets graduated either way from the central. After an interval this series is
continued along margin adjacent to first adambulacral in five or six small papilliform
spinelets. About eight to ten spinelets stand on edge of median suture and five or
six, one of which may be enlarged, form an intermediate series.

The postadambulacral pedicellaris or fascioles are somewhat different from those
of type, and are intermediate in some respects between the latter and the ordinary
form. They extend about a third the length of furrow. At the base of the series
each consists of two or three flattened low jaws on either plate, forming a sort of
compound pedicellaria. One pedicellaria has a single long jaw on one plate and
opposing it two jaws, the whole occupying the entire width of the intermediate
plate. At the distal end of the series the pedicellariz are single and like those of
type. It seems probable that as growth proceeds two opposing jaws develop, while
the others atrophy. The figures will give a better idea of these pedicellarie than a
description will. (Pl. 57, fig. 7a.)

The intermediate plates are armed with short papilliform spinelets, not so long
nor so slender as in type. There is usually one enlarged spinelet to each plate on
the disk. Papule extend 0.6 length of R on radial area; not so far on sides, and are
lacking in a triangular interradial space.

Young.—A specimen with R=28 mm. is doubtfully referred to this species. It
has not yet developed the characteristic pedicellarie. The character of the spinula-
tion, armature of mouth plates and of adambulacral plates has led to the supposi-
tion that the specimen belongs here.

Type.—Cat. No. 27882, U.S.N.M.

Type-locality.— Albatross station 3346, off Oregon, 786 fathoms, green mud.

Distribution.—Bering Sea to Oregon, 786 to 1,064 fathoms.

Specimens examined.—Five: the type; an immature specimen from station 3601,
Bering Sea, between St. George Island and the Bogoslof Islands, 1,044 fathoms, green
mud, fine sand; a young specimen doubtfully referred to this species, station 3075,
off Sea Lion Rock, Washington, 859 fathoms, green mud; two specimens from station
3788, off Cape Blanco, Oregon, 1,064 fathoms, green mud.

Remarks.—This species is characterized by postambulacral fascioles of a bivalved
form, which may consist, however, of multiple opposing jaws, and by the details of
abactinal, marginal, and adambulacral armature, as noted below in the comparison
with Ps. pectinifer, to which it is closely related.

I have two specimens of Ps. pectinifer Ludwig? from station 2807, Galapagos
Islands, 812 fathoms, globigerina ooze, coral mud. It is very evident that dissonus
bears the closest sort of relationship with pectinifer and probably has a continuous
distribution with it. No specimens of an intermediate character or from interme-
diate localities are at hand however.

@ Mem. Mus. Comp. Zodl., vol. 32, 1905, p. 106, pl. 7, figs. 34, 35.

196

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

The following apparently constant differences separate the north Pacific species
from pectinifer, the larger specimen of which has R=95 mm. (the same as Ludwig's

type from the Gulf of Panama):
pectinifer.

Granules of central portion of tabulum only half
as broad as those in an equal-sized specimen of dis-
sonus.

Superomarginal plates at base of ray twice as
wide as long; granulation of even length, or only
one or two slightly enlarged granules near outer
end of plate; granules very short, truncate, or
slightly convex at tip.

Inferomarginal spinelets slenderer and more
numerous.

Actinal intermediate spinules slenderer and
longer.

Postadambulacral pedicellariz pectinate only.

Adambulacral plates proximally as wide or wider
than long, with truncate or blunt untapered furrow
spinelets; numerous actinal spinelets; intermedi-
ate plates and pedicellarize normally placed, not
encroaching upon adambulacrals and furrow mar-

gin.

dissonus.

Notably larger granules to the abactinal paxillx,
and fewer on surface of tabulum.

Superomarginals at most one and one-half times
as wide as long; outer granules conspicuously
longer than inner, and one or more in the form of
short spinelets shaped something like a grain of
wheat; other granules much fewer than in pectinifer,
and more or less elongate.

Inferomarginal spinelets coarser and fewer.

Actinal intermediate and
stouter.

Postadambulacral pedicellarize bivalved, but
with sometimes a valvate pectinate form.

Adambulacral plates proximally much longer
than wide, with tapering sharp furrow spinelets;
few actinal spinelets; intermediate plates and their
pedicellarize (proximally) encroach upon adambu-
lacrals and sometimes push between them and
border the furrow.

spinules shorter

In pectinifer the adambulacrals are joined by their transverse margins, which
are well marked; but in dissonus, as shown in the figures, there may be no true lateral
sutures between the plates, as the first row of intermediate plates encroaches upon

them.

furrow between two adjacent but partially separated adambulacrals.

A portion of the inner edge of an intermediate thus comes to border the

There is no_

hint of this in pectinifer, the surface of whose adambulacrals is much more spacious

and spiny.

Subfamily GONIASTERIN 4 Verrill, 1899 (extended).

Goniasterine+ Mediasterine VERRILL, Trans. Conn. Acad., vol. 10, 1899, p. 200.
Pentagonasterine SLADEN, Challenger Asteroidea, pp. xxxi, 262, 264.—PERRIER, 1894, p. 337.

Genus MEDIASTER Stimpson.

Mediaster Strupson, Boston Journ. Nat. Hist., vol. 6, 1857, p. 530. Type, M. xqualis Stimp-
son.—SLapEN, Challenger Asteroidea, 1889, p. 263.—Prrrrer, Expéd. sci. du Travailleur et du
Talisman, 1894, p. 377; Résult. camp. sci. du Prince de Monaco, fasc. 11, 1896, p. 45.—VER-
RIL, Trans, Conn. Acad., vol. 10, 1899, p. 178.

Isaster VERRILL (not of Desor), Proc. U. S. Nat. Mus., vol. 17, 1894, p. 257. Type, Archaster

bairdii Verrill.

Pentagonaster SLADEN (not Gray) part, Challenger Asteroidea, 1889, p. 277.

Diagnosis.—Goniasteride having astellate form, witha broad flat disk, wide inter-

brachial ares and evenly tapered rays; with tabulate, regular more or less paxilliform
abactinal plates which meet the terminal plate in from one to three longitudinal
series; with internal radiating ossicles connecting the abactinal plates of papular

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 197

areas; papule single or grouped in twos or threes; marginal plates commonly not
tumid; actinal interradial areas extensive; adambulacral plates with straight
furrow margin bearing three to eight rather slender, often prismatic spinelets, the
actinal surface with a rosette or double row of shorter spinelets which may be
angular and obtuse, and toward tips of rays some of them may become larger and
longer. Some of these spinelets may be replaced by spinuliform or clavate, two-
or three-bladed pedicellariz. | No odd median mouth spine as in Pseudarchaster.
Abactinal and marginal pedicellariz, low valvate foraminate; actinal intermediate
pedicellariz like the abactinal, or with slender spatulate jaws. Rudimentary
superambulacral plates present.

Remarks:—This genus is closely related to Ceramaster, as here limited, but
differs in having internal radiating ossicles connecting the slightly lobed well-spaced
abactinal plates, and in possessing also rudimentary superambulacral plates.
These two characters will also separate Mediaster from Nymphaster and Nereidaster
both of which I have examined with reference to these features. The analytical
key will serve to contrast the present genus with others more distantly allied.
The Nectria of Gray which bears a certain resemblance to Mediaster is probably
not so close as the abactinal surface would lead one to expect. There are inter-
marginal papule, superambulacral plates, and sometimes pincer-shaped pedicellariz.
with several jaws.

The first adequate account of this genus was given by Verrill (1899) who ranked
in it the following forms: M. xqualis Stimpson, M. bairdii Verrill (olim Isaster),
M. agassizii Verrill, M. pedicellaris Perrier (olim Goniodiscus), M. arcuatus (Sladen),
M. japonicus (Sladen) M. patagonicus (Sladen). The last three species were
described under Pentagonaster by Sladen. A very important feature of the genus,
the rudimentary but constantly present superambulacral plates, was overlooked
by Verrill, who was the first to discover the internal abactinal ossicles. The three
species of Sladen I have since examined with special reference to the presence of
the internal ossicles and superambulacral plates, as these forms were not accessible
to Verrill. Arcuatus is a Mediaster, but patagonicus and japonicus lack the diagnos-
tic characters of the genus, and fall within the limits of Ceramaster. Both pedicel-
laris and agassizii have not been dissected so that they must still remain question-
able members of the genus, as also M. florifer (Alcock). It is impossible to identify
a Mediaster from an external examination only. The following species belong to
Mediaster as here defined: xqualis Stimpson, bairdii (Verrill), including stellatus
Perrier, arcuatus (Sladen), tenellus Fisher, transfuga Ludwig, ornatus Fisher. The
additional problematical species are: pedicellaris, agassizii, and florifer. Mediaster
elegans Ludwig (1905) lacks the diagnostic features of the genus; there is no advan-
tage in placing it in the present group.

KEY TO THE SPECIES OF MEDIASTER HEREIN DESCRIBED.

Papule grouped in twos and threes; pedicellarize bivalved, longer than high; pedicellarie of actinal
interradial areas like those of the abactinal surface; adambulacral spinelets rather robust and
IRIN ULC nee olen eee eon eno dnle dei aan a een Cee en ae ee ene Ds

Papulz single; pedicellariz broadly spatulate, slightly higher than wide, those of actinal interradial
areas with long slender jaws, unlike the abactinal; furrow spinelets long, compressed, rather
TIGLICHULC Re. eet ae res ate eI Nee wise cals tuiein aR as alc Sl eee eae wk Od tenellus, p. 202.

198 BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

MEDIASTER ZQUALIS Stimpson.
Pl. 35, figs. 1-8; pl. 59, figs. 1, la-c.

Mediaster xqualis Stimpson, Boston Journ. Nat. Hist., vol. 6, 1857, p. 530, pl. 23, figs. 7-11.—
J. F. Wurreaves, Trans. Roy. Soc. Canada, vol. 4, sec. 4, 1886 (1887), p. 117 (Malcolm Island,
Queen Charlotte Sound, British Columbia)—StapEn, Challenger Asteroidea, 1889, p. 572.—
VerriLL, Trans. Conn. Acad., vol. 10, 1899, p. 179, pl. 24, figs. 10-12.

Description.—Rays five. R*=84 mm.; r=32 mm.; R=2.6r. Breadth of ray
at base, 36mm. R°=56 mm.; r=22 mm.; R=1.54 r. Breadth of ray at base,
26 mm. Interbrachial ares wide, rounded or subangular; rays regularly tapering,
rather slender toward tip, blunt. Marginal plates conspicuous, but as a rule not
encroaching much on abactinal area; sides of body rather evenly rounded.
Superomarginals, twenty-eight to a ray (in a specimen with R=84 mm.), almost
plane to slightly tumid, wider than long, covered with many small pentagonal to
roundish, close set, but nevertheless distinctly spaced, flat-topped granules, of which
the marginal series is larger, and transversely oblong; a plate here and there may
bear a typical low bivalved pedicellaria with two or three jaws, these not extending
above the level of the granulation, but with a base four or five times as wide as
height of jaw. Inferomarginals similar to superomarginals, the suture between
the two series being more or less zigzag.

Abactinal surface usually slightly convex, covered with large tabulate plates or
parapaxille well spaced as a rule, and bearing a flat-topped ornate crown of up-
wards to twenty-four central, trapeziform, pentagonal, or more rarely roundish,
truncate or convex slightly spaced granules, and upwards to 25 slightly longer
peripheral granules, with an upper face four-sided and beveled inward; frequently
some of the granules are replaced by a large low bivalved pedicellaria; median
radial row of plates attain the terminal plate and the adradials frequently, but not
always. Papular areas extensive; papule two or three in a cluster, about six
clusters around a plate; lacking at tip of ray and from a small triangular interradial
area near margin; abactinal plates roundish, with six short lobes, well spaced
(except off the papular regions), and connected by small independent internal
radiating ossicles, six to a plate.

Actinal interradial areas extensive, the intermediate plates extending nearly
to tip of ray; they are close set, broadly ovate, slightly imbricated and bear roundish,
rhombic, or elliptical groups of two or three central (also one or none) and five to
ten peripheral prismatic, elongated or spiniform granules. A few plates especially
near furrow, bear a good-sized bivalved pedicellaria with two, three, or four jaws,
sometimes higher and not so long at the base, as abactinal pedicellarie.

Adambulacral plates squarish, not large, with a furrow series of three to five,
oblong, more or less prismatic or compressed blunt spinelets, subequal in size; external
to these, two successively shorter series form a rosette-like group, those nearest
the furrow series being thicker than the latter, prismatic, or quadrate in section
and about three in number; the outermost series usually has four or five granules
like those of adjacent intermediate plates; occasionally on the outer half of ray an
adoral spinelet of the inner series is replaced by a pedicellaria with two or three

@ Example from Albatross station 4228—S. E. Alaska.
> Example from Albatross station 3159—off Point Reyes, California.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 199

short, stout, truncate, slightly tapering, spiniform jaws. On the distal part of
the ray one or two of the spinelets on the central part of the plate become con-
siderably longer and thicker than the rest.

Mouth plates, rather small and not prominent actinally, with a furrow series
of five to seven prismatic compressed spinelets (the innermost largest and most
compressed) and a parallel actinal series of three to five thicker prismatic spinelets
all similar to but larger than the adambulacral spinelets; outer part of plates with
granules.

Color in life, abactinal surface deep brilliant poppy red to vermilion; actinal
surface “salmon” orange; tube feet scarlet with sucking disks of flesh color.

Anatomical notes.—Abactinal plates, when viewed from the inner surface, are
rather small and roundish with short blunt lobes, widely separated in fully grown
specimens, and connected by radiating independent ossicles. (See fig. 1, pl. 59.)
These ossicles leave triangular papular spaces in which two or three papule com-
monly occur. The papular area is very extensive, comprising practically all the
abactinal surface except the very restricted interradial areas (where the plates are
close together and lack ossicles) and a small space at the tip of ray. The papular
area is more extensive, especially interradially, than in MV. bairdii, M. ornatus, or
M. arcuatus, all of which possess the internal radiating ossicles.

I was considerably surprised to find rudimentary superambulacral plates in
this species. Mediaster bairdii, M. arcuatus, and M. ornatus were examined and
found to possess them also. Ceramaster lacks any trace of the plates, as does also
Nymphaster ternalis and Nereidaster bowersi. I think it probable that all species
of Mediaster have rudimentary superambulacrals. In J. xqualis they begin with
the second ambulacral ossicle and extend far along ray, or until the lower end of
ambulacral ossicles touch the inferomarginals. The plates are small and oval in
form (PI. 59, fig. 1c), are entirely covered by connective tissue, so that the skeleton
must be treated with caustic potash before they can be seen. In M. arcuatus
(Sladen), WM. bairdit (Verrill), and M. ornatus Fisher the ossicles do not begin so
near the mouth, but start at varying distances therefrom, and die out under the
same conditions as in xqualis. In MM. arcuatus the plates are slender and longer
than in any other species.

The actinal intermediate plates, from the internal view, are broadly oval and
slightly imbricated.

Intestinal ceecum consisting of five irregular, unequal, sparsely lobed divisions,
the longest reaching to margin; no central portion, the divisions meeting at the
intestine; anus present; hepatic ceca long; stomach small. Gonad in three tufts
on either side of the uncalcified interradial septa. Tube feet with sucking disks.

Variations—When compared with specimens from Chignik Bay, Alaskan
Peninsula, Southern Californian examples present no greater differences than such
as result from individual variation. Northern specimens appear to be better sup-
plied with pedicellarix, especially on the distal adambulacral plates, but even in
this respect Chignik Bay examples vary somewhat.

Among the Californian specimens the most important differences are due -to
age. The small examples have more prominent marginal plates than the medium-
sized and large examples, and much fewer pedicellariw#, these being absent from
the distal adambulacral plates. The adambulacral pedicellarie are rare in Cali-

200 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

fornian specimens. Some specimens have mostly four furrow spines to a plate,
others three, while still others both numbers; four is the usual number for typical,
large examples, though some have three, and less commonly five. There is also
variation in the number of granules to the abactinal paxilliform plates, especially
of the radial areas. For instance, a specimen with R=42 mm., on the midradial
parapaxille, at base of ray, has about eighteen to twenty peripheral granules and
fifteen to seventeen central. Another specimen with R=59 mm.—consequently
larger—has only fourteen or fifteen peripheral and ‘five to ten central granules.
The granules vary a little in size, being slenderer and more spiniform in the
peripheral series in some examples, crowded and more granuliform in others.

The rays vary in thickness, especially distally, according to whether one, two,
or three rows of plates attain the terminal plate. In some examples the actinal
intermediate pedicellarize are higher than the abactinal, and have a shorter base
line; in others they are scarcely different from the abactinal.

A remarkable variation is presented by one specimen from the type-locality,
Puget Sound. In this example there are numerous small secondary abactinal
parapaxille on the median radial areas (proximally). Practically all the adam-
bulacrals have five furrow spinelets and the first actinal series is set obliquely.
There is one six-rayed and one four-rayed specimen in the collection.

Young.—The smallest specimen (station 2934) has R=8 mm.; marginal plates,
eight; papule restricted to radial areas; tabula low or absent; granules, two or
three central and about eight peripheral on proximal radial region; two or three
abactinal pedicellariz; three furrow spinelets; one or two actinal pedicellarie.

Types—In Coll. Suckley (Puget Sound), and in U.S. National Museum (Cali-
fornia), Cat. No. 1977.

Type-locality —Puget Sound, Washington, or San Francisco, California.

Distribution—From the Alaskan Peninsula (Chignik Bay) south to northern
Lower California, in shallow water (9 to 160 fathoms).

Specimens examined.—About two hundred and thirty-one specimens from the
following stations:

Specimens of Mediaster xqualis examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
2873......| Off Cape Flattery, Washington . 40 ROCKS enawecscnciete see seco 3 U.S. Nat. Mus.
Wi aoosess| ener OOS oe socne cern te eee ee 27 _~_—*| rocks and shells............-- 8 Do.
S75 tee | ee DO oe Ie Ao) jes 8tdowse eee ee ee 4 Do.
2886...... Heceta Bank, Oregon ........... 50 Locky-aeses--= sso 2 Do.
294 S555 Off San Miguel Island, California. 53 broken shells, sand...-......- 1 Do.
2906......| Santa Barbara Channel, Cali- 96 sand MU paces ecco 4 Do.
fornia.
D0 imowsetnceas GL ee eran eae an seen oe 44 fine gray sand.........---.... 16 Do.
2008 cane 2 |=22me GO sea seew ns coset nee aoee 31 gray sand, broken shells...... 1 Do.
2922......) Cortez Bank, off San Diego, Cali- 47 fineisraySAnd-< co.e cee sees 1 Do.
fornia.
GO)" |e dos. 28st SS Ee 1 Do.
36 STAY SANG eee eee cee 3 Do.
ATM) Wilesane GOsste sees See eee 2 Do.
27 fine gray sand_.....:......... 14 Do.
| 20 gray sand, broken shells...... 4 Do.
2944.2... Off Anacapa Island, California...| 30 TOCK Y= occacecen cst ed ene 3 Do.
2950..... | Between Santa Cruz and Santa 21 gray sand, broken shells... .-- 1 Do.
Rosa Islands, California.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER.

Specimens of Mediaster xqualis examined—Continued.

Station. Locality. Depth.
Fathoms.

55
Near Santa Barbara, California. | 21
21
27
61

7
Off Anacapa Island, California... | 36
46
$059:....-. Off Siletz Bay, Oregon.......... 77
3068. ..... Admiralty Inlet, Washington....) 135
3080...... OP Orepon aeres+<acade cece ewes 93
3087...... Off Alseya River, Oregon........ 46
BOnSaeee as anes ober eseceesenassh aps seenne 46
800555 --.2 Off Heceta Bank, Oregon... .... 42
3106...-.-. Southwest of Half Moon Bay, | 7

California. ;
eens Iran One ee eeeeeaneeeereeee as 43
Monterey Bay, California....... ] 9
Off Point Reyes, California ..... 27
eee tnen Onn merece ee as 39
.| South of Monterey Bay.........- 77
Off Avalon, Santa Catalina Is- 47
land, California.

Off Santa Catalina Island, Cali-
fornia.
Off Santa Cruz Island,California .

Monterey, California............. |
Puget Sound, Washington.......
Bea eens! Near Naha Bay, Behm Canal,
southeast Alaska. |
4281...... Chignik Bay, Alaska. ........-.-
asilenceee Near San Diego, California... ....
4343.....- Near Los Coronados Island,
Lower California.
4414...... Between Santa Catalina and San-
ta Barbara Islands, California.
4420...... Between Santa Barbara and San
Nicholas Islands, California.
M431. .<2.. Near Santa Cruz Island,California
4439...... Monterey Bay, California......--
MADD tease el see's MD eoictans does fe sam acon ce eee
Gia oeeiese. MO erento oan acne nae
DOSE ena elnw ssn QO ver san eM ace~ see tne cnn
Q5Gl oo vanslan mae DOpseh reece eeeeuesuce wuss |
4552 Seales -2= Odweet saqeteecereeheenaesaetor
S553 8.0 Sclcoan ois a aaa de teen

4stations.; Off San Pedro, California

160-141

38-32

3041
42
55-167
5451
86-76
56-46
83-66
74-05

Yea Bay, Alaska. . 5.020... e ces ca]esncsscsss|saccsencecsess

Nature of bottom.

green mud, gray sand, broken
shells.

rocks, stones, broken shells...
fine gray sand................

mud, gray sand, rocks.......-
gray sand, broken shells......
green mud, gravel............
soft dark green mud..........

green mud and sand

Number.

ee OMe ON NH eRe ee

om wt eonwwhrs-

ete

oe

to

se Pere nee Re ee Re eo

|

201

Collection.

U. 8. Nat. Mus.

Do.
Do.
Do.

No. 8764, D. S. Jordan.
Chas. M. Drake.
Albatross, 1903.

Do.
Albatross, 1904.
Do.

Albatross, 1905.

University of California.
Do.

Stanford collection,

202 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Remarks.—This species is occasionally caught by fishermen in Monterey Bay
and Puget Sound. Its life colors are very bright. It seems to inhabit several
kinds of bottom ranging from rocks and shells, hard sand, gravel, clay, and pebbles,
to green mud, and a mixture of mud and rocks.

MEDIASTER TENELLUS Fisher.
Pl. 36, figs. 1, 2; pl. 57, figs. 3, 3a-d; pl. 59, fig. 2.

Mediaster tenellus FisHer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 307.

Diagnosis.—Rays five. R=58 mm.; R=19 mm.; R=3r. Breadth of ray
between first and second superomarginals, 20 mm; at middle of ray, 7 or 8 mm.
General form flattened; disk large; rays fairly long and slender, tapering abruptly at
base, then more gradually; interbrachial ares very wide, and rounded; abactinal
surface inflated on radial areas, sunken on interradial areas, the rays especially
being convex; actinal surface considerably inflated on disk; marginal plate small,
confined nearly to sides of body; abactinal area with tabulate plates or parapaxille
having an ornate crown of numerous prismatic spinelets, and often a small upright
two-jawed pedicellaria; A4dambulacral plates with five long strongly compressed
furrow spinelets, and three actinal spinelets about three-fourths as long; a few
adambulacral and actinal intermediate pedicellariz with high slender jaws. Rudi-
mentary superambulacral plates present; internal ossicles connecting abactinal
plates of papular area; papule single.

Description.—Abactinal area covered with ornate, regularly spaced para-
paxille or tabulate plates, largest in a regular median radial series, decreasing thence
toward tip of rays and margin of disk; paxille of median radial series elongated
transversely, the others roundish; the former bearing on the periphery of the tabu-
lum fifteen to seventeen prismatic, blunt spinelets, as long as, or slightly longer
than, the lesser dimension of the tabulum and decidedly longer than is usual in this
genus; in the center are six or seven irregular, prismatic, pointed granules, much
shorter than the peripheral spinelets; ornamentation of the other paxille differing
only in having fewer spinelets and granules. At the tip of the ray the plates lose their
tabulate character. Many of the paxille bear on the edge or nearer the center, a
small upright pedicellaria, whose two broadly spatulate concave jaws are slightly
higher than wide and very much larger than the central granules. They are nearly
as long as the marginal spinelets, and curve inward slightly toward each other.

Abactinal plates, viewed from the celomic side, are seen to be well spaced,
transversely elliptical or roundish with slight indications of lobing. The median
radial plates are considerably wider than long. All plates of the papular areas are
connected by internal radiating ossicles, but these are less regular than in xqualis.
Between the radial and either adradial series there are sometimes two or even three
shorter ossicles end to end. There is only one conspicuous papula to an area, not
two or three as in equalis. They are absent from the tip of rays and a small inter-
radial triangular area adjacent to the marginal plates.

Marginal plates rather smaller than in any other species of the genus, and
throughout most of the ray the superomarginals are confined to the side wall;
thirty in number from median interradial line to extremity of ray, thin, slightly

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 203.

longer than high throughout most of ray, and forming a slightly arched, steep bevel
to side of body; first plate conspicuously higher, or wider, than long, and longer
on inner than lower edge. General surface covered with evenly spaced thimble-
shaped granules which increase in size toward the center of each plate, the periph-
eral being rather small and pinched. A few plates bear pedicellarie similar to
those of the paxilla. Inferomarginals beyond the second plate alternating with
superomarginals and wider than high, forming a narrow border to the actinal area,
thin, slightly longer than broad, and the upper edge angular, like the lower edge of
the superomarginals; granulation rather coarser than that of superomarginals, square
or prismatic, and very few pedicellarie on the series.

Adambulacral plates rather regularly quadrate and remarkable for the long
furrow spinelets, of which there are five in each series, strongly compressed, regular,
round-tipped, the adoral slightly shorter than the other four, these spinelets reach-
ing nearly across the furrow and tapering slightly as regards their thickness, the
tip being often beveled; on the actinal surface a longitudinal series of three spine-
lets, the central longest and about three-fourths the length of the furrow spinelets,
the series standing midway between furrow margin and outer margin, upon which
are three or four less regular granules similar to those of the actinal intermediate
plates; first plate and a few others with a peculiar foraminate pedicellaria on actinal
surface consisting of two or three slender upright spiniform jaws, occasionally
faintly spatulate, nearly as high as the inner actinal series. (Pl. 57, fig. 3b.) These
pedicellariz do not have a depression into which the jaws fit when opened.

Mouth plates slightly convex actinally and the furrow armature consisting of
eight spinelets, similar in character to those of the adambulacral plates; the two
inner larger than the others, and those between inner and outer members of the
series slightly shorter, or, in other words, graduated in size from the middle toward
either end of the series; on actinal surface a number of elongate granules, and one
of the peculiar pedicellariz, the latter situated near the median suture at about
middle of plate.

Actinal interradial areas large, the plates extending to ninth inferomarginal or
about half the length of ray, arranged quite regularly in series parallel to the furrow,
slightly convex and armed with six to eight spaced, elongate, prismatic granules,
usually forming an ornate rosette. A few plates have a pedicellaria with slender,
high, jaws. (Pl. 57, fig. 3c.)

Madreporiec body convex, oval, situated nearer than midway to margin;
striations coarse, undulating, interrupted, centrifugal.

Anatomical notes.—This species possesses rudimentary superambulacral plates.
It is necessary to treat the ambulacral and adjacent plates with caustic potash
solution. After this is done the small independent oval ossicles are seen at the
lower end of the ambulacral plates. They are very similar to those of M. xqualis.
They were examined on eleven plates at about the middle of the ray. It was not
thought desirable to mutilate the type further to ascertain if the superambulacrals
extended to the very base of the series, as this point is not of great importance.

Y oung.—A young specimen from station 4421 has the following dimensions: R =
13 mm.; r=6mm. It has the general facies of the adult, but all the paxille are

204 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

roundish, the median radial with one central and six to nine peripheral terete blunt
spinelets. There are four long furrow spinelets and one of the two or three actinal
adambulacral spinelets is longer than the rest. The marginal plates are more
conspicuous in the adult.

Type.—Cat. No. 22337, U.S.N.M.

Type-locality— Albatross station 4427, off Santa Cruz Island, California, 447
to 510 fathoms, black mud, rocks. A young specimen from station 4421, between
Santa Barbara and San Nicholas Islands, 291 to 229 fathoms, gray mud, rocks.

Distribution.—Southern California south probably intergrading with M.transfuga
Ludwig of the Panama fauna (the two species may be the same). Found in 291 to
510 fathoms.

Remarks.—The species most closely related to the present is Mediaster transfuga
Ludwig (1905) from the vicinity of Acapulco, 902 meters. It is probable that the
two forms are merely extremes of the same species, as the outward habit and most
of the structural details are essentially the same. JM. transfuga has slightly longer
rays and longer lobes to the dorsal plates; the area of the internal connecting
ossicles is somewhat more restricted (possibly due to younger specimens); there are
fewer spinelets to the dorsal paxille; adambulacral spinelets are shorter (?), especi-
ally on the actinal surface of plate; the actinal pedicellariz when present are like
the dorsal; in tenellus they are much higher, almost spiniform. Probably rudi-
mentary superambulacral plates are present in transfuga and escaped notice. It is
impossible to tell how much the characters of typical tenellus vary, as only one adult
was taken by the Albatross.

The Mediaster elegans described by Ludwig in the same paper (1905, p. 125)
is not a Mediaster, as it lacks the distinctive characters of the genus.

Mediaster tenellus is remarkable for the small marginal plates, the high, slender,
pedicellarie of the actinal surface, and for the unusually long furrow spinelets.
The peripheral spinelets of the abactinal paxille are also more elongate than is
usual in the genus. The species is therefore far removed from gxqualis, the type
of the genus.

Genus CERAMASTER (Verrill).

Ceramaster VeRRILL (Sec. C of Tosia), Trans. Conn. Acad., vol. 10, 1899, p. 161. Type, Asterias
granularis Retzius.—FisueEr, Bull. U.S. Fish Comm, for 1903, pt. 3, 1906, p. 1054 (subgenus).

Pentagonaster AUTHORS (not Gray) in part.

Tosia VeERRILL (not Gray) in part, 1899.—FisHeEr, 1905.

Mediaster VeERRILL in part, 1899.

Philonaster (subgenus) Ka@Hurr, Deep Sea Asteroidea collected by the Roy. Ind. Mar. Surv.
Ship Investigator, 1909, p. 74. Type, Pentagonaster (Philonaster) mortenseni Koehler.

Diagnosis.—Goniasteridx with a pentagonal, arcuate pentagonal or very short-
rayed form; with the abactinal plates entirely covered with granules, those of
papular areas with more or less elevated (sometimes almost paxilliform) tabula,
the base being slightly to conspicuously lobed; marginal plates granulated, some-
times with a plane or tumid naked area; actinal intermediate plates always gran-
ulated; adambulacral plates with two to eight furrow spinelets, their actinal gran-
ules frequently in series, graduated in size between furrow spinelets and actinal

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 205

intermediate granules; the first actinal adambulacral granules or spinelets some-
times slightly spaced from furrow series as in Mediaster. Pedicellarim bivalved
or with two or three spatulate jaws; absent in type. No trace of rudimentary
superambulacral plates, such as occur in Mediaster; no internal abactinal radiating
ossicles connecting the plates.

Remarks.—This genus is better characterized than any other nearly related
group, and while the species included in it are quite diverse in some particulars,
they hang together fairly well in respect to the structure of the abactinal, mar-
ginal, and actinal intermediate plates. The genus Tosia, as limited by Verrill in
1899 and adopted by me in 1906, is too comprehensive. Plinthaster, a section
of Tosia under Verrill’s classification, is very different from Ceramaster, as I found
after having examined Plinthaster perrieri, P. comptus, P. nitidus, and P. dentatus.
As explained above (p. 165), Pyrenaster does not seem to me to be separable from
Plinthaster. Plinthaster has flat, naked abactinal plates, while in Ceramaster the
plates are granulated and there is a distinct tabulate structure. The Tosia (Cera-
master) micropelta described by me in 1906 is not a Ceramaster, as it has small, round,
flat, granulated plates, with intermediate smaller plates on the central part of the
disk and proximal radial regions. It belongs to Peltaster or a closely related genus.

Of the list of species given by Verrill (1899) as belonging to Ceramaster I would
not include, without personal examination, mammillatus, hxsitans, and pulvinus;
placenta is in many respects an aberrant member of this group, though I have not
seen specimens. The tabulate form of the plates of the papular areas seems to
be minimized. In formulating our conception of a genus we hark back to the
type, in this case granularis. Typical species, as granularis and patagonicus, stand
about midway between such extremes as japonicus and placenta. C. leptoceramus
with its secondary abactinal plates in adults is decidedly aberrant, though appar-
ently near to japonicus. It might be made the type of anewsubgenus. C. arcticus,
with the exceptionally extensive papular areas, short adambulacral plates, and
few furrow spinelets, is still another variant from the typical form of structure.

Perhaps the nearly related genera which can be grouped about Ceramaster
are Nereidaster, Mediaster, Tosia (as here used, a very restricted austral genus),
and Peltaster, with Plinthaster, Eugoniaster, and Pentagonaster in an outer circle.
Mediaster arcuatus (Sladen), which has usually been classed as Pentagonaster, but
which possesses the internal independent radiating ossicles of Mediaster zxqualis
shows how near the two groups approach in external appearance. [From the
original figures and description Professor Verrill was led to place @. japonicus and
patagonicus in Mediaster, but they are more nearly allied to @. granularis and
must therefore be classified in Ceramaster. C. placenta seems to tend toward Tosia.
This small group, as already explained, is homogeneous enough to form an independ-
ent genus. Plinthaster and Eugoniaster are its nearest relatives, next to Penta-
gonaster (restricted). Peltaster has small flat (not tabulate) granular abactinal
plates, with few to many secondary ones among them. If the aberrant P. micro-
pelta (Fisher) is admitted to the genus the form of the pedicellarie and the adam-
bulacral armature will have as great a range as in the most diverse species here
classified under Ceramaster.

206 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

KEY TO THE SPECIES OF CERAMASTER HEREIN DESCRIBED,

a. A well-defined more or less triangular interradial area free from papule; papule single; interradiai
abactinal plates never lobed; furrow spinelets four or more.
b!. Purrow spinelets numerous, six or more proximally; abactinal tabula well spaced, rather convex.
with many small crowded granules, of which the marginal series is sensibly smaller.

c!, No abactinal secondary plates; actinal intermediate granules flat and closely crowded; furrow

spinelets about six proximally; margin of disk typically thicker; pedicellariz low, bivalved.
japonicus, p. 206.

@. Secondary abactinal plates on radial areas; actinal intermediate granules well spaced, bead-
like; furrow spinelets seven or eight proximally, slender; disk very thin and margin thin;
pedicellarie with short spatulate jJaws......-------------+------------4- leptoceramus, p. 210.

b?. Furrow spinelets usually not more than five; proximally abactinal tabula closer, with larger, fewer
granules, the marginal granules being equal to or larger than the central ones.

c!, Superomarginal plates usually with a tumid bare spot; the granules small and crowded, form
ing a smooth pavement; abactinal tabula with numerous central granules; adambulacral
plates proximally as wide as long and first actinal series not enlarged into conspicuous stubby
tubercularsepinelets =~ cit aa tel <n ela ee ee ais eel patagonicus, p. 214.

2, Superomarginal plates small, only a few distals with a tumid bare spot, their granules coarse
and hemispherical; abactinal tabula with few granules; adambulacral plates proximally
narrower than long, with four long furrow spinelets and a series of actinal spinelets.

clarki, p. 217.

a?, Papule distributed all over abactinal surface, usually two to four in a group; all abactinal plates
lobed; furrow spinelets two or three; adambulacral plates very short, more than twice as wide

US LOD Sassi ra to wre lars ie asim a ete pep rate te Leal m eat) weasel te eee etef tenet arcticus, p. 219.

CERAMASTER JAPONICUS (Sladen).
Pl. 37, figs. 1-3; pl. 59, fig. 3.
Pentagomaster japonicus SuaDEN, Challenger Asteroidea, 1889, p. 272, pl. 46, figs. 1 and 2; pl.

49, figs. 1 and 2.
Mediaster japonicus VERRILL, Trans. Conn, Acad., vol. 10, 1899, p. 179.

Diagnosis.—One of the largest if not the largest species in this or any closely
alhed genus. R=102 mm.; r=65 mm.; R=1.56 r (varying to R=1.36 r). Gen-
eral shape pentagonal or arcuate pentagonal, depressed, the edges of body more
or less thickened; abactinal surface usually swollen on the radial areas; actinal
surface nearly plane. Abactinal surface covered with fairly regular, spaced,
hexagonal to quadrate tabula crowned with a convex group of numerous polygonal
granules, the marginal series regular, sometimes compressed, and when dried show-
ing a pit in the top. On the interradial areas tabula are quadrate and smaller.
Many of the tabula with a bivalved pedicellaria, having thin, lamelliform jaws
which are wider than high. These vary in size, equaling one-third or one-fourth
the width of the tabulum. Marginal plates variable, covered with small flat crowded
granules, forming an evenly rounded, or a decidedly tumid margin to ray; a bare
space sometimes present on plates of both series, each superomarginal, with one
to four small bivalved pedicellariz, whose thin jaws equal two to four granules
in width. These are situated usually on the abactinal surface of the superomar-
ginals near the margins; inferomarginals with usually only one pedicellaria or
none. Actinal intermediate areas very extensive, the plates decreasing in size
toward margin, and covered with crowded low flat polygonal granules, of which
the marginal are slightly the largest; individual plates perfectly flat and vari-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 207

ously four-sided; actinal intermediate pedicellarie in variable numbers. Adam-
bulacral plates usually slightly wider than long, with a slightly oblique furrow
series of five or six (four to seven) round-tipped, four-sided, often somewhat flat-
tened, stout spinelets, frequently graduated in length toward the adoral edge of.
plate, or all but the shorter adoral number subequal. Actinal surface with three
to five irregular longitudinal series of low granules, those nearest the furrow spine-
lets being largest. A bivalved pedicellaria is normally present on most of the
plates, the jaws being usually wider than high, but exceptionally higher than wide.
The granules of actinal surface of adambulacrals vary greatly in number and those
on the outer half are not in very definite series.

Madreporic body pentagonal situated one-third distance from center to inner
edge of marginal plates.

Anatomical notes—Anal opening present, subcentral surrounded by three
to six plates larger than those adjacent. Intestinal cecum large, with eight long,
unequal glove-finger-like radiating divisions. Intestine spacious, stomach small;
hepatic ceeca long, extending two-thirds to three-fourths length of R; quite a
sharp distinction between dorsal and ventral divisions of stomach. Gonads very
large, bushy, each with four short ducts and four apertures in a line parallel to and
a short distance from the tough, uncalcified interradial septum. Dorsal muscle
bands stout, joining at center of disk. Each band gives off numerous lateral
smaller branches, obliquely like the barbs of a feather, on the side toward margin.
Polian vesicles large, one in each interradius except that of the madreporic
canal; ampulle double; tube feet with large sucking disks. No superambulacral
ossicles.

When the plates which bear the tabula are examined from the inner surface
of the stout dorsal integument they are seen to be arranged in very regular rows
parallel with the median radial. The radial row and the two on either side have
plates with six short, broad, rounded, or truncate lobes. On either side of these
the lobes are very soon lost and the plates become circular. All plates are slightly
spaced and there are no internal connecting ossicles as in Mediaster. Papule are
single and distributed all over the abactinal surface, except in a narrow interradial
band reaching toward center of disk as far as the madreporic body; they are absent
also at the tip of ray; four to six, or even more, occur around each plate, according
to its size and shape.

Variations.—The examination of seventy specimens of this species, including
one from off Misaki, Japan, 640 meters, kindly sent by Dr. Seitaro Goto, of Tokio,
reveals a considerable range of individual variation. The differences which the
American specimens seem to present are greatly overshadowed by individual
differences occurring in examples from the same station. Sladen apparently had
only one specimen, and of course could give no hint of this variation in his excellent
description.

General form, etc—Some specimens are more arcuately pentagonal than
Sladen’s figure and of the same form as his figure of patagonicus. The Misaki
specimen is about intermediate in form. The Washington specimens are less
arcuate than many of those from Bering Sea. The abactinal area varies in the

208 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

amount of inflation, and corresponding to this the interradial sulcus is more or less
evident. There is one four-rayed example.

Abactinal plates—Two large specimens from stations 3330 and 3331, Bering
Sea, in practically the same locality exhibit very nearly the extremes of variation
in the size of plates (PI. 37, figs. 1 and 2). That from station 3330, specimen A (fig. 2)
has R=92 mm., while B has R=91 mm. In A the largest tabula of the median
radial region are 3 mm. in diameter, in B 3.5 mm., but in A there are relatively
much fewer large tabula and the spaces between them are much wider than in B,
as will be seen from the photographs. This difference in the size of the tabula
and in the extent of space between is duplicated in other localities. The larger
tabula have larger granules, not a correspondingly greater number; eighteen to
twenty-four marginal granules and eighteen to thirty in the central group is the
range in large specimens. In dried specimens the marginal granules are concave
on the top, the others are flat. Pedicellarize are usually very numerous, occurring
on a majority of plates of the papular areas.

Marginal plates.—The greatest variation of all is in respect to the marginal
plates. Typically these form a tumid rounded border, but exceptionally are
nearly as thin as in leptoceramus. Again, the two specimens mentioned above will
serve to illustrate the extremes of size. In A, which is less typical, the plates
are entirely covered with close, flat granules and measure in the interradius 4 mm.
by 4 mm., being less regular in size and not at all tumid. The thickness of the
edge of the disk is 5 mm., which includes the height of the combined plates. In B
the plates are very regular, 5 mm. long by 6 wide, and the edge of the disk is 8 mm.
thick. The superomarginals are strongly tumid, but entirely covered with gran-
ules except for a few distal plates. In A there are forty-two to forty-four supero-
marginals and about fifty inferomarginals to a side of the pentagon; in B thirty-
six superomarginals and forty-two inferomarginals. In typical japonicus as
described by Sladen the superomarginals all had an extensive tumid bare space on
the abactinal surface, and the inferomarginals ‘a smaller one on the actinal surface.
The size of this spot is variable even when it is present. In my Misaki specimen
the spots are smaller than those described by Sladen. One of the Washington
examples has the bare spaces (on superomarginals only) about as in the Misaki
specimen; three others lack them. Out of twenty-four specimens from station
3608 only three have a bare spot on the superomarginals and inferomarginals. A
small specimen from 3331 has fairly large tumid bare spots, while a large example
(“B”) lacks them. The shape of the plates is variable. In specifnens with large
marginals the width is greater than length, but when the plates are small and not
tumid the proximal plates may be square or even longer than wide. The extreme
in this respect is a specimen from station 4775 which has proximal plates 4 mm.
long by 3.25 wide. Japanese specimens vary in this respect, as the type had
longer proximal superomarginals than the Misaki example, which has the plates
decidedly wider than long throughout. A remarkable variation is exhibited by a
specimen from 4774, which has the marginal plates slightly concave, the transverse
sutures being on the summit of a tumidity. The number of pedicellarie on the
superomarginals varies greatly, two or three are commonly present, sometimes
three or four on all except the most distally situated. The inferomarginals some-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 209

times lack pedicellariz, but usually most of the plates have one or even two which
are slightly larger than the superomarginal ones.

Actinal intermediate plates.—A characteristic of this species is the very crowded
polygonal, flat-topped, smooth actinal intermediate granules. They are not at all
globular, and this character seems fairly constant. It is well marked in the Wash-
ington specimens and serves admirably in connection with other features to separate
japonicus from leptoceramus. The latter species has spaced hemispherical granules.

The small bivalved pedicellaria vary considerably in numbers. They may
be restricted to the angle adjacent to mouth plates, or scattered part way or all
along the row of plates adjacent to adambulacrals. Many specimens, but not a
majority, have them also sparingly elsewhere on the intermediate areas (Misaki,
Washington, Bering Sea.)

Adambulacral plates.—In proportions the plates vary from being as wide as
long to wider than long. The outer edge is frequently oblique or obtusely angular.
The furrow spinelets are prismatic or subquadrate in section, blunt, usually but
not always with a slightly curved, oblique, base-line. The Washington examples
have proximally four or five, rarely six, furrow spinelets, distally three to five; the
Misaki specimen proximally six or seven, distally six spinelets, and the base line is
more evidently oblique. Large Alaskan examples have proximally five or six
spinelets, distally four or five. Young Alaskan specimens (R=30 mm.) have four
or five furrow spinelets throughout. The Misaki specimen has rather shorter
spinelets than the Alaskan and Washington examples. The first series of actinal
granules varies from three to five. Far along the furrow one of these granules is
enlarged into a blunt tubercle. The bare space between the furrow spinelets and
first series of granules is variable in width, wider in the Japanese specimen than in
the Alaskan or Washington examples. The number of granules exclusive of the
first series varies from about twelve to twenty-two. The small number occurs
when there is a pedicellaria on the plate. In specimen A, before cited, very few
of the adambulacrals have pedicellarie. In the Misaki example nearly all have
one. Nearly as great a range of difference is present in the Washington examples,
and in Bering Sea specimens, as, for example, station 3608.

Young.—The young bear a great resemblance to GC. granularis (Retz.). In
this species, however, the superomarginals are longer than wide, whereas all the
young japonicus have them wider than long. Granularis lacks pedicellariw alto-
gether, the abactinal tabule are flat-topped with large marginal granules, and the
actinal intermediate granules are roundish and not crowded. The adambulacral
armature is coarser and the spinelets and granules fewer. In young japonicus
the papule are restricted to broadly ovate areas on the radii, the tabula are lower
with fewer granules, and no abactinal pedicellarie or only few. The marginal
plates are sometimes relatively thicker than in the adult.

Type.—In the British Museum.

Type-locality.— Challenger station 232, south of Yokohama, Japan, 345 fath-
oms, green mud.

Distribution.—Japan (Misaki, and south of Yokohama) to southern Bering
Sea, thence south along the American coast to Oregon.

57444°—Bull. 76—11——14

210 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Specimens examined.—Seventy-three, from the following localities:

Specimens of Ceramaster japonicus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
lr, 1 ww cy 4 ta See ip re 7?) eke
| Fathoms.
3227 225 QreeI MU eececce ence 4 U.S. Nat. Mus.
3330 351 black sand, mud......... 1 Do.
3331 350 INU Stocoacine scan sissies 5 Do.
3346. 786 greenimudircssessececsee 4 Do.
3488...... 106 green mud, gray sand.... 1 Do.
3489 eee | Soe GOtee recesses eee 2 Do.
S502) eee 368 green mud, dark sand.... 26 Do.
860822<--~ | Bering Sea between St. George and Unalaska. . 276 QTaY, SANG); nen eee sea 24 Do.
4768...... Bowers Bank, Bering Sea, 54° 12’ N.; 179° 07’ ea preenimudeseccsoraeciones 1 Albatross, 1906.
4774...... Bowers Bank, Bering Sea, 54°33’ N.; 178° 45’ E. 557 green mud, black speck, 1 Do.

| foraminifera.
AIG someee Howes DO cen- opel ences caw da seaaee ecase ees jase Ea) besa QOseas--oseueeeeeses 3 Do.

| Misakf,Japan’=2. -:c-scc.2.0cee cesecesee este 3208) | toesnan neecee se eee 1 Stanford.

Remarks.—In lieu of giving a detailed description of this species, I have listed
the chief variations presented by a large series of specimens. The various figures
will serve sufficiently, in connection with Sladen’s excellent description of the type-
specimen, for purposes of identification. Except in the case of young specimens
there is no danger of confusing this species with C. granularis, to which it seems
to be rather closely related. The salient differences have been already pointed out.
The differences between japonicus and leptoceramus are mentioned below under the
latter species.

CERAMASTER LEPTOCERAMUS (Fisher).
Pl. 39, figs. 1-3; pl. 58, figs. 3, 3a; pl. 60, fig. 2.
Tosia leptocerama FisHeRr, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 306.

Diagnosis.—Related to C. japonicus Sladen, but with secondary abactinal
plates, thinner disk and margin, spaced bead-like actinal granules, and shorter and
more numerous furrow spinelets. Type, R=63 mm.; r=44 mm.; R=1.43 r.
Disk thin, general form pentagonal, marginal plates small, closely granular, some-
times with a bare spot; superomarginals not in contact distally; radial areas inflated;
body very thin interradially (about 1.5 mm.). Abactinal plates stellate on papular
areas, raised into spaced tabula bearing numerous granules and usually also a small
two-jawed pedicellaria; numerous smaller secondary plates intercalated between the
larger ones on proximal part of papular areas. Actinal intermediate plates with
spaced bead-like granules and two-jawed pedicellarie. Adambulacral plates with
a regular furrow comb of seven or eight short compressed quadrate slender spine-
lets, and two rows of actinal granules.

Description.—Abactinal surface covered with spaced, low, tabulate plates,
arranged with great regularity in a radial series, which is largest, and numerous
other parallel series, decreasing rapidly in size toward the margin, where the plates
are small and more crowded. In center of disk the plates are arranged without
order, and are more or less irregular in a narrow interradial area, which is free from
papule. Bases of plates of radial areas, when viewed from the inner side, in a
prepared specimen are substellate, regularly with six truncate or blunt processes,

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 211

a shallow sulcus running from the center of the plate to each interradial angle.
The plates are entirely free from one another, those of the radial and either adradial
series being lengthened transversely, the others rounder. In the vicinity of the mar-
ginal plates they become more crowded and lengthened longitudinally. In the
center of disk the plates are roundish, and in the narrow interradial area quadrate,
hexagonal roundish, lozenge-shaped, or, near margin, oblong. In the proximal
portion of the radial areas, small secondary plates or paxille, with narrow ossicle-
like bases, are interspersed rather irregularly, but form a fairly regular series between
the radial and either adradial row of plates. The low tabulum surmounting each
plate is slightly convex, hexagonal to elliptical on radial areas, roundish to irreg-
ularly quadrate interradially, elliptical and compressed near border, and bears a
central group of robust, flat-topped, quadrate or polygonal granules (about ten
to fifteen on radial plates) surrounded by a peripheral series of smaller, square-
tipped, flattened regular granules. The granulation does not appear crowded.
Nearly all the plates bear one or two delicate pedicellarie with thin wide-spatulate
jaws, higher than the granules. At the edge of the disk where the plates are very
irregular, only the marginal series of granules may be present on account of the
compression of the plate. The papulz are conspicuous on account of the open
character of the tabulate armature, and are distributed as in C. japonicus. The
abactinal membrane is rather flexible, the plates not being in contact.

Marginal plates more or less elongate and, owing to the thinness of the disk,
rather small; except for the first two or three plates, the members of the two series
alternate. Superomarginals, fourteen in number from each interradial line to
extremity of series, or twenty-eight to each side of disk, about as high as broad, and
much longer than high, except at the end of the series, where they are shorter.
These plates form a narrow border to the abactinal area and are covered with a
rather uniform, dense, polygonal granulation, a peripheral series being readily
distinguishable. The outer plates, or sometimes all, have a small naked area on
abactinal face, bearing one to three small, two-jawed pedicellarie, larger than the
granules. Sometimes on the proximal plates only a narrow area about each is
left free by the granules; lower margin of each superomarginal angular, forming
two facets for articulation with two inferomarginals.

The inferomarginals, sixteen to a ray or thirty-two to a side, are much wider
than high, and encroach farther upon the actinal area than do the dorsal series
upon the abactinal. The total width of the plate in large specimens is often less
than the length, but in smaller specimens the width equals, or exceeds the length.
This is true of some large examples. In the middle of the actinal surface is a small
naked area, increasing in size toward outer end of series, which bears one to three
small, two-jawed pedicellarie, but slightly larger than the granules. There is consid-
erable variation as to the extent of the naked areas, which are much reduced on some
specimens. Granulation of inferomarginals similar to that of superomarginals.

Adambulacral plates nearly square with a straight or slightly curved furrow
margin. Armature consists typically of a furrow series of eight or seven untapered,
compressed, square-tipped spinelets about half as long as the length of the plate,
subequal or slightly shorter at either end, very regular and standing parallel.
These spinelets are four-edged, or quadrate in section. On the actinal surface are
two irregular longitudinal series of smaller granules, the inner well spaced from

919 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

the furrow spinelets as in Mediaster, and consisting of three or four compressed,
square-tipped, sharp-edged granules, larger than those of the outer series. These
are very irregular in distribution, eight or nine in number, and similar to, but a
trifle larger than the actinal intermediate granules, which they resemble but are
usually less regular in form. There are sometimes two irregular rows of these
smaller outer granules. At adoral end of inner actinal series is a pedicellaria with
two broadly spatulate jaws, slightly larger than the granules of the series, which
the pedicellaria greatly resembles when the jaws are closed; at tip of ‘‘ray” there
are four to seven spinelets in the furrow series, and on the actinal surface a blunt,
prominent tubercle, surrounded by several granules, this tubercle grading into the
granules of the inner actinal series. The pedicellaria persists to within two or three
plates of the tip.

Mouth plates triangular, with a longer furrow margin than the edge adjacent
to first adambulacral; furrow series consisting of ten to thirteen spinelets similar
in character to those of adambulacrals, but increasing in size toward inner angle,
where there is an oblanceolate, blunt, flattened, or compressed tooth, the other
spinelets square in section or slightly compressed. On actinal surface is a linear
series of low, squarish granules adjacent to the median suture, several others along
the aboral edge, and two or three intermediate between the superficial and marginal
series.

Actinal interradial areas very extensive; intermediate plates quadrate,
roundish, or irregular, those adjacent to the adambulacrals much larger than any
of the others, and usually oblong in shape, the short end toward furrow. The
plates are arranged in rows parallel to furrow, considerable irregularity existing.
Plates are covered with numerous uniform, hemispherical, bead-like granules dis-
tinctly spaced. Pedicellariz similar to those of adambulacral plates scattered
here and there, especially near the furrow, but their numbers subject to great
yariation. They are not so numerous as in C. japonicus. The appearance of the
spaced, round, bead-like granules is strikingly different from the crowded, flat,
polygonal granulation of these areas in C. japonicus.

Madreporie body unsymmetrically pentagonal, larger than any of the plates,
situated one-third distance from center to margin. It is slightly convex and has
branching ridges radiating from an eccentric point.

Color in life, vermilion, yellowish on actinal surface.

Anatomical notes.—The internal anatomy is very similar to that of C. japonicus.
The intestinal coecum, in the specimen examined, is large with four principal divi-
sions, two of which are divided at the end. The gonads form a series, on either side
of, and parallel to the interradial septum. There are five to eight tufts in each
series. The dorsal muscles are arranged as in C. japonicus but are much weaker.

Variations.—The variations in this species parallel to a considerable degree,
those of C. japonicus. The general shape ranges from a nearly straight-sided
pentagon (R=1.31 r) to very broadly stellate (R=1.5 to 1.73 r) the latter being
young specimens; there are also pentagonal examples among the young. The tabula
of the abactinal plates are broader on the radial areas of some examples; corre-
sponding to this, the typically wide spaces between the tabula are narrower. The
tabula of the radial area are sometimes less regular in form, being nearly elliptical.
In young specimens the secondary intermediate plates of the abactinal radial

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 213

areas (with their small tabula and few granules) are confined to the vicinity of
the proximal radial plates. As the size increases the secondary plates extend
farther toward tip, and spread out laterally till they are scattered over nearly all
the proximal half of the papular areas. These intermediate plates have roundish
to triangular groups of granules, frequently bear a pedicellaria, and are never
half as large as the adjacent plates. When a large primary tabulum bears
fourteen central and about twenty-four smaller peripheral granules, a secondary
tabulum has one to eight central and seven to twelve peripheral granules; the
smallest have as few as four or five peripheral granules only.

_ While abactinal pedicellariz are variable in numbers they are always numerous.

Young specimens usually have a thicker margin in proportion to size than
the large ones. The proportions of the plates vary from wider than long to longer
than wide in the interradial region; distally they are always wider than long.
Near the inner edge of the plates more or less of a bare space is generally present
which increases in size toward the end of the ‘‘ray.”” Sometimes only the distal
plates of the series have it. ‘

The range of variation in the thickness of margin is not so great as in C.
japonicus; the margin is normally thin and the plates are not tumid.

The granules of the actinal intermediate plates are low hemispherical, bead-
like, and always spaced. The pedicellarize may be confined to the angle adjacent
to mouth plates, and hence few, or else scattered all along the first row of inter-
mediate plates, and sometimes a few occur here and there elsewhere on the inter-
radial regions.

Normally the furrow spinelets form regular combs which have a gently con-
vex distal contour, there being a well marked bare space between them and the
first actinal series of granules, which are low, compressed, and elongated longi-
tudinally, three to five in each series.

Type.—Cat. No. 22336, U.S.N.M.

Type-locality.— Albatross station 4378, off Point Loma, near San Diego,
California, 376 to 594 fathoms, green mud and sand.

Distribution —Off southern California, between San Diego and Point Con-
ception, 216 to 638 fathoms.

Specimens examined.—Sixty-two from the following localities:

Specimens of Ceramaster leptoceramus examined.

Station. Locality. | Depth. Nature of bottom. Number Collection.
ea ee SS Ss es
Fathoms. |
2037.....< Off Southern California, east of San Clemente 464 =| green mud...............| 10 U. S. Nat. Mus.
Island. |
NMR WG eaten Seles Uaek< sew ekin| etna aut elic nau ah swe xa sam ueaea ilaie ae 8 Do.
4378...... Off Point Loma, near San Diego, California.....| 376-594 | green mud and sand... .. 1 Albatross, 1904.
4407......| Between San Clemente and Santa Catalina | 334-600 | rocks, shells, fine gray 2 Do.
Islands, California. } | sand.
4415...... Between Santa Catalina and Santa Barbara | 638-302 | green mud............... | 1 Do.
Islands, California. |
4416...... Between Santa Barbara and San Nicholas | 448-223 | dark green mud.......... 4 | Do.
Islands, California. |
BETES ocaafrouan UO i uecaccarenneatanendes dea apestodesas -.--| 330-216 shells, black pebbles, | 1 Do,
green sand.

A i ante Off Santa Cruz Island, California............. 475-510 | black mud, broken stones. 35 Do.
|

914 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Remarks.—This species probably ranges considerably north and south of its
present known distribution. It shows no signs of intergradation with @. japonicus
to which it is more nearly related than to any other known species. From japonicus
it differs in having secondary intermediate abactinal plates on the radial areas,
hicher tabula, typically thinner margin, spaced bead-like actinal intermediate
granules, shorter and more numerous furrow spinelets, much more numerous
actinal adambulacral granules. The same characters will serve to distinguish
the young also. Small leptoceramus have much thinner margins, more numerous
abactinal pedicellarie, more distinctly spaced tabula, and more numerous furrow
spinelets than young japonicus.

The pedicellarix, especially the actinal, of leptoceramus are higher and have
more delicate jaws than those of japonicus.

CERAMASTER PATAGONICUS (Sladen).
Pl. 37, fig. 4; pl. 38, figs. 1, 2; pl. 60, fig. 3.

Pentagonaster patagonicus SuapEN, Challenger Asteroidea, 1889, p. 269, pl. 46, figs. 3, 4; pl. 49,
3 and 4.

Astrogonium granulare? Wuirnaves, Trans. Royal Soc. Can., vol. 4, 1886 (1887), p. 117 (40
fathoms, Strait of Georgia, off mouth of Qualicum River).

Mediaster patagonicus VerrtLL, Trans. Conn. Acad., vol. 10, 1899, p. 145.

Diagnosis.—Pentagonal, with straight sides and short rays, or arcuate pentago-
nal. R=76 mm.; r=52 mm.; R=1.46 r. Abactinal area flat or slightly inflated;
body fairly thick; marginal plates massive, the superomarginals more or less tumid,
covered with small polygonal crowded granules, except for a bare spot sometimes
occupying the whole abactinal face of plate. This bare space may be quite tumid.
Inferomarginals with a bare space on actinal surface. Abactinal surface with
regular hexagonal tabula radially, square or rhomboid interradially, these crowned
by a very slightly convex group of convex (truncate off South America) polygonal
granules, the marginal series being slightly larger; edge of tabula sharply cut (South
America) or slightly scalloped (Alaska). A pedicellaria with two narrow to broadly
spatulate jaws stands on a few of the tabula. The jaws fit into special depressions.
Actinal intermediate plates rhombic, squarish, or oblong, covered with coarse
slightly to distinctly spaced, quadrate, polygonal, or roundish convex (flat-topped,
South America) granules. Broadly spatulate two- or three-jawed pedicellarie in
variable numbers on actinal surface. Adambulacral plates wider than long. Fur-
row spinelets four or five, robust, more or less compressed, untapered, sometimes
subquadrate in section, or with broadened tips, subequal, or graduated toward a
shorter adoral member. Back of these on actinal surface is a longitudinal row of
two or three heavy granules or short stubby spinelets, sometimes compressed in a
longitudinal plane, sometimes subquadrate, with rounded tips. Outside of these
is a variable number of granules (about five to eight) similar to but more irregular
than those of neighboring actinal intermediate plates. A two-jawed spatulate
pedicellaria stands on some of the plates. Mouth plates elongate, plane, with a
long furrow margin bearing eight or nine quadrate or prismatic blunt spines (the

inner two quite heavy) similar in character to those of adambulacral. Armature
of actinal surface variable.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 215

Variations.—The specimens here considered differ in a few minor details from
the typical examples dredged by the Challenger from near the Atlantic entrance
to the Straits of Magellan. It is possible that these represent a new and very closely
allied species, but in view of the individual variations exhibited, and the absence of
typical examples for comparison, I would feel considerable hesitation in instituting
a new species. The specimen from the Gulf of California points to a continuous
distribution along the west coast of the Americas, but there is too little data for
any generalization.

The Alaskan examples have fewer and convex granules on the abactinal
tabula (not truncate and crowded) and the edges of the tabula are not so straight
cut but are a trifle scalloped owing to the slightly convex outer edges of the periph-
eral granules. The actinal intermediate plates have more distinctly spaced and
round-topped granules, appearing hemispherical in some cases. In Sladen’s figure
the granules appear to be fairly close together and truncate. In the Alaskan
specimens the first series of actinal adambulacral granules are slightly longer and
more spiniform, but as there is some variation even among the few specimens avail-
able it is probable-that South American specimens are also variable. The Alaskan
examples have with one exception more tumid marginal plates; but the size and
tumidity of the bare area is quite variable.

Taking into consideration the eight specimens, the contour varies as Sladen’s
figures of C. patagonicus and C. japonicus (PI. 46), all the Alaskan specimens being
like C. japonicus with the exception of two young, which are more arcuate. The
Gulf of California specimen is also arcuate but not quite so much so as Sladen’s
figure of patagonicus. The superomarginals are wider than long, that is, are trans-
versely oblong, but two specimens have the plates narrower than usual. The size
of the bare space varies from a third to over a half of the exposed surface of the
plate. Usually but a single row of regular granules borders this space on the inner
and one-half the lateral margins. One large specimen, however, has the areas
nearly as small as figured by Sladen. A small pedicellaria with two tapering or
spatulate jaws stands on the edge of the bare space of a few plates. The bare
space is smaller on the inferomarginals and the pedicellariz when present have
shorter jaws and are sunken in shallow pits.

The abactinal surface has very regular low tabula, slightly spaced and hexago-
nal on the radial papular areas, less regular and tending to roundish on center of
disk, square, rhomboid, or even elongate interradially. The primary basal plates
are clearly discernible and are larger than any of the others except the central plate,
which with three or four slightly smaller companions surrounds the anus. The
radial tabula bear a peripheral series of twelve to fifteen and about four to twelve
central round-topped low granules, the former slightly the iargest. The central
granules are polygonal or-roundish and do not touch each other. The peripheral
have the outer margin longest. In typical patagonicus as figured by Sladen the
central granules are more numerous, crowded and flat-topped, and the edges of the
peripheral are more squarely cut, as if ‘‘dressed” with a knife. In Alaskan exam-
ples the outer edges of the granules are usually fairly straight, but the granules are
not crowded, so that the edge of the tabulum is not such astraight line. In one speci-
men (station 3257) there is a row of scattered pedicellariw parallel to and about

216 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

5 mm. from the marginal plates. The tabula which bear these are nearly bare,
except for the peripheral and a few additional granules. Usually the pedicellarie
have special depressions into which the jaws fit when open.

A specimen from station 4228 has furrow spinelets with thick spatulate tips so
that the spinelets taper toward the base. Two very thick, short, stubby spinelets
stand in the first actinal row. A comparison of Sladen’s figure (pl. 49, fig. 4) with
those here published will demonstrate better than a description the range of varia-
tion of the adambulacral armature.

Young.—Two small specimens from station 3223. The smaller has R=15mm.;
r=10 mm. Form arcuate pentagonal; marginals with conspicuous bare spot;
tabula of midradial series with two to six central and nine to eleven peripheral
granules which are more flat-topped than in adults. Primary basal plates large
conspicuous with a central bare spot. Furrow spinelets three or four, truncate or
round-tipped, prismatic or four-sided; two or three compressed spinelets or gran- —
ules in first actinal series, three to five granules in outer part of plate. Mouth
plates with nine or ten marginal spinelets.

Anatomical notes.—Intestinal cceca very large. There is a small central por-
tion, from which radiate four narrow thin-walled tubes. These each divide into
two distal thin-walled pear-shaped sacs, which resemble Polian vesicles. In two
interradii, these sacs reach as far as the marginal plates, or as far along each radius
as the hepatic ceca. Gonads very small, in two tufts on either side of interradial
septum and about one-third distance from margin to center. Polian vesicle in
four interradii; lacking in that of the madreporic canal. Dorsal muscles weak;
arranged as in japonicus, but much less conspicuous. Stomach with strong
retractors. Ampulle double, tube feet with sucking disks.

The papular areas are not so extended in large specimens as in C. japonicus.
They are found in center of disk, inside a circle bounded by the primary basal plates,
and on a fairly broad petaloid radial area. They are absent on an extensive trian-
gular interradial area. The abactinal plates of the radial areas show faint lobes
and are closely juxtaposed, allowing only a small space for the papule to pass
between. Interradially the plates are regular, closely joined, and in form, square
near center, then hexagonal or pentagonal, and finally rhomboid near margin.

Type.—In British Museum.

Type-locality— Challenger station 313, near the Atlantic entrance to the
Straits of Magellan, 55 fathoms, sand.

Distribution —Vicinity of Cape Horn, Gulf of California, southern Alaska to
the southern part of Bering Sea (east to Commander Islands).

Specimens examined.—Nine from the following localities: Albatross station
3223, Bering Sea, near Unalaska Island, 56 fathoms, black pebbles, two; 2856,
near Kadiak Island, 68 fathoms, gray shells, one; 3257, Bering Sea, near Unalaska
Island, 81 fathoms, gray sand, gravel, two; 3548, north of Unimak Island, Aleu-
tians, 91 fathoms, black sand, one; 4228, vicinity of Naha Bay, Behm Canal,
southeast Alaska, 41 to 134 fathoms, gravel, sponges, one; 4791 near Bering Island,
76 fathoms, rocky, one; small bay north of Carmen Island, Gulf of California, one
(Stanford collection).

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 217

CERAMASTER CLARKI Fisher.
Pl. 38, figs. 3, 4; pl. 58, figs. 2, 2a-b; pl. 59, figs. 6, 6a.

Ceramaster clarki Fisuer, Zool. Anz., vol. 35, March 29, 1910, p. 552.

Diagnosis. —R =53 mm.; r=33mm.; R=1.6 r. General form stellato-pentag-
onal, but last two or three superomarginals in contact medially. General aspect
somewhat resembling C. patagonicus, but margins much thinner, marginal plates
smaller, their granules coarser, hemispherical, and slightly spaced ; the abactinal plates
with fewer and larger granules, more numerous and slightly larger pedicellarie;
the actinal intermediate areas with smaller and less regular plates (especially in
the series adjacent to adambulacrals), and larger, less regular, thimble-shaped
granules; adambulacral plates with subequal, robust, truncate, tubercular spinelets
in the first actinal series instead of granules, and the plates narrower than long
instead of wider than long.

Description.—The tabula of the abactinal plates are slenderer than in either
patagonicus or japonicus and the marginal granules overhang. One to six convex,
roundish or faintly polygonal, unequal, spaced granules occupy the center of tabulum,
while seven to fifteen more flattened ones occupy the periphery. The latter have
a faintly curved or straight beveled outer margin. The edge of the tabulum is not
so square cut as in patagonicus, and all the granules are larger and more distinctly
spaced. Nearly all the tabula have a pincer-shaped pedicellaria with two or
three spatulate or truncate-oblong or even slightly tapering jaws. The figure of
the paxillar tabula will give a better idea of their structure and appearance than
will a description. Papule are absent from near tip of ray and from a small trian-
cular interradial area. On this area the plates are abruptly smaller, with very low
circular groups of six peripheral and one or no central granules. This difference
in the size and spacing of the plates (these being farther apart on the papular
areas) is more conspicuous in clarki than in either patagonicus or japonicus.

The abactinal plates when viewed from the inner or cwlomic side are seen to
have a very concave under surface. The plates are strongly lobed and the lobes
are turned downward, forming the concavity mentioned. The median radial
plates have shorter, broader lobes than either adradial series (where the lobes are
spatulate and of unequal length). These lobes appear in many cases to have a
suture at the base as if to form incipient independent ossicles. The figure will give
a better idea of the form. The high tabula give to the plates of radial areas a
decided paxilliform appearance.

Margin of disk evenly rounded, the plates being small and low, the superomar-
ginals forming a roundish bevel. In both series the proximal plates are slightly
longer than wide, but about the last five or six are wider than long; and the last
two or three superomarginals are conspicuously swollen, with most of the surface
bare. In one case the last plate is the largest, but usually it is the penultimate
plate. The distal one to three superomarginals are in contact with their fellows
of the opposite side of the ray. These distal five or six plates are larger (with the
exception usually of the last plate) than the other plates of the series, which are
covered with spaced hemispherical granules. Sometimes one or two pedicellarie

218 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

similar to those of abactinal surface stand on the superomarginals. The plates
of the two series do not exactly correspond, the lower edge of the superomarginals
being, in consequence, frequently angular. Superomarginals fourteen, infero-
marginals sixteen to the ray.

Inferomarginals are about the same size as the superomarginals, but the
distal ones decrease regularly in size, and the last four or five have a tumid bare
spot. The granulation is the same as on the upper series, and a pedicellaria is
sometimes present. The marginal plates are about as small as in leptoceramus.
In a small specimen from southern California, very doubtfully referred to this
species, the marginals are slightly larger, are all wider than long, and the distal
superomarginals are not enlarged.

Adambulacral plates narrow, slightly longer than wide, with a furrow series
of four or five (sometimes six), distally three, equal, rather long, compressed, blunt
spinelets. These are considerably longer than width of plate and longer even
than the length of plate. The two median spinelets are frequently more com-
pressed than the laterals. Immediately behind the furrow spinelets is a longitu-
dinal parallel series of two or three much stouter, cylindrical, blunt, tubercular
spinelets about two-thirds the length of the furrow spinelets. Proximally these
spinelets are often compressed and quadrate in section. Far along the furrow
there are two in the series, and the adoral is reduced in size, the aboral becoming
conspicuous and equaling the furrow spinelets in length. The outer margin of the
plate is occupied proximally by an irregular row of three to six bluntly pointed
granules; distally there are two rows or a group. The adambulacral plates, far
along furrow, gradually become square.

Mouth plates large, rather plane, with a long margin to furrow bordered by
about ten spinelets, increasing slightly in size toward the inner angle, the innermost
spine being strongly compressed and leaf-like, and those adjacent three-sided in
section. About eight shorter, blunt, irregular, tubercular spinelets form a parallel
series which is continued along the first adambulacral margin. On the outer part
of the plate three or four elongated granules stand close to the median suture
(which is not visible), and the series of spinelets just mentioned may be continuous
with these instead of with the spinelets standing on margin adjacent to adambu-
lacral.

Actinal intermediate plates numerous, but the sutures are not readily seen.
The series of plates adjacent to adambulacrals are not larger than the next series,
are obliquely elliptical in general form, but have several sides. The other plates are
elliptical or irregularly rhombic, sometimes roundish. They have not the straight-
cut sides found in granularis, japonicus, and patagonicus. In the previously men-
tioned young specimen from southern California the plates are much more regular;
it is possible that the irregularity comes with age. The plates bear spaced, large,
hemispherical granules, those occupying the margin being sensibly smaller and
more or less pointed. Many of the plates adjacent to the adambulacrals bear a
conspicuous pedicellaria, with two oblong, slightly spatulate, round-tipped, or sub-
truncate jaws. Near the margin many plates have a similar but smaller pedicellaria
with two or three jaws. The granulation is much coarser than in leptoceramus,
where the granules are also spaced and uniform in size.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 219

Madreporic body circular, convex, situated one-third minor radius from center
of disk; striations interrupted and meandering.

Young.—The specimen from station 4407 has R about 28 mm. and r 18 mm.
It may be an extreme variant of C.leptocerama, but differs markedly from that form
in the spaced marginal granules and the enlarged actinal spinelets of the adam-
bulacral plates. The granulation of the actinal intermediate plates is like that
of clarki, but the plates are more regular in form. The abactinal paxilliform plates
are essentially like the type, but have fewer granules.

Type.—Cat. No. 27883, U.S.N.M.

Type-locality.— Albatross station 4772, Bowers Bank, Bering Sea (lat. 54° 30’ 30’
N.; long. 179° 14’ E.) 344 fathoms, greenish-brown sand, one specimen.

Distribution —Southern Bering Sea to southern California, and from 334 to
600 fathoms.

Specimens examined.—The type, and an immature specimen from station 4407,
between San Clemente and Santa Catalina islands, California, 334 to 600 fathoms,
rocks, shells, fine gray sand. The latter was in company with C. leptoceramus.

Remarks.—This species is characterized by the small marginals, with coarse
spaced granulation, by the coarse spaced actinal granulation, by the armature of
the adambulacral plates, and by the rather small abactinal paxilliform ornate
tabula. The species differs from C. leptoceramus in lacking secondary abactinal
plates, in having coarser actinal and marginal granules, much coarser and fewer
abactinal granules to each plate, fewer and longer furrow spinelets, and in having
the first series of the actinal adambulacral spinelets well developed and tubercular.
From C. japonicus, clarki differs in all these points; especially striking are the
differences in abactinal and actinal granulation and in the form and armature of
the adambulacral plates.

Unfortunately there is only one undoubted specimen of this somewhat aber-
rant species. It is aberrant in that it departs distinctly from the type facies of
the genus (as exhibited by @. granularis and C. patagonicus) and shows a
tendency toward Mediaster. C. japonicus and C. leptoceramus are similarly aberrant
in a slightly different direction, possibly also toward Mediaster. Mediaster, Nymph-
aster, and Nereidaster are closely related and the gap between them and such genera
as Goniaster, Pentagonaster, Tosia, Peltaster, Eugoniaster, and Plinthaster is bridged
by the various species of Ceramaster. So it does not appear possible to separate the
groups into two subfamilies, although the ‘extremes, such genera as Mediaster or
Nymphaster on the one hand and Goniaster or Pentagonaster on the other, are
strikingly different.

Named for Dr. Hubert Lyman Clark.

CERAMASTER ARCTICUS (Verrill).
Pl. 40, figs. 1, 2; pl. 58, fig. 1; pl. 60, fig. 1.
Tosia arctica Verritt, Amer. Journ. Sci., vol. 28, July, 1909, p. 63, figs. 8, 8a.
Diagnosis.—Arcuate pentagonal; R=55 mm.; r=34 mm.; R=1.6 r (also
1.5 r). The short rays, or corners of disk, bluntly rounded. Marginal plates
forming an evenly rounded margin to disk, massive, with usually a slightly tumid

bare spot, and with large flat crowded polygonal granules; bare spot sometimes
lacking on all but distal plates; either the penultimate or the last two superomar-

220 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

ginals in contact medially, terminal plate large, ovoid, usually separating the last
superomarginals of either side. All abactinal plates with very strongly stellate
bases, and all with stout, low tabula bearing, as a rule four to twelve large over-
hanging, crowded triangular, quadrate, or pentagonal marginal granules, fre-
quently surrounding one to three smaller polygonal granules; often no central
granules; the crown of granules circular or more often irregularly polygonal and
resembling, with tabula, a miniature toadstool; primary apical plates, and also
sometimes others in center of disk, with numerous central granules. Actinal in-
termediate plates with unequal, large triangular, trapeziform, square and polyg-
onal, very slightly spaced nearly flat-topped granules; plates quadrate with usually
a rhomboidal form. Adambulacral plates short and wide, transversely oblong in
form, with a furrow series of usually two, less commonly.three, short swollen, blunt,
frequently compressed upright spinelets, which do not extend much if any above
the general surface of the actinal granulation. Actinal surface with three to six
granules, one of which immediately behind the furrow series is larger than rest;
frequently an upright bivalved pedicellaria with oblong, truncate or round-tipped
jaws stands among the granules, above which it does not extend.

Description.—The abactinal paxilliform plates are very characteristic in their
armature. The granules are slightly convex, and the marginal series overhang so
that from the side a paxilla has the form of a miniature toadstool. The granules
are large, few in number, and frequently the central granules are lacking. Super-
ficially the tabula bear a great resemblance to Sladen’s figure of Gnathaster elonga-
tus. The outer or free edge of the granules is straight, curved, or angular. The
primary basal plates have much smaller and numerous granules so that they are
conspicuous among the other plates. In a specimen from the Shumagin Islands
a curious coalescence of the granules has taken place on some of the tabula. In
one case all but two, in another all but three, and in still another all the granules
have fused, leaving no sutures. The latter looks like a nail head. The plates are
in fairly regular series parallel with the median radial. No abactinal pedicellarie.

The papule seem to be generally distributed except for a narrow line where
the interradial septum meets the abactinal wall. The plates have strongly stellate
bases, the development of the lobes being on the side toward the median radial line.
Thus most of the plates have two or three lobes only, which impinge upon two or
three neighboring plates, as shown by the figures. The great extent of the papular
area causes all the plates to have these lobes, which sometimes show sutures at
their base, as if on the road to form an independent ossicle.

Superomarginal plates eighteen to twenty-four to a side; inferomarginals, two
more. The superomarginals are nearly square in the interbrachial are, the width
being slightly greater than length; distally the width greatly exceeds the length. A
little over half the exposed surface is usually bare, except in a large specimen from
Bering Island, in which the distal plates have a small bare spot. A single series
of large, squarish, close-set granules borders the inner and lateral edges of the bare
area, the outer portion of the plate being closely set with large rhomboid, trapezi-
form, and polygonal granules, those nearest the outer edge being the largest. The
inferomarginals are similar to the superomarginals, but the bare spot is smaller.

« Challenger Asteroidea, pl. 49, fig. 5.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 221

The adambulacral plates are very short and comparatively broad. The fur-
row spinelets (two or three in number) are short, thick, sometimes compressed or
three-sided in section, and round-tipped. The first actinal series of elongated
granules or short spinelets (one or two in number) are thick and intermediate in
size between the furrow spinelets and about four angular granules which occupy
the outer portion of plate. The two-jawed upright pedicellaria occurs on nearly
every plate in a specimen from station 3212 and from Bering Island, but are very
few in the Shumagin Island examples, and not numerous in that from Unalaska.

Mouth plates not prominent, the combined pair rhomboid in outline, with
straight margins. Armature similar in character to that of adambulacrals. Fur-
row series, seven or eight; about two series of granules (seven to nine in each)
occupy most of the surface of plates.

Madreporic body situated one-third to nearly one-half r from center. Stria-
tions interrupted, often in the form of perforations.

Type.—In U.S.National Museum.

Type-locality.— Bering Island (N. Grebnitsky).

Distribution —From Bering Island along the Aleutians to Kadiak Island, Alaska,
low tide to 102 fathoms.

Specimens examined.—Ten, from the following localities:

Specimens of Ceramaster articus examined.

Locality. | Depth.

Collector. Number. Collection.

Bering Island, Commander Islands...............-..- Rea Saas ose ae aeeeens N. Grebnitsky . ; 1 fuss U.S. Nat. Mus.
Round Island, Coal Harbor, Shumagin Islands dewase Beach « «sscests~sc200 ie WE. Dalle-= 2 Do.
Sanborn Harbor, Shumagin Islands.................. | Lowest water........-]..... OO eawaeca i 3 Do.
Unalaska and Unga, Aleutian Islands..............- Ze. anne eas |idishigge 2, 7 eee 1 Do.
AeabanitenlinivA neem: Seppe cel! eased bles 2d Loeet eae | W. J. Fisher... 2 Do.
1 Do

oe ee ae ag

Station 3212, off St. Paul, Kadiak.................... 102 fathoms, blue pul Albatross .......

Remarks.—This species is characterized by the few large granules of the abac-
tinal tabulate plates, the central granules being often lacking; by the very exten-
sive papular areas, comprising practically the whole of the abactinal surface; by
the imbricating abactinal plates which are strongly lobed on one side only; by the
short and comparatively wide adambulacral plates which have only two or three
thickened furrow spinelets.

Genus CLADASTER Verrill.

Cladaster VERRUL, Trans. Conn. Acad., vol. 10, 1899, p. 175, pl. 28, figs. 2,2a-2c. Type, C. rudis
Verrill.

Diagnosis.—In general appearance most closely resembling Hippasteria, but
differing in lacking secondary intermediate abactinal plates, prominent marginal
tubercles or spines, and abactinal tubercles or spines. Abactinal plates hexagonal
or roundish, with spaced granules and sometimes with spatulate pedicellariz ;
papulee single at the rounded corners of the plates; marginal plates large, few, and
granular, not increasing in size distad, and without spines or pedicellarise; actinal
interradial areas extensive with rather large angular, roundish or irregular plates
bearing spaced equal or unequal granules and spatulate pedicellariz ; adambulacral
armature similar to that of Hippasteria: two prominent furrow spines, a large
heavy subambulacral spine, and several granules.

222 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

“=

Remarks.—Cladaster is a puzzling genus. Abactinally it reminds one strongly
of some form allied to Tosia, as it has no prominent spines nor any intermediate
secondary plates. But the prominent adambulacral spines are strikingly similar
to those of Hippasteria. There are, however, no prominent marginal tubercles nor
bivalved pedicellarie, so that the similarity is not maintained to the ambitus.
Cladaster seems to be unmistakably allied to the Hippasteriine, but it lacks what
are believed to be important characteristics of that group, the secondary abactinal
plates and the tuberculate or spiny marginals.

CLADASTER VALIDUS Fisher.
Pl. 41, figs. 1, 2.
Cladaster validus Fisuer, Zool. Anz., vol. 35, March 29, 1910, p. 552.

Diagnosis —R=17 mm.; r=10 mm.; R=1.7 r. Arcuately pentagonal with
slightly convex abactinal surface ; broad marginals with tumid naked abactinal sur-
face; regular abactinal plates bearing spaced, deciduous obovoid granules and spatu-
late pedicellariz of small size; with two furrow spines and one larger, heavier clavate
subambulacral spine; with unequal actinal intermediate granules, and relatively
large broadly spatulate pedicellarie.

Description.—Abactinal plates regularly arranged in series parallel to the
medioradial, which fails to attain the terminal plate by the width of two supero-
marginals; from exterior the abactinal plates appear polygonal with rounded
corners, or roundish. Papule single, six about each plate (or five interradially),
absent only from a small interradial area adjacent to margin. Covering of plates,
relatively large obovoid or ovoid well-spaced granules, usually one or two in center
and six, eight, or more on periphery; some of the larger plates in center of disk
have four to seven central granules. Scattered over abactinal surface are about
seventeen two-jawed broadly spatulate pedicellariz which have no depressions into
which they fit when open. The jaws are one and a half to twice as high as
granules, and have curved ends.

Superomarginals six (or twelve to a side), and subequal except the last, which
is very small. They encroach on abactinal area conspicuously, having there a
quadrate contour, the proximal ones being nearer square than the distal; abactinal
surface of plates tumid and bare, except for a marginal series of spaced granules
similar to those of adjacent abactinal plates. Lateral surface of plates is covered
with spaced granules similar to latter but much larger.

Inferomarginals seven (or fourteen to side) and similar to superomarginals,
except that the actinal surface has a small and less regular bare space. The inner
edge of the plates has two rows of granules, where the superomarginals have one,
and these are much smaller than granules of lateral surface of plates, the latter
being just like the superomarginal granules. Terminal plate tumid and almost
globose; naked above, but granular on the distal face.

Actinal intermediate plates with coarse unequal granules, those on the outer
or marginal edge being larger than those on the side toward furrow or mouth angle.
Center of many of the plates bears a conspicuous two-jawed broadly spatulate
pedicellaria, similar to but much larger than the abactinal, or in lieu of this occa-
sionally a granule. The plates without pedicellariz are, however, smaller, and

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 223

there is little central space. The enlarged granules are globose, obovoid, ovoid,
or thimble-shaped. Pedicellariw# are higher than any of the granules.

Adambulacral plates with two blunt slightly compressed or cylindrical promi-
nent furrow spines which interlock with those across the narrow furrow, and behind
them a slightly longer but much heavier clavate, often swollen-tipped spine with a
small (adoral) companion granule. The tip of the subambulacral spine is often
obliquely dressed or bent toward furrow. Outer edge of plate has three or two
granules larger than those immediately adjacent on actinal intermediate plates.

Mouth plates with five furrow spines, of which the innermost is largest and
most compressed, and a large suboral situated at about the middle, behind which
are three smaller tubercular granules or spinelets, besides which about six unequal
granules border the median suture, the innermost being sometimes enlarged and
standing in line with the suboral.

Madreporic body small, triangular, with relatively coarse striz.

Type.—Cat. No. 27884, U.S.N.M.

Type-locality.— Albatross station 3480, Amukta Pass, Aleutian Islands (lat. 52°
06’ N.; long. 171° 45’ W.), 283 fathoms, black sand, coral, rocky (one specimen).

Distribution —Known only from the type-locality,

Remarks.—The only other described species of this genus is the type, C. rudis,
taken by the Albatross off Florida, at station 2415 (30° 44’ N.), 440 fathoms, coarse
sand and shells.

C. validus differs from C. rudis in having a definite tumid naked area on upper
surface of each superomarginal, abactinal pedicellarie, very unequal actinal inter-
mediate granules of larger size, larger actinal pedicellariw, and, relatively to the
abactinal granules, larger granules on lateral face of marginals. C. rudis has more
definite rays, the last four marginals being in contact medially, but this may vary
with age. The adambulacral spines seem to be relatively larger in validus, but it
is difficult to be sure of this point without direct comparison of specimens.

The type-specimen is probably young, but there are so many positive charac-
ters that it seems worth while to describe the species, especially as the genus is

little known.
Subfamily HIPPASTERIIN 22 Verrill, 1899.

Hippasteriine Verrmt, Trans. Conn. Acad., vol. 10, 1899, p. 174.—Fisuer, Bull. U. 8. Fish
Comm., 1903, pt. 3, 1906, p. 1165.
Anthencide PERRIER, in part, and most authors.

Genus HIPPASTERIA Gray.

Hippasteria Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 279. Type, H. europra Gray=
H. phrygiana (Parelius).

Diagnosis.—Goniasteride with large disk, short rays, bivalved pedicellari«,
one to three elongated thick furrow spines and only one or two similar actinal
adambulacral spines; with two sorts of abactinal plates: (1) Large roundish tumid
primary plates bearing a central granule, tubercle, spine, or pedicellaria, and a
peripheral series of granules, and (2) more numerous small, irregular, crowded
intermediate plates which join the larger plates, and bear externally granules,
pedicellariz, and sometimes small spines and tubercles. Marginal plates with

224 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
one to six tubercular spines, a peripheral series of granules, and occasionally bivalved
pedicellarie; general surface of marginals not covered with granules. Actinal inter-
mediate plates armed with pedicellariz and various-sized granules and tubercles.
The body surface is covered by a skin which, however, usually does not obscure the
plates or granules, and is typically only evident by the thin covering of the granules
and by wrinkles between the plates; exceptionally it is thick enough to obscure
the borders of the plates, which with the granules then appear immersed. Tube
feet with heavy sucking disks. No superambulacral plates.

KEY TO THE SPECIES OF HIPPASTERIA HEREIN DESCRIBED.

a', Marginal plates well developed, subquadrate, not separated by encroaching abactinal or actinal
intermediate plates; granules fairly or quite smooth.

b!. With conspicuous abactinal spines and abactinal surface of marginals not smooth and unarmed,
c!. Pedicellarize shorter and higher; dorsal surface very spiny; as a rule no pedicellarize on mar-
ginal plates; actinal pedicellarize with oblong or subquadrate jaws; actinal intermediate
plates not conspicuously tuberculate; abactinal granules in definite marginal series about the
plates, abactinal skeleton not fenestrated....-......---------------+-------- spinosa, p. 224.
c2, Pedicellarize low, long; a well developed pedicellaria on the proximal superomarginals, and
on most of the inferomarginal plates; papule conspicuous, bag-like; actinal intermediate
plates tuberculate; actinal pedicellarize very low, long; abactinal granules small conical,
scattered, abactinal membrane thick; abactinal skeleton when viewed from inner side honey-
combed\or fenestrated 3° 5.2 et -asaee ee ac eee eae ee eee eee eee heathi, p. 231.
b?. No abactinal spines or tubercles (or only exceptionally and abnormally present, and then very
few); abactinal plates mostly smooth, bordered by a single row of stout granules; marginals with
all except lateral surface bare; lateral spines small and unequal, one or two to a plate or absent;
adambulacral spines two, clavate, in a transverse series............-.-.------ leiopelta, p. 227.
a?. Marginal plates weak, irregular, oval, or elliptical, the proximal usually separated by encroaching
plates from the dorsal and ventral surfaces; actinal pedicellarie high, rather delicate, flaring at
base, and with narrow curved serrate tips; granules rugose or denticulate....californica, p. 233.

HIPPASTERIA SPINOSA Verrill.
Pl. 42, figs. 1-3; pl. 43, figs. 1, 2; pl. 60, fig. 4.

Hippasteria spinosa VERRILL, Amer. Journ. Sci., vol. 28, July, 1909, p. 63.

Diagnosis.—Similar to H. phrygiana of the north Atlantic, but primary abac-
tinal and marginal plates with prominent, often long, tapering blunt spines rather
than the elongate tubercles of phrygiana.

Description.—Rays five. R=110mm.;r=56mm.;R=2r. Breadth of ray at
base,64mm. General form same as phrygiana, but rays a trifle broader. Majority of
the larger or primary abactinal plates convex and bearing in center a stout, upright,
rigid, tapering truncate spine, one to two times width of its plate in length; mid-
radial spines slightly the longest, thence decreasing in length to margin of area;
same plates in phrygiana bearing a very much shorter, stouter tubercle, truncate
or rounded at apex. General surface of plates smooth, but rim encircled by a
single series of irregular, small, subconical or roundish granules heavily invested with
membrane. Many large and smaller plates bear a large central bivalved pedicel-
laria, the jaws of which are only slightly wider than high; sometimes the two
dimensions are equal. These pedicellarie are usually narrower and higher than
those of phrygiana and have thinner blades. Distal edge of jaws is slightly curved,
either serrate or smooth. Small intermediate plates intercalated between the

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 225

larger ones bear one to three or four conical skin-covered granules, or a small cen-
tral spinelet or pedicellaria, surrounded by irregular granules, depending upou
size of plate. As in phrygiana, the surface of granules is smooth, and the free
edge is rounded, or the whole granule may be subconical and pointed.

Marginal plates slightly tumid as in phrygiana. Superomarginals sixteen or
seventeen to ray, corresponding with inferomarginals except on distal part of
ray. Plates of both series surrounded by a single row of squarish blunt or
flattened skin-covered granules, which in one specimen are more pinched with
rounded top. Proximal eight or nine plates of each series with two (less com-
monly only one, or as many as three) rigid spines, similar to abactinal spines, but
usually a little larger; remainder of plates with one such spine. In phrygiana
marginal plates have only robust tubercles. Occasionally a two-jawed pedicel-
laria occurs on marginal plates, but widely scattered along the series.

Actinal intermediate plates arranged as in phrygiana, but pedicellarie with
higher jaws which sometimes taper to a narrow tip; these pedicellarie frequently
have serrate jaws, especially in southern examples. Granules pointed, thickly
covered with membrane. Tubercles variable, subtruncate, thimble-shaped, or
subconical and sharp, especially in southern examples.

Adambulacral spines arranged practically as in phrygiana, but longer and
stouter. Actinal spine at least as long as width of plate and subequal to but much
stouter than the two furrow spines, which are slightly tapering and compressed at
tip. Frequently a pedicellaria stands on the plate.

Mouth plates with four or five furrow spines slightly larger than those of
adambulacral plates.

Color in life-—A specimen from off Point Pinos, California, was bright scarlet
vermilion when taken from the water; this color disappeared almost at once in
alcohol.

Anatomical notes—No rudimentary superambulacral plates. Intestinal co-
cum large, composed of about four radiating branches which immediately sub-
divide irregularly into numerous slender tubes reaching nearly to margin. Gonad,
a thick tuft of branching tubules on either side of the interradial septum, about
one-third r from margin. Stomach capacious, the thick hepatic coeca reaching into
ray about one-half R.

Variations.—The chief variations concern the length of the abactinal spines
and the height of the pedicellarie. A specimen from station 4292, vicinity of Kadiak
Island, and three from station 3080, coast of Washington, have considerably shorter
abactinal spines than all the others, but they differ still from typical phrygiana in hay-
ing prominent marginal spines, longer adambulacral spines, and higher pedicellarie.
The actinal pedicellarie are the most variable, in some cases tending toward the
form characteristic of californica. The jaws have either smooth or serrate edges.

In Californian specimens the spines are all very prominent, more so than in the
northern examples, and the actinal pedicellariw are plainly serrate. The Monterey
Bay (station 4552) example has more numerous marginal spines than the example
from off southern California (station 3664). In both, the rays are notably stouter
and shorter than in typical phrygiana. The specimen from station 4552 has the

57444°—Bull, 76—11 15

226 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

“as

following characters: R=71 mm.; r=40 mm.; R=1.75r. Breadth of ray at base,
48mm. Height of longest abactinal spines,6 mm. Disk very rigid, marginal plates
massive, the upper with proximally two or three spines as long as those of abactinal
plates, distally with one or two; more of the inferomarginals have three spines, occa-
sionally four. Marginal spines of both series stand in line, forming a perpendicular
series of six in interradial are (sometimes seven), then five, four, three, and finally
two. Pedicellaris numerous, the jaws slightly wider than high; actinal a trifle
higher than abactinal; none on marginal plates. Actinal intermediate plates
with short subconical spines forming a group of four or five (in addition to the
granules), or they may surround the pedicellaria when that is present. They are
equal to or only slightly higher than pedicellarie. All granules skin-covered,
depressed or conical, but not denticulate or rugose as in H. californica. Color in
life, a bright scarlet vermilion.

Ty pe-locality —Departure Bay, British Columbia, 18 fathoms.

Distribution —From southern California to Alaska (Kadiak Island) in 27 to
121 fathoms.

Specimens examined.—Kighteen specimens from the following stations:

Specimens of Hippasteria spinosa examined.

Station. Locality. | Depth. Nature of bottom. Number. Collection.
|
| Fathoms.
38 gray. sand... - oc cncn ene 1 U.S. Nat. Mus.
sone | 27 rocks and shells.....-..-. 1 Do.
Q050 cesses Oft Siletz Bay, Oregons a )ajerectsetssise wae serene | 77 mud... Seeeeeteeee ee 1 Do.
S0G0 scone Off Tillamook Rock, Oregon........-.-------- | 28 brown saad Renee ete 1 Do.
S080 Ee soos Off Heceta Bank, Oregon..........--..-.----- 93 BreeNIMNUG ore stetste eee sa 4 Do.
3445 Straits of Fuca, Washington.............---- 100 TOCKYy $52). tyadtecseeecceee 1 Do.
3500......| South of St. George Island, Bering Sea. . 121 fine gray sand.......-.-. 1 Do.
3664 Near Santa Catalina Island, California 80/5 eee Gone Beeson 1 Do.
MBSTe cose Near Yes Bay, Behm Canal, Alaska. - - --| 380-45 soft gray mud.......-..-- 2 Albatross, 1903
PAS eee aol Kasaan Bay, Prince of Wales Island, Alaska. - .| 42-47 green\mud se--==256--- 2 Do.
4202... <2 | Shelikof Strait, near Kadiak Island, Alaska..-. 66 graymud)-2S2==2625-254: 1 Do.
Aongen re |\Shelikot Stralter sastes oo tee a 92 | soft gray mud.......----- 1 Do.
4552...... Off Point Pinos, Monterey Bay, California..... 73-66 green mud, rocks. -...---. 1 Albatross, 1904.

HIPPASTERIA SPINOSA KURILENSIS, new subspecies.
Pl. 43, fig. 2; pl. 121, fig. 1.

Diagnosis.—Similar to H. spinosa, but with shorter rays, and relatively larger
pedicellarie ; abactinal granules typically pointed; margin massive, spinous; small
pincer-shaped marginal pedicellariz; adambulacral plates with two unequal com-
pressed furrow spines or only one; one prominent subambulacral spine shorter than
the longer furrow spine. R=55 mm.; r=32 mm.; R=1.7 r.

Description.—Abactinal surface spiny, the prominent spines being more coni-
cal and pointed than in typical spinosa, where the spines are usually subeylindrical
with a blunt or truncate tip. Granules in a single series surrounding plates and
ending in an abrupt usually sharp point, instead of being rounded. Abactinal
pedicellarie, numerous, large, as high or higher than broad, and wider at top than
at base; edge of jaw undulating or slightly denticulate. Papule numerous, all over
abactinal surface except a very narrow interradial band. Superomarginals tumid,

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 227

ten to twelve to each ray, each plate with two, and inferomarginals with one or two
stout conical spines, and in addition a single row of pointed marginal granules. Pedi-
cellariz, higher than wide and with denticulate jaws smaller than those of abactinal
surface, are scattered here and there, usually on the lateral face of the plates. Such
pedicellarize do not occur in typical spinosa.

Actinal surface with numerous pedicellariz having higher jaws than in leiopelta.
A prominent series occurs on the row of intermediate plates adjacent to the adam-
bulacrals. The dimensions of these pedicellariz are variable in typical spinosa.
The actinal granules are similar in distribution to those of spinosa, but often (though
not always) have several points, or only one point, rather than a simple low conical
form. They are, therefore, somewhat similar to those of H. californica. Many of
the plates have one or two low thimble-shaped tubercles in the center, in place of
pedicellariz.

Usually the adoral furrow spine is the shorter of the two, and both are strongly
eompressed at the blunt tip. The adoral spinelet is sometimes wanting on the
outer part, or along most of the ray. Actinal spine shorter than the longer furrow
spine and usually compressed at the base in a plane parallel to furrow. The outer
part of plate is covered with granules. Mouth spines four, strongly compressed,
chisel-shaped; usually but one suboral standing near margin, and in line with
the marginal spines.

Type.—Cat. No. 27885, U.S.N.M.

Type-locality.— Albatross station 4804, off Simushir, Kuril Islands, 229 fathoms,
coarse pebbles, black sand.

Distribution.—Known only from the vicinity of Simushir.

Specimens examined.—Eight; from type locality four, and from station 4803,
same locality, depth, and bottom, four. (Albatross, 1906.)

Remarks.—This race differs considerably in general appearance from typical
spinosa of the North American coast, as the photograph of the type-specimen will
reveal. The actual differences are not great, however, the most striking being the
form of the granules, the marginal pedicellariz, and adambulacral armature.

Along with these specimens were taken one small and four large examples of #7.
leiopelta armata. Three specimens listed as kurilensis are possibly hybrids of this
race and the typical form of kurilensis.

The differences between leiopelta and kurilensis are at once apparent on a com-
parison of figures. The latter has much higher pedicellarie, heavily spined abacti-
nals and marginals, marginal pedicellariz, and compressed, rather spatulate furrow
spines, not stumpy conical ones. The naked surface of the marginals in leiopelta
is also quite different from the condition of kurilensis.

HIPPASTERIA LEIOPELTA Fisher.
Pl. 46, figs. 1-4; pl. 58, fig. 6.
Hippasteria leiopelta Fisner, Zool. Anz., vol. 35, March 29, 1910, p. 553.
Diagnosis.—Rays five. R=30 mm.; r=20 mm.; R=1.5 r. General form
stellato-pentagonal to arcuate pentagonal, the latter being usual for small speci-

mens, which have R=about 1.45 r. The longest-rayed specimen has R=31.5 mm.;
r=18 mm.; R=1.7+r. Abactinal plates smooth in the middle surrounded by

928 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

a single series of granules, and bearing besides occasionally a central granule,
rarely a tubercle, and (interradially) here and there a large low bivalved pedicel-
laria. Marginal plates regular, large, dorsally and ventrally naked except for a
peripheral series of granules; a cluster of granules with sometimes a spine or tubercle
on outer face; no marginal pedicellarie; scattered large bivalved actinal pedicel-
larie; actinal granules large; adambulacral spines two, large and clavate in a trans-
verse row on plate, the furrow spine slightly the larger.

Description.—Abactinal plates large and small intermingled, the larger or pri-
mary plates not arranged in very regular order, the median radial coming nearest
to forming a series; and that only on outer part of major radius. Periphery of
plates surrounded by a single series of rather large ovate, acorn-shaped, or com-
pressed granules blunt or pointed and of unequal size. The surface of the plate is
smooth and flat, but occasionally the center of a primary plate has one or two
thimble- shaped eranules, those on the radial series very rarely enlarged into short
tubercular spines. A specimen more often has no abactinal tubercles, or only one,
two, or three. The specimen photographed has more than any other. Scattered
over the dorsum chiefly on the interradial areas is a variable number (fifteen
to sixty-five) of low bivalved pedicellarize occupying whole width of plate, the
jaws lower than the combined breadth of the two jaws. The flat or very gently
convex bare surface of the abactinal plates (except those bearing pedicellarie# or
an occasional central granule) is one of the most characteristic ene of this
species.

Marginal plates conspicuous, encroaching conspicuously upon both surfaces.
Superomarginals six to nine; inferomarginals one more. Most of the superomargi-
nals wider than long, the abactinal surface shghtly tumid and bare, with a periph-
eral series of granules similar to those of neighboring abactinal plates but usually
more compressed and wider than high. Actinal surface of inferomarginals similar
but the naked surface a little larger and less convex and the granules even more
regular. Outer face of both series covered with unequal ovoid, subglobose, round-
tipped or pointed tubercular granules larger than those of dorsum. The granular
surface of superomarginals larger (that is, higher) than is that of inferomarginals;
one to four granules are enlarged, and frequently one or less often two are
developed into short blunt tubercular spines, the superomarginal spines being the
longer. There is great irregularity in the occurrence of these spines which are more
numerous, irrespective of age, on some specimens than on others. Near the end
of ray there are fewer granules, and the plates frequently bear only the peripheral
series and a tubercle with several granules near the aboral edge, or only one to
several granules. Terminal plate subglobose fairly large with or without granules
and one tubercle.

Actinal interradial plates irregular with large unequal peripheral granules,
those on the edge toward marginal plates being largest as a rule. In each inter-
radial area adjacent to adambulac -rals are from two to nine long low bivalved pedicel-
larize, with an undulating smooth or sparsely serrate edge to jaws. These are longer
but not higher than the dorsal pedicellarie, the largest as a rule being near the
base of furrow.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 229

Adambulacral plates with typically one prominent club-shaped furrow spine
about as long as width of plate and behind it on the surface of plate another similar
but slightly shorter one, with two to five additional granules along outer edge, or
outer half of plate. Proximally a small adoral companion furrow spine is sometimes
present, or more rarely two equal spines. The first plate may bear a pedicellaria
on its surface, in which case there are two furrow and no subambulacral spines,
but only the marginal granules. The spines are sometimes not so markedly clavate,
but more nearly cylindrical or even slightly tapered.

Mouth plates with three to five spines along furrow margin, continued to outer
end by three or four granules; several granules and sometimes a tubercular spinelet
stand along the suture margin.

All the abactinal, marginal, and actinal spines and granules are immersed in a
thin membrane which is not at all apparent on the surface of the plates and does not
in any way obscure the outlines of plates. This disappears when specimens are dried.

Madreporic body small, midway between center of disk and inner edge of mar-
ginals; striations few, coarse, and radiating.

Color in life, vermilion or orange vermilion apparently, as specimens when
received were so tinted.

Anatomical notes.—Intestinal caecum large, consisting of three principal
slightly lobed divisions lying in the trivium and a short lobe lying in the left bivium
radius. Gonad in an interradial tuft on either side of membranous interbrachial
partition.

An examination of the dorsal plates from the inner or ceelomic surface shows
that they are similar in arrangement to those of H. spinosa and are fitted together
even more closely. The larger and smaller plates are everywhere intermingled, but
on outer part of ray the secondary plates form more definite connecting pieces
between the primary plates. The large primary plates are polygonal with blunt
corners or irregular, the secondary plates are elongate to roundish. Papul single
and everywhere except a small interradial area near border. Inner surface of
skeleton fairly even, without the honeycombed structure of H. heathi.

Type.—Cat. No. 27886, U.S.N.M.

Ty pe-locality.— Albatross station 4795, off Kamchatka, south of Petropavloysk
(lat. 52° 46’ 50’’ N.; long. 158° 44’ 30’’ E.), 69 to 48 fathoms, green sand, pebbles.

Distribution.—Sea of Okhotsk to southern Bering Sea.

Specimens examined.—Twenty-eight; from type-locality, sixteen; from station
4794, same locality, 58 to 69 fathoms, sand, pebbles, eleven; from station 5024,
off eastern coast of Sakhalin, vicinity of Cape Patience, Okhotsk Sea, 67 fathoms,
sand, pebbles, one. (Albatross, 1906.)

Remarks.—This well-marked species differs from phrygiana and allies in the
absence of abactinal spines or tubercles, the center of the plates being usually
smooth and bare except for an occasional granule or pedicellaria. Likewise, the
marginal plates lack the characteristic spines of phrygiana, although spines may be
present on the lateral face of the plates. The greater part of the surface of the
marginals is bare. The adambulacral spines are heavy and reduced to two, forming
a transverse series.

230 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

HIPPASTERIA LEIOPELTA ARMATA, new subspecies
Pl. 121, fig. 2; pl. 122, figs. 1, 2.

Diagnosis.—Intermediate in general appearance between leiopelta and phrygiana
but nearer the former; differing from leiopelta in having one, and sometimes un the
interbrachial ares two, stout rigid conical spines to each marginal plate, and in
having much more numerous, and higher, abactinal pedicellariz ; actinal pedicellarie
shorter and higher; adambulacral spines longer, the single furrow spine tapering
and blunt, not clavate, as in leiopelta. R=73 mm.; r=42 mm.; R=1.7+r.

Description.—Abactinal plates bear more numerous pedicellariz than typical
leiopelta, and the jaws are relatively higher in proportion to the length, about as in
spinosa and higher than in phrygiana. The median radial plates bear a low thimble-
shaped tubercle, scarcely higher and often lower than width of plate. Many of
the other abactinal plates bear one or two central granules which may become
enlarged into a low tubercle, never conspicuous. In leiopelta such tubercles occur
only rarely or are entirely absent, the surface of the plate being smooth. A single
series of smooth granules surrounds each plate. Papulsx, as in leiopelta, all over
abactinal surface except a narrow interradial line.

Marginal plates more tumid than in leiopelta. On the lateral face on a raised
boss is a stout conical rigid spine, and sometimes in the interbrachial arc a second
occurs in a transverse line. However, most of the abactinal surface of the super-
omarginals and the corresponding face of the inferomarginal is bare. Scattered
bivalved pedicellariz# smaller than those of abactinal plates also occur in the inter-
brachial are. These are not present in leiopelta.

The actinal pedicellariz are shorter and higher than in leiopelta, as well as
much more numerous. In leiopelta the pedicellariz have the appearance of being
immersed in the granulation, while in armata they stand well above it (as in spinosa).
The adambulacral spines (one furrow and one actinal) are longer and relatively
slenderer than in Jezopelta and taper slightly to a very blunt tip. In leiopelta they
are thick and often clavate. The actinal spine is the shorter. Occasionally an
adoral spinelet is present on the furrow series, or merely a granule. Irregular
granules cover the remainder of the plate.

Type.—Cat. No. 27887, U.S.N.M.

Ty pe-locality.—Albatross station 4804, off Simushir, Kuril Islands, 229 fathoms,
coarse pebbles, black sand.

Distribution.—Known only from the type-locality.

Specimens examined.—Five; from the type-locality, three; from station 4803,
same locality, depth, and bottom, two. (Albatross, 1906.)

Remarks.—A small specimen with R 32 mm. and hence about the same size
as the type of leiopelta has the characters of the race as well marked as the type, with
the exception of the marginal pedicellaria. This shows that armata is not simply
a fully grown leiopelta. (Pl. 122, fig. 1.)

Armata differs from kurilensis in lacking the conspicuous abactinal spines and
pointed granules, as well as in the armature of the adambulacrals. It differs from
typical phrygiana in being more pentagonal in shape, with notably shorter rays,
in haying only one furrow spine, higher pedicellaria, and more massive marginals

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 231

bearing pedicellarix and longer, sharper spines. The unarmed surface of the mar-
ginal plates is much more extensive in armata.

In some respects armata may be considered as an intermediate form between
phrygiana and leiopelta. It is certainly much nearer the latter, however, and per-
haps should be ranked as a full species. It shows such obvious resemblances to
letopelta, which furthermore exhibits certain characters of armata (as, for instance,
the marginal spines) in a rudimentary condition, that a subspecific designation is
believed to best indicate the relationship.

The relationship between armata and kurilensis is not clear. They both range
together at stations 4803 and 4804, and several specimens of kurilensis seem to have
an infusion of armata blood, as if hybridism had occurred. These intermediates
are nearer kurilensis, however, and I do not think that they are genuine intergrading
forms between kurilensis and armata. The possibility of leiopelta being connected
with spinosa through this series must be considered, however improbable it may
seem from the great difference of the extreme forms. Future exploration will prob-
ably tend to multiply rather than lessen the number of species of Hippasteria in the
North Pacific.

HIPPASTERIA HEATHI Fisher.
Pl. 44, figs. 1, 27 pl. 58, figs. 5, 5a; pl. 60, fig. 6.
Hippasteria heathi Fisner, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 309.

Diagnosis.—Rays five. R=78 mm.; r=39 mm.; R=2 r. Breadth of ray at
base, 42 mm. General form robust; disk large, rays broad, evenly tapered from
a wide base to a blunt recurved extremity. Abactinal surface considerably inflated;
a well-defined interradial sulcus leading from marginal plates three-fourths of
distance to center of disk. Interbrachial ares wide, shallow. Abactinal surface
covered with robust widely spaced conical spines and more numerous large, low,
bivalved pedicellarie; abactinal granules unequal, small, scattered, immersed in a
soft meMbrane that obscures entirely the outlines of plates; abactinal skeleton
fenestrated; papule all over abactinal surface large; marginal plates small, most
of them with a large bivalved pedicellaria, one to four prominent conical spines
and unequal conical granules in one or more series about the margin. Adambulacral
plates with two stout blunt furrow and two actinal spines. Actinal intermediate
areas with unequal, irregular plates bearing large bivalved pedicellarie# and stout
tubercular spines and granules.

Description.—Abactinal surface beset with widely spaced, robust, rigid, taper-
ing, upright, bluntly pointed spines, 3.5 or 4 mm. in length; one, or rarely two,
to the larger rather widely separated primary plates; the median radial and either
adradial series of spines extending to tip of ray, but very irregularly; the other
spines too irregularly distributed to form rows. Scattered all over abactinal
surface are many sessile, large, low bivalved pedicellarie with smooth edges to the
jaws. These pedicellarie (which have a base 1.5 to 3.5 mm. in length, a diameter
of 1.5 mm., and a height a trifle less than 1 mm.) are more numerous than the
conical spines, and are borne on the primary abactinal plates, usually occupying
all the elevated surface. A comparatively few small ones are borne on the larger
secondary plates. The primary plates are not well defined and circular with a

232 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

marginal series of granules as in the two preceding species, but are elliptical as a
rule (most of the spine-bearing plates are subcircular) and immersed in a soft mem-
brane, from which the spines, pedicellarie, and elevated portions of the plate
emerge. The small unequal, conical, subspherical, and acorn-shaped granules are
immersed in this membrane, and are scattered irregularly and rather sparsely over
the exposed portions of the plates and intermediate ossicles. On some of the
secondary plates a granule assumes the proportions of a small acorn-shaped tubercle.

When the abactinal skeleton is. viewed from the inner or celomic surface, it
presents a very characteristic appearance. The primary plates are connected by
irregularly radiating ridges composed of ossicles placed end to end. The bottom
of the intermediate ossicles or plates is much compressed, so that the ridges formed
by them are narrow. The height of the ossicles causes the enclosed space to appear
as a pit, the skeleton assuming a honeycombed or fenestrated appearance. The
meshes of this skeleton are further subdivided by a few smaller low trabecule
near the surface. These separate the papule, are incomplete, and do not interfere
with the honeycombed structure mentioned ahove. (See PI. 60, fig. 6). The papule
are very conspicuous in the undried state (2 mm. long) and are distributed all over
the abactinal surface.

Marginal plates not particularly conspicuous. Superomarginals, fifteen in
number from median interradial line, slightly convex, and confined to side of body;
first plate larger than the rest, which are rather longer than high except at end of
ray, where the reverse is true; upper edge of the series largely obscured by abactinal
membrane. Each plate on proximal half of ray (first to fifth or sixth) bears a
sessile bivalved pedicellaria and in addition one to four spines like those of the
abactinal surface. Edge of plate (excepting usually the upper) is armed with one
or two interrupted series of stout conical granules of various sizes. Outer plates
of series have a single central spine and marginal granules in one or two irregular
series. Inferomarginals corresponding in number to superomarginals and oppo-
site them in proximal third of ray, but alternate on outer part; they encroach some-
what upon the actinal area, the lower margin being obscured on the disk by the
membrane covering the general surface of the plates. Nearly all the plates bear
a bivalved pedicellaria and, in addition, on the first two plates of the series a circle
of four or five spines (like the superomarginal spines) surrounding it, on the rest
two or three, which are situated on the aboral side of the plate. There are also on
the border of each plate numerous stout conical tubercular granules mixed with
small ones, the former occasionally having one or two enlarged into conical spinules,
intermediate in size between the spines and larger granules. The pedicellariz of
the marginal plates are exactly like those of the abactinal surface.

Adambulacral armature, consisting of two heavy cylindrical or shghtly com-
pressed blunt furrow spines, the adoral a trifle the smaller on outer half of ray; and
on actinal surface a transverse series of two similar spines rather more tapered, the
outer the shorter, occupying all the exposed surface, the outer spine reduced in size
on the first few plates. Margin of plate surrounded by conical granules, one or two
on the outer edge larger than the rest.

Mouth plates small, narrow; furrow margin very extensive; furrow spines four,
similar to those of adambulacral plates, the innermost slightly longer and most

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 233

compressed; on actinal surface opposite the outer furrow spine one compressed
spine nearly as long, and on the inner part of the plate usually another, considerably
shorter, connected with the first by a linear series of several spaced, unequal, conical
granules, which are continued beyond the outer spine to the end of the plate, there
forming a group rather than a row.

Actinal interradial areas large, three series of intermediate plates extending
to the fourth inferomarginal, two to the fifth, and one to the ninth; each of the
plates adjacent to the adambulacrals bears a large, sessile, bivalved pedicellaria,
4 mm. in length, similar to those of dorsal surface, placed usually obliquely cross-
wise. Numerous other interradial plates also have a pedicellaria, usually some-
what smaller. Plates also armed with one or two conical spinules or tubercles
standing on the edge, often on either side of the pedicellaria, in line with a peripheral
series of large and small hemispherical or acorn-shaped granules. Plates without
pedicellarix bear one or two unequal thimble-shaped spines surrounded by a
peripheral series of small granules, with two or three here and there larger than
the others. All the actinal interradial spines are shorter than the marginal.

Madreporic body small, circular, situated slightly nearer center than midway
to margin; striations fine, ridges rather wide.

Type.—Cat. No. 22338, U.S.N.M.

Type-locality.— Albatross station 4239, Clarence Straits, Alaska, 206 to 248
fathoms, coarse sand, rocky. (Cruise of 1903.)

Distribution —Known only from the type-locality.

Specimen examined.—The unique type.

Remarks.—This species differs from all the preceding, as well as from phrygiana,
in having the abactinal granules, which are small and more or less conical, scattered
over the plates and not arranged in a definite marginal series; by the preponderance
of the large low bivalved pedicellariw over the spines; by the characteristic form of
the abactinal skeleton and the soft investing membrane; in having large and
numerous marginal pedicellarie; in the presence of two actinal adambulacral
spines instead of one; in having conspicuous stout actinal spinelets. The pedi-
cellariz on heathi are very large and have lower jaws than in any other species of
Hippasteria. The abactinal spines are also slightly different, being thicker and more
conical, and the papule are very conspicuous, being distributed all over the abactinal
surface. The scattered granules and rather thick soft abactinal membrane make
it impossible to distinguish the outlines of the plates, even in the dried condition.
No such difficulty is encountered in other species of Hippasteria.

HIPPASTERIA CALIFORNICA Fisher.
Pl. 45, figs. 1-4; pl. 60, fig. 5; pl. 111, fig. 1.
Hippasteria californica Fisner, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 310.

Diagnosis —Disk broad; rays broad and rather short, tapering from a broad
base to a bluntly pointed extremity; interbrachial ares very wide, shallow, and
rounded; abactinal surface subject to inflation; disk thinner than in phrygiana.
Abactinal surface spiny, as in spinosa, but typically with fewer spines. Marginal
plates small, oblong-elliptical, often separated by encroaching abactinal and actinal
intermediate plates, typically with one conspicuous spine. Adambulacral plates

934 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

with one furrow and one actinal spine and, usually, also a broad fairly high serrate
pedicellaria. Abactinal pedicellariee with quadrate rounded serrate jaws, usually
wider at tip than base and either wider than high, or the reverse; actinal pedicel-
lari with a broad base, tapering abruptly to a narrow, more or less truncate,
incurved, serrate margin; jaws usually higher than wide. All granules denticulate,
or more or less rugose, as if covered with minute unequal points and confluent
knobs. R=about 2r. Type, R=73 mm.; r=35 mm.; breadth of ray at base,
40 mm. Another specimen: R=100 mm.; r=55 mm.; breadth of ray at base
64 mm.

Description.—Abactinal surface beset with numerous rigid, upright, tapering,
sharp, or bluntly pointed spines, which are variable in length and number, but are
usually about as long as in spinosa* (3.5 to 5 mm.). These spines, which are
borne on a boss or convexity of the primary abactinal plates, are nearly twice as
numerous in some specimens as in the type. There is a median radial series of
about twelve to over twenty spines, and on either side of these two to five very
irregular longitudinal series of shorter spines. In the especially spiny individuals
some of the secondary plates have short spinules. On a few small specimens the
spines are all very short. Abactinal pedicellariz numerous, usually found on the
secondary plates, of divers sizes, according to size of plate, typically with broadly
spatulate denticulate jaws, which, however, are quite variable in shape. The jaws
may be nearly square, or higher than wide, rarely wider than high, strongly curved
to truncate at the biting edge. The teeth are also very variable in length. In
the type they are conspicuous and slender, but in a specimen from station 4530
(otherwise also not typical) the edges of the jaw are smooth (the actinal pedicel-
larie are toothed, however). The spaced granules, which border all the slightly
raised plates, instead of being smooth, as in phrygiana and spinosa, are beset with
tiny prominences, giving them a conspicuously rugose or denticulate appearance;
in extreme cases multifid. The membrane investing the surface of body is thinner
than in spinosa, so that even before the specimen is dried the irregularities of the
granules are easily detected.

Abactinal skeleton more open than in spinosa. The primary plates are
roundish or slightly lobed and well separated. The much more numerous, irregu-
larly disposed, trilobed, roundish, oblong, and very unequal secondary plates form
a network between the primary plates by touching or overlapping each other. In
those specimens with especially weak marginals the abactinal surface is also thinner
and weaker. This is caused by the fact that the larger secondary plates are only
loosely joined, leaving comparatively wide meshes, in which are very small inde-
pendent (that is, separate) grain-like ossicles or rudimentary plates. Papule
numerous, often conspicuous, emerging between the plates singly—that is, a papula
is usually separated from its neighbor by an ossicle, sometimes very small. In
large specimens the papule extend all over the abactinal surface; in smaller speci-
mens there is a small interradial area near margin free from them.

Marginal plates small and rather thin, irregularly oval or elliptical, longer than
high, except at end of ray, where the two dimensions are nearly equal. Supero-

@ Jn the original description this species was referred to as phrygiana; of course phrygiana, having
short spines or tubercles, is not at all similar to californica.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 235

marginals, twelve to seventeen in number from interradial line to extremity of ray,
confined entirely to side of body, and rather inconspicuous; very irregular as to
shape; successive plates frequently separated by small intermediate encroaching
abactinal plates, each plate with a tapering blunt spine articulated rigidly to a cen-
tral boss and a marginal series of rugose granules, there being also one or two
smooth subconical granules on the general surface, occasionally, also, a pedicel-
laria just beneath the spine. Inferomarginals very similar to superomarginals, of
the same number, but not opposite to them; similarly armed; but the pedicellarie
very rare, actinal intermediate plates encroaching and separating most of the
proximal inferomarginals. The marginal plates are very characteristic in this spe-
cies, being remarkably small in most specimens and usually more or less irregular
in shape and disposition. In many cases the actinal and abactinal plates are inter-
calated between a few or many of the marginals. In one extreme case (station
3344) several superomarginals are separated by two abactinal plates in longitudinal
series, so that normal superomarginals seem to alternate with two very much
smaller ones. In young specimens this encroachment does not take place. A
curious variation is presented by a specimen (which is otherwise not typical) from
station 4530, where there are several intermarginal plates on each ray. These are
simply abactinal plates, around which the marginal plates have grown, inclosing
and segregating them, as it were. Exceptionally, the marginal plates bear two or
even three spines. One spine is more usual, however, and in a specimen with
R=113 mm., is 4 to 5 mm. long, measured from the general level of the plate. In
large specimens pedicellariz are rare on the marginals.

Adambulacral armature, consisting of (1) a single blunt spine, usually consid-
erably compressed and somewhat tapering, situated on the angular furrow margin,
with occasionally a shorter similar adoral companion; or, one or two very short
spinelets or compressed granules on either side of the single spine; (2) on the actinal
surface a more robust, upright, tapering, blunt spine, situated just external to the
furrow spine, and just adoral to this there is frequently a medium-sized pedicellaria,
one jaw of which is broad serrate and rather deeply scoop-shaped, closing over the
other, which is smaller, thicker, and not hollowed out; margins of plates bordered
by numerous peculiar rugose granules, the general surface of the plate being covered
with the same. There is some variation in the thickness of the spines. In a large
specimen from station 3344 these are swollen and blunt, and between the furrow
and actinal spine there is usually a large bivalved pedicellaria considerably lower
and broader than in the type.

Actinal intermediate areas large, paved with unequal irregularly roundish,
polygonal, or elliptical plates, two series of which extend three-fourths the length
of ray, and a single series nearly to tip. Most of the plates bear a central high
pedicellaria of a peculiar shape, so far as this genus is concerned; jaws rather thin
and wide at base, abruptly narrowing into the distal portion, which bends toward
the other jaw, and has a truncate serrated edge. This base varies in width, but is
commonly one-half to three-fourths as wide as the plate. The pedicellarie are
wider than high to slightly higher than wide. Plates are bordered with a series of
rugose or denticulate granules, with several of the same size on general surface,
where there are, besides these, a number of considerably larger, smoother, subcon-

236 BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

ical, or cylindrical granules, immersed like the rest in thin membrane. On some
specimens these are almost wanting; on others they are prominent and compressed,
sometimes subtubercular and thimble-shaped.

Mouth plates prominent actinally; furrow spines five or six, considerably com-
pressed, the inner spines the larger and heavier; on actinal surface an incomplete
series of three spines, parallel with the furrow, usually one near the inner angle and
one or two near the outer end of the furrow series; remainder of surface covered
with rugose granules, often curiously compressed or pinched, with denticulate edges,
these forming a row on the median suture margin and another adjacent to first
adambulacral plate.

Madreporic body variable in size, ranging in two large equal-sized specimens
from 4.5 to8mm. in diameter; situated nearly midway between center and margin
of disk; striations radiating and often interrupted.

Color in life coral red. A large aberrant specimen from station 4530 was col-
ored as follows: Abactinal surface salmon orange; the marginal plates and actinal
surface yellowish buff.

Anatomical notes.—The inner organization is very simular to that of H. spinosa.
The intestinal ececum is, however, much smaller, consisting of short-branched radi-
ating, irregular, and unequal tubes. There are five Polian vesicles.

Type.—Cat. No. 22339, U.S.N.M.

Type-locality.— Albatross station 4429, off Santa Cruz Island, California, 680
fathoms, green mud, black pebbles, shells.

Distribution.—Southern California to Washington, 266 to 847 fathoms.

Specimens examined.—Twenty from the following stations:

Specimens of Hippasteria californica examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
2892......| Santa Barbara Channel, California...... 284 yellow mud cose ewer ete esis | 2 U.S. Nat. Mus.
BOs s-- Off Santa Rosa Island, California....... OO Olean GO cceeeoee races eeccereaees | 1 Do.
2048-2232 Off Santa Cruz Island, California... ... 266 gray sand, gravel, broken shells. . 1 Do.
280 ES cee] dates CO ssasn Maa ee see eee eee aaseeeeee 603 green MUG eo enenc~ cee ee ee 1 Do.
S071. oa | or Sea Lion Rock, Washington......-. B85 Fe ease GO. a sac ee seco sens hale ee | 1 Do.
3104s 22: | Off Half Moon Bay, California.......... 391 elayet iio aoc ceee eeaees wea ee | 1 Do.
3112......| Off Pigeon Point, central California... .. 296 fine'gray sand... 0 --2:--dsenaece 1 | Do.
Bh fT eee | Off Point Sur, central California ........ 328 black sand, mud.... 1 Do.
$844.2... | Off Cape Elizabeth, Washington....... 831 preensMuds veseccse eee 1 | Do.
3348. ....- | Off Point Arena, California............. fine gray sand....... 1 Do.
4345?a...; Off San Diego, California. ............. ETAY: SANG eiiccccn tee eee aac e eal 1 Albatross, 1904.
4427...... Off Santa Cruz Island, California....... blackimudisss-32=.<422 eee oe 3 Do.
BOD eee |e C0 enone one ere eee eae n nea, PreeMIMud owe va cers eee 3 Do.
4430......|.....d0..... 22-22. 22. 02sec cece eee ee ee black sand, pebbles...........-- | a Do.
4530...... | Off Point Pinos, California............. | 847-755 | soft gray mud............------- 1 Do.
| ‘

@ Probably incorrectly labeled.

Remarks.—This species is remarkable for the small marginal plates and for the
shape of the pedicellarize of the actinal surface. The granules and adambulacral
armature are also different from those of phrygiana and spinosa. The abactinal
skeleton is weaker than that of phrygiana, but does not present the fenestrated
appearance of heathi.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 237

The ceelomic cavity of a specimen from station 4429 was nearly filled with
a large parasite, the Ascothoracid crustacean Dendrogaster, a figure of which is
given. (Pl. 111, fig. 1.) This specimen has a very weak dorsal skeleton and the
hepatic coeca are small. Many of the abactinal plates are disconnected and inde-
pendent, a state of affairs which is not found in a normal example.

Genus CRYPTOPELTASTER Fisher.

Cryptopeltaster FisHer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 311. Type,
C. lepidonotus Fisher.

Diagnosis.—Nearly related to Hippasteria, from which it differs in having the
whole abactinal surface covered with numerous flat, circular, quadrate, oval, elliptical,
triangular, and polygonal granules attached to the plates and secondary intermediate
ossicles in such a thanner that there is a very narrow free projecting edge which some-
times overlapsthenextgranule. The primary circular or elliptical abactinal plates are
spaced, and the small intermediate ossicles are quite independent and not arranged
so as to form connecting pieces, as in /ippasteria. Primary abactinal plates with
low central spine or tubercular granule, and many with a large low bivalved pedi-
cellaria occupying entire width of plate. Actinal and marginal plates covered with
flat large granules similar to those of dorsum, but usually without the free rim.
Marginal plates with a central short tubercular spine. Actinal interradial areas
extensive, the intermediate plates adjacent to adambulacrals with large bivalved
pedicellariz; the others with one or more enlarged central granules. Adambulacral
plates with two heavy furrow spines and one shorter actinal spine, similar in char-
acter to those of Hippasteria; but, unlike that genus, the furrow series is often
replaced by a large bivalved pedicellaria, occupying entire length of furrow margin.

Remarks.—There are three features in which this genus differs from all species
of Hippasteria—the character of the granulation, the arrangement of the abactinal
secondary plates, and the substitution of a large bivalved pedicellaria for the
furrow spines on some of the adambulacrals. The last I do not consider as of so
great importance as the first and second. In this genus there is no indication of the
abactinal membrane which is always more or less evident in Hippasteria and in
which the granules are immersed. Sometimes this membrane is thick and com-
pletely hides the underlying plates (as in 7. heathi); it always forms a thin coating
on the granules. In Hippasteria the granules form a border about the larger plates
of the abactinal surface as well as surrounding the marginal plates. Even the
intermediate plates are defined by a marginal series of granules. In Cryptopeltaster
the entire surface is covered with flat valve-like granules, and the intermediate
ossicles usually bear but one large flat valve, like the flaring head of anail. It is on
the intermediate plates that the free edge of the granule is most evident.

Ludwig (1905, p. 140) has indicated that the disposition of the abactinal
secondary intermediate plates of his pacifica (which he figures) is not unlike that of
Hippasteria phrygiana. I have examined the dorsal skeleton of two specimens
of the latter species and find that the intermediate plates act as binders or joiners of
the primary plates. The character of the dorsal skeleton is the same as in /7. spinosa,
and quite unlike that of Cryptopeltaster, where the intermediate plates do not in any
way connect the primary dorsal plates. '

238 BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

CRYPTOPELTASTER LEPIDONOTUS Fisher.
Pl. 47, figs. 1, 2; pl. 58, figs. 4, 4a; pl. 60, fig. 7.
Cryptopeltaster lepidonotus Fisuer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 311.

Diagnosis—Rays five. R=105 mms; r=51 mm.; R=2 (+) r. Breadth
of ray at base, between second and third superomarginal, 50 mm. or less, according
to degree of inflation of abactinal area. Disk large, rays well developed, tapering
to a blunt recurved tip; interbrachial ares very wide and rounded; abactinal area
much inflated on rays and radial areas of disk, also in each interradius adjacent
to marginals. Abactinal plates covered with flattened, irregularly quadrate, cir-
cular, oval, elliptical, and polygonal granules, with usually more or less projecting,
often overlapping, edges. Primary abactinal plates circular or elliptical, well
spaced, with many independent, small intermediate ossicles. Primary plates bear
low conical spines or large low bivalved pedicellariz. Marginal and actinal plates
completely covered with flat granules; each marginal plate with a single tubercular
spine. Adambulacral plates with two thick furrow spines or sometimes a large
pedicellaria on the margin, and one similar actinal spine just behind the furrow
series.

Description.—Abactinal surface is covered with peculiar, flattened, scale-like
granules, which are irregularly quadrate, circular, oval, elliptical, triangular, polyg-
onal, boomerang-shaped, and of several other shapes which defy description, of
greatly varying sizes and so closely placed that they often overlap a trifle. They
are attached to the larger plates of the skeleton in such a manner that usually a
narrow free projecting edge is discernible, or they form the flaring summit of many
variously sized ossicles packed between the regular rows of rather widely separated
primary plates. There is some variation in the amount which the edge of the
granules extend beyond the attachment area. Overlapping of edges is common.
The granules covering the intermediate plates are larger than those surrounding
the spines and pedicellariz.

The exposed surface of many granules is raised into a low tubercular eminence.
Primary plates superficially marked by a robust, low, conical spine, about the base
of which is a series of elongated granules often curiously excavated on the edge, these
spines decreasing in size toward edge of disk and end of ray, and grading into broad
conical granules in the interradial areas, where the primary plates are small, closely
packed, and the secondary ossicles nearly wanting. On the ray is a radial series of
spines and on either side three or four parallel series, all Jow (1.5 mm.), scarcely
more than tubercles. Long, low, bivalved pedicellarie (2.5 to 4 mm. in length) are
numerous on the abactinal surface, especially on interradial areas, center of disk,
and proximal radial areas. Each is surrounded by a series of quadrilateral granules
of various sizes; papule numerous, especially on rays, but apparently absent from
a very small interradial area adjacent to marginal plates.

The abactinal plates when viewed from the inner or ccelomic side are seen to
be either circular or elliptical, there being occasionally an indication of one or more
short lobes. They are arranged in spaced, not very regular, longitudinal series and
the interspaces are filled with independent, small, very unequal ossicles, the upper
ends of which bear large scaledike granules, usually about one (sometimes three or

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 239

four) to an ossicle. The papule are scattered abundantly among these ossicles. The
figure will show more clearly the arrangement of the dorsal plates.

Superomarginal plates rather small, irregularly quadrilateral, higher than long
in middle of interbrachial arc, but longer than high throughout most of ray except
at tip. On account of the inflation of the abactinal surface these plates are mostly
confined to side of ray, and the abactinal edge of each is arched. Each plate is
tumid and bears in the center a rigid acorn-shaped or conical tubercular spine
slightly larger than those of the median radial series. The general surface of plates
is covered with polygonal granules similar to those of abactinal surface, the periph-
eral granules being elongated and a number on each plate convex or low. The
granules are attached by nearly the entire under surface, there being very little
free edge. Superomarginals twenty-six or twenty-seven in number from median
interradial line to extremity of ray. Inferomarginals slightly larger, and more
nearly square on the ray where the upper series is oblong. In general each infero-
marginal corresponds to a superomarginal and is similarly covered with granules
(most of which, exclusive of the peripheral series, are prominent or subconical).
Each bears one or two, rarely three, tubercular spines in the center, all short and
stubby. Smaller plates intercalated here and there in the inferior series are appar-
ently due to injury of some sort.

Adambulacral plates nearly square, each bearing two large, heavy truncate
furrow spines, usually compressed at tip. Sometimes the tip is conspicuously
widened and occasionally grooved or even incipiently bifid or trifid. A large sessile
bivalved pedicellaria occupies the furrow margin of first plate, extending the whole
length of plate; a similar pedicellaria occurs frequently on other plates of the series
but there is no regularity of occurrence. On actinal surface of each plate is a
robust spine just behind the furrow spines, which it greatly resembles, though a
trifle shorter; this spine is either bluntly tipped, compressed, or curiously wrinkled.
It is reduced to a small conical tubercle, a mere specialization of one of the granules,
on those plates in which a pedicellaria replaces the furrow series. General surface
of plate is covered with irregular plate-like granules similar to those of the actinal
intermediate plates.

Mouth plates fairly large. Furrow spines three, flattened, the inner flaring
and hatchet-like, though irregular; an odd spine at inner angle of the combined
plates; one much smaller spine on actinal surface adjacent to outer furrow spine;
surface of plate covered with two longitudinal (interradial) rows of flattened plate-
like granules which diverge and partially surround the actinal spine, being much
smaller toward the inner angle; these granules are quadrate or pentagonal, and
their surface sunken a trifle, then raised in a small low boss in the center. The
actinal spine may be absent.

Actinal interradial areas large; a single row of intermediate plates extending
very nearly to tip; a second series two-thirds the length of the ray; a third very
nearly as far; a fourth nearly half, and so on. Plates adjacent to adambulacrals
largest, elongated transversely, and more regular than the others, though far from
regular themselves; most of them bear a long bivalved sessile pedicellaria (4 mm.)
not more than 0.5 mm. high, placed transversely or obliquely transversely, forming
a very conspicuous series on either side of the adambulacral plates and extending

240 BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

about half the length of ray. Flat, oblong, or irregular granules surrounding them,
the whole being somewhat convex in appearance. A number of the granules have
a subconical prominence in the center. The other intermediate plates bear a
central conical tubercle, surrounded by several low conical granules which grade
into the flatter, plate-like granules toward the periphery of the plate, these granules
resembling those of the abactinal surface in shape but not free on the edges.

Madreporic body fairly large, somewhat convex, irregular in outline, situated
nearer center than midway to margin; striations deep, irregularly interrupted;
anal opening surrounded by a cluster of bead-like granules.

Color in life, scarlet vermilion.

Type.—Cat. No. 22340, U.S.N.M.

Type-locality— Albatross station 4430, off Santa Cruz Island, California, 281
fathoms, black sand, pebbles, rocks.

Distribution.—Known only from the vicinity of Santa Cruz Island, California,
but probably extending to the Tres Marias Islands, Mexico. (See below.)

Specimens examined.—The type, and a specimen in the National Museum
collection from station 2948, off Santa Cruz Island, California, 266 fathoms, gray
sand, gravel, broken shells.

Remarks.—Hippasteria pacifica Ludwig (1905, p. 138, figs. 56, 57, 132) is, I
believe, the same as C. lepidonotus, though in view of some slight differences it may
prove to be a geographical variety. It was taken near the Tres Marias Islands
(station 3425, 680 fathoms). If the two forms are identical, then the range of
lepidonotus will be from southern California to the Tres Marias Islands, Mexico.

The differences seem to be as follows: In pacifica, more clearly defined abactinal
plates, less prominent abactinal tubercles (perhaps due to smaller size of specimen),
and less crowded and overlapping abactinal granules. All these differences may
be due to the small size of Ludwig’s specimen (R=48 mm.). The description of
pacifica otherwise fits lepidonotus very well.

Family LINCKIID Perrier, 1875.

Ophidiasteride © Verri1, Trans. Conn. Acad., vol. 1, pt. 2, 1867, p. 344.

Linckiide Perrier, Réyision des Stellérides, 1875, p. 117.—StapeEn, 1889, p. 397 (includes also
Chetaster).

Linckiade Vicurer, Squelette des Stellérides, 1879, p. 144 (includes Chextaster, Linckia, and
Nardoa only).

Phanerozonia with usually small marginal plates, slender, fingerlike rays and
small disk; skeleton tessellate; tegumentary developments granulate; superam-
bulacral plates usually present but small; pedicellarie, when present, foraminate
or excavate; papule sometimes occurring on actinal surface.

Remarks.—Members of this family are usually found in tropical or subtropical
regions. Only one species reaches California. There is no authentic record of
either Pharia or Phataria from the State, although both occur on the coast of Lower
California and southward.

@ Verrill’s name is the oldest, but Linckiidse has been used by every writer since 1875, and Verrill
has himself adopted it. Ophidiasteridx should have received the recognition that was accorded Echin-
asteride.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 241

Ludwig (1897) has made Chetaster the type of a special family.
KEY TO THE KNOWN GENERA OF LINCKUD®.
a‘. Plates simply granulose, or covered with a thick granular or smooth skin, but not spiny.
b!. Abactinal plates not forming regular longitudinal series.

c!, Adambulacral armature spiniform, often subprismatic, in from one to three series and sometimes
resembling that of Goniasterinze; disk usually well developed, with rays tapering (Nardoa
frequently with small disk and narrow rays).

d', Papule on actinal surface as well as on abactinal.
e!, Adambulacral armature in two or more series; papule isolated; no superambulacral ossicles;

ravemattened sec ss scceh ase Soh 8 cscs eee eads Sacetenaee meade Fromia Gray.
e?. Papulie in areas; superambulacral ossicles; rays rounded; adambulacral armature in three
GE CS ei oe aoe iS Se oc a Ne cae ten tr ater rae Nardoa Gray.

d@. No papule on actinal surface.
e!, Adambulacral armature in one series; marginal plates not larger than others; no super-
ampbiaeral Platessaccns.-usetse = see one Ts eee cee ameter Ferdina Gray.
e’. Adambulacral armature in two or three series (very like the Goniasterinz); marginal
plates conspicuous; rays rather triangular in section; superambulacral plates present.
Narcissia Gray.
c?, Adambulacral armature granuliform in two or three series; rays slender, about same width
throughout; disk always small; superambulacral plates.
d', No papulz on actinal surface; papule in areas.
e!, Rays more or less cylindrical; papule in areas irregularly distributed among the abactinal

BLE LeB ete sjeraiain taste aoe elae ae ee eee are te ara ee ore Linckia Nardo, p. 242.
e?. Rays trigonal; papul in one or two continuous uninterrupted rows on either side of ray.
Phataria Gray.

@. Papule single, on both abactinal and actinal surfaces; adambulacral granules in two series,
the outer much larger than the inner as in Ophidiaster; marginal plates separated near angle
of disk by intermediate intermarginals; all plates bossed, separated by granules.

Bunaster Diderlein.
b?. Abactinal plates in regular longitudinal series; adambulacral armature in two, rarely three,
unequal series, the outermost largest, sometimes spiniform.
c!. Papul on actinal surface; superambulacral plates.
d'. Plates all granulose; that is, underlying a membrane of variable thickness covered with fine,
or coarse, often squamiform granules or papillee.
e', Papular areas not confluent laterally; madreporic body simple.
Jf}. Between the inferomarginal and adambulacral plates three series of actinal intermediate
plates and two series of papular areas; papular areas of the inner series twice as numer-

COTTE BTR LG QUOT, RODLGR sel ere neh a aire nae oe cia mina i ne Hacelia Gray.

f?. Only one ventrolateral (actinal intermediate) series of papular areas which are not more
numerous than in the other longitudinal series ..........--.-..--- Ophidiaster Agassiz.

e. Papular areas confluent laterally; madreporie body large and compound...... Pharia Gray.

d?, Whole test covered with a thick smooth membranous investment.............Leiaster Peters.

c?. No papule on actinal surface; superambulacral plates sometimes very small and covered by
RC Ke UT Gp WIEN ty ses onc he oiolein tk aes nai sachin APE wenn nd anak saan Ophidiaster Agassiz.

a?. Abactinal and marginal plates covered with membrane, the former, and occasionally the latter,
bearing isolated skin-covered spinelets.....:....-...--.-.-------------------d Metrodira? Gray.

The above key differs from that of Sladen in the addition of two genera since
instituted and in the subordination of the superambulacral plates. These are diffi-
cult to find in small specimens, and to base primary divisions on them is likely to
cause confusion. Perrier’s key (1894) is difficult to work, having the same fault as
Sladen’s, in addition to ambiguity of expression in ‘ Section IT A.
are Che aiide te bo xapnsara os Rosier 2 Astéries »3 et ophiures d des iles Aru et Kei, Abb.
Senckenb. Naturf. Ges., vol. 33, 1910, p. 284.

57444°—Bull. 76—11—16

242 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
Genus LINCKIA Nardo.

Lin{clkia Narpo, Oken’s Isis, 1834, p. 717. Type, L. typus (by designation) =L. lxvigata
(Linnzeus).—Gray, Ann. and Mag. Nat. Hist., vol. 6, Dec., 1840, p. 234.

Cribrella AGAssiz, Mém. soc. sci. nat. Neuchatel, vol. 1, 1835, p. 191.

Ophidiaster (part) Miitrer and TroscuEeL, Monatsber. preuss. Akad. Wiss. Berlin, April, 1840,
p. 103.

Acalia Gray (subgenus), Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 285.

Diagnosis.—Linckiide with small disk and long cylindrical arms; abactinal
plates irregularly disposed, not in regular longitudinal series?; papule in areas,
irrezularly scattered, except between the marginal plates; no actinal papule; no
pedicellaria; adambulacral armature granuliform in two series.

LINCKIA COLUMBI® Gray.
Pl. 48, figs. 1-7.

Linckia columbie Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 285; Synopsis, 1866, p. 14.—
SLapEN, Challenger Asteroidea, 1889, p. 784 (Colombia, W. coast).

Linckia ornithopus Verrw1, Trans. Conn. Acad., vol. 1, pt. 2, 1867, p. 330.

Linckia diplax Perrier, Révision des Stellérides, 1875, p. 144 (California).

Linckia pacifica, var. diplax SLaDEN, Challenger Asteroidea, 1889, p. 786 (California, listed with ??).

Phataria (Linckia) fascialis Monks, Proc. Acad. Nat. Sci. Phila., vol. 55, 1903, p. 351 (San Pedro,
California).

Phataria (Linckia) unifascialis Gray, var. bifascialis, Monxs, Proc. Acad. Nat. Sci. Phila., vol. 56,
1904, p. 596.

Diagnosis.—Rays one to nine, usually five; R extremely variable, no two rays
ordinarily the same length; R=90 mm.;r=8 mm.; R=11 r; width of ray at base=r.
Another average specimen: R=60 mm.; r=6 mm.; R=10 r. (Both measured by
longest ray.) Resembling closely L. guildingii Gray, from which the species differs
in having two instead of three regular lateral rows of plates; in having not more than
five prominent abactinal plates on a transverse line across dorsum between the
superomarginals (instead of not less than seven in an adult of guildingii) ; in having
coarser abactinal granules; in having papular areas noticeably smaller than the
primary abactinal plates (larger in guildingii).

Description.—The arrangement of plates on ray is as follows: between adambu-
lacrals and inferomarginals three longitudinal series of actinal intermediate plates
which are lengthened transversely to ray; then two regular series of marginal
plates, irregular or with four or five incipient lobes, the two series being joined by
small secondary ossicles; the dorsum between the two series of superomarginals is
paved with relatively few large plates with four or five incipient lobes, joined by
much smaller irregular oval or somewhat three-sided secondary plates. In a trans-
verse line across the ray between the two series of superomarginals there are from
three to five of the larger plates. Thus there are 3+3 actinal intermediates, 2+2
marginal regular series, +3 to 5 irregular abactinals—thirteen to fifteen plates
around the arm, omitting the adambulacrals and ambulacrals.

Whole body is covered with rather coarse hemispherical granules, which on the
abactinal and lateral surface are fairly uniform in size (being, however, slightly

aExcept in the decidedly aberrant Linckia marmorata (Michelin).

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 243

smaller and less uniform in the papular areas), but on the actinal surface are not at
all uniform and increasing in size toward the furrow, in the vicinity of which they are
larger and often more globose than on the abactinal area. Papular areas small,
always smaller than the adjacent slightly convex abactinal plates. There are few
pores, usually not more than ten, and commonly as few as three or four. In some
examples the granules are not markedly smaller in the papular areas than on the
adjacent plates, but usually they are considerably smaller, with numerous very
small granules intermixed.

Adambulacral armature nearly identical with that of L. guildingii. On the
furrow, two granuliform spinelets, the adoral the larger, with a flaring truncate-
rounded tip about half as wide as the height; next to this (aborally) a shorter, slen-
derer spinelet, either slightly tapering and bluntly pointed, or somewhat clavate and
truncately rounded. Directly back of the larger spinelet is a tubercular granule
more nearly circular (that is, not much compressed) in a transverse section than the
furrow granule, round-tipped, and about the same height as the latter.

Madreporie bodies, usually two, sometimes one, or three to five; flat, varying
greatly in size, shape, and position. Usually two anal apertures, less commonly one,
rarely three or four; rarely two mouths. (Monks.)

Color in life, ‘‘mottled reddish brown and ash color, with the tips of the rays
and the small new rays, or new portion of the ray a brighter and uniform red.”’
(Monks. )

Variation.—Miss Sarah P. Monks studied this species with reference to the varia-
tion and autotomy. She examined over four hundred specimens. The following
is quoted from her paper:*

In over 400 specimens examined not more than four were symmetrical, and no two were alike [in
regard to the size of the rays]. * * * Theanimals vary in size from 4 to17 centimeters. This differ-
ence may in some measure be due to age, but there is evidently also considerable difference in size in
adult specimens.

The normal number of rays is five, but some specimens have only one, while others have four, six,
seven, or even nine; four, five, or six being most common numbers. Of 248 specimens examined 29 had
one ray, 34 four rays, 135 five rays, 44 six rays, 5 seven rays, and 1 nine rays.

There is a marked irregularity in the size of the rays, for the breaking and renewing occur at all ages,
or at least at allsizes. Single living rays without any external sign of diskare notuncommon. Twenty-
two in a lot of 240 were single. In animals measured the rays varied in length from 22-75 mm., and were
in all conditions, from specimens which showed by the raw surface that the breakage was recent, through
all grades of healing surface to stars where new rays appeared as mere buds, to the comet series and to
the various rays of adult Phatariz [i. e., Linckia].

Comets are frequently found; there were forty-one of them in 240 specimens. These are stars with
one long ray and a number of small rays extending from a minute disk. They are rays that have made
a new body.

From the great discrepancy in the size of the rays it follows that one or two rays may be sexually
mature, while others are small or even buds. In two individuals sexually mature Aug. 4, 1901, the
measurements of rays show: male, 77, 70, 58, 20 mm.; female, 70, 65, 52, 13 mm.

There is a great difference in the size of the madreporic body and frequently, but not always, it is
larger in large specimens. In 174 specimens 6 had one, 154 had two, 11 had three, 2 had four and 1 had
five madreporites.

@ Variability and Autotomy of Phataria, Proc. Acad. Nat. Sci. Phila., vol. 56, 1904, p. 596, pl. 42.

244 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

The ani are only visible in live animals. In 68 specimens 15 had one, 48 had two, 3 had three, 1 had
four ani.

There may be some connection, as has been suggested, between madreporites and breaking plane,
but I have failed to find anything satisfactory on the subject. Sometimes the largest ray, or a large ray,
is between two madreporic bodies. In 132 specimens 87 long arms and 45 short arms were between
madreporites.

I have also found this year (August, 1904) that a Phataria may have two mouths.

Type-locality.—‘ West. coast of Colombia” (Gray).

Distribution.—California (Santa Catalina and San Clemente Islands, San
Pedro, San Diego) and Lower California (La Paz); Colombia, west coast (Gray);
Galapagos Islands (California Academy of Sciences coll.).

Specimens examined.—Forty-one from the following localities:

Specimens of Linckia columbiz examined.

Locality. | Depth. | Collector. Number. Collection.
Santa Catalina Island, California............- | Shorestessec eee | Wi. He Dall2-255-= 5 U.S. Nat. Mus.
30 to 40 fathoms. . | nose doses eee 1 Do.
Sete oat alae eee | SRR aes as Berryscas-et 1 Stanford Univ.
Santa Catalina|(Atvalom)=-<--<22-2 eesec nanan seco ete mee eee | Se oleae eta 2 Univ. of Cal.
a:Jolla, San Diego; California. 5972-22523 --|-heuaowen acces eseee| se eeee eens ee 16 Do.

San Clemente Island, California............-. [dN OO Rc Me ee a 12 Do.

SanyDiegox Galiforniameces--a-see eas ce ae a [ieee ke ce Jot ats nea] tare teepoce eee nee 1 Mus. of Comp. Zool.
1 Do.
2 Cal. Acad. of Sei.

La Paz, Lower California. Q
Galapagos Islands.................--------- W. H. Ochsner...

Remarks.—Two questions are suggested: (1) Is this form really distinct from
Tinckia guildingii of the Atlantic and L. pacifica of the Pacific? (2) If distinct,
is the name chosen the correct one?

I may say at the outset, that with the limited material at my disposal, and from
an examination of specimens in the Museum of Comparative Zoélogy, I am unable to
find any very satisfactory differences between L. guildingii, L. pacifica, and L. diplax,
This opinion is also held by Dr. H. L. Clark. The three forms have been differ-
entiated on the basis of the most variable characters, namely, length of ray and
number of madreporic bodies. Perrier says (Révision, p. 143) the only characters
which seem to him, after an examination of numerous specimens, to distinguish
diplax from pacifica are: smaller size, slenderer arms, of a number varying from
four to seven, and the existence of two madreporic bodies. He says further (p. 147)
that he is unable to find any distinctive character between Linckia guildingii (from
the Atlantic) and a specimen collected at Tahiti. All writers (including myself,
1906) have placed faith upon the number of madreporic bodies, whereas in the present
group of species it is worthless for classificatory purposes, as also I think are the
proportions and number of rays.

We have to deal in the present connection only with Linckia guildingii. It
may be assumed that the same characters hold good for ‘“diplax.’’ They do for
two specimens of very different size in the Stanford Collection. One is young, the
other full grown. The following differences hold good for all the specimens I have

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER.

245

examined; six guildingii from Brazil, Tortugas, St. Kitts, and Florida, and forty-one
columbiz from California, Lower California (La Paz), and the Galapagos Islands.

guildingii.

Along side of ray three regular longitudinal rows
of plates, clearly distinguishable except in injured
arms; then they are distinguishable on distal por-
tion of ray.

Abactinal plates numerous; in large specimens,
and also in some small ones, the midradial region
with conspicuously smaller plates.¢ In a trans-
verse line across dorsum, between superomarginals,

columbiz.

Only two regular longitudinal rows of plates
along side of ray (superomarginals and infero-
marginals).

Abactinal plates fewer; the plates ® of the mid-
radial region are never conspicuously smaller
than the marginals. In a transverse line across
dorsum, between superomarginals, one can count

one can count in medium-sized individuals ten to
twelve primary abactinal plates (rarely so few as
seven in small examples).

In mature specimens the papular areas of dorsum
are larger or equal’to the surrounding plates, and
the papule are fifteen to forty to an area.

(In young specimens which have not attained
their growth the papular areas are naturally smaller
than the plates; there is a time when only one
pore to an area obtains.)

in large individuals five primary abactinal plates.

In all specimens the papular areas are very much
smaller than surrounding plates.

Papular pores in fully grown specimens eight to
twelve or fewer to an area.

The following differences may or not be constant, and are not of so much im-
portance as the above: in columbie the granules are coarser, especially in the
papular areas, and the marginal and abactinal plates show four or five slight lobes.
In guildingii the plates are more often three-lobed, or roundish. Columbix is much
smaller than guildingii.

There is no constant difference in proportions, number of madreporites, number
of rays, nor in color. The unreliability of these characters is sufficiently brought
out by Miss Monks’s notes quoted above.

The question concerning the name is satisfactorily answered by two specimens
from Indefatigable Island, Galapagos, collected by Mr. W. H. Ochsner during a
recent (1905-1907) expedition under the auspices of the California Academy of
Sciences. These specimens, the largest of which has R, 26 mm., agree in all par-
ticulars with Californian examples, and exhibit all the differences from guildingii
detailed above. Although not topotypes, they are from a relatively near locality
and make certain the identification of the Californian examples.

L. columbiz is confined pretty closely to continental shores. Is it not likely
that the species may have been derived from ancestors of the present guildingii
which spread through the “Straits of Panama” in Tertiary times and ranged north
and south from that point ?

Autotomy.—Three specimens collected by Dr. W. H. Dall show a single arm
regenerating the entire animal. In one there is just an indication of three buds,
representing future rays; in the others, the buds of four rays are well started; in

2 It is necessary to clean an arm with sodium hypochlorite or caustic potash to make out this point

in small specimens.
b This means primary plates, not secondary connecting ossicles.

246 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

one of these regeneration is sufficiently advanced so that the beginning of a disk is
to be seen. The mouth is developed, but as yet no madreporic bodies. It is evident,
therefore, that this species may regenerate an entire animal, in nature, from an
arm, when no part of the disk is present. Miss Monks has conducted experiments
on this species. The following extracts are from her paper, already cited:

I have been fortunate in a series of experiments extending to the present time, August, 1904, in
having a number of single rays, cut at various places, regenerate the disk and other rays. I cut off rays
and in that way the stretching of the pyloric ceca and the consequent loss of time for its withdrawal
were avoided and after a number of months the ray had reached the comet stage. In one case two arms
started to grow from the place from which one had been cut. * * *

The breaking plane in the skeleton may occur near the disk, from 6 to 12 mm, from the angle of the
ray, or at almost any point on the ray. I have never seen a break that was decidedly inside the disk.

The cause of breaking isobscure. * * * Ifany external force bearsa part in breaking the animal,
it is probably that the creature is surprised when limp and relaxed, but I am inclined to think that
Phataria [i. e., Linckia] always breaks itself, no matter what may be the impulse. They may break
when conditions are changed, sometimes within a few hours after being placed in jars, or a few days,
or even not for months. Some never break, but stand all kinds of inconvenience of heat and cold and
stagnant water. * * * Whatever may be the stimulus, the animal can and does break of itself.

During the years 1901-1904 I have watched more than fifty starfish disunite. * * * The ordinary
method is for the main portior of the starfish to remain fixed and passive with the tube feet set on the side
opposite the departing ray, and for this ray to walk slowly away at right angles to the body, to change
position, twist, and do all the active labor necessary to the breakage. If there are more rays than are
needed to resist the pulling arm, they take a negative part and position. Frequently they hang limp,
with their ambulacree withdrawn.

There seems to be the same coordination in this work as in the walking of Asteroidea, for in the six
or eight positions taken during the severing of a ray the larger portion of the star always keeps pulling
in a direction opposite that of the parting ray, and not in any, or all, of the four, five, or six directions of
radiation.

The skeleton breaks in one of two ways. In one way the ray swells about its midpoint and grows
smaller at its proximal portion as it pulls away from the body. In the region of strain the ossicles become
separated for a space of about 25 mm., so that the white connective and muscular tissues shine between
them till finally, just before parting, they stand out like beads on a network of shiny white floss.
There is much more elasticity in the animal at this time than would seem possible in a starfish.

Immediately after the break the ossicles recover their compact position and the two parted edges
show a more or less jagged surface. This way is rapid, the body wall being pulled apart in from three to
thirty minutes after the arm becomes narrowed.

In the other way of breaking only a small area is affected by the strain, and the breaking place shows
from the first as a fracture. There is no narrowing of the ray, no strained surface and no gradual separa-
tion of ossicles. A small opening, something like a crack, appears on the dorsal side of the ray, and this
increases with an uneven edge till the body wall is parted around the entire ray. The first visible evi-
dence of this kind of division is, either the ray is in a strained position opposite the body or it lops over
and hangs limp, letting its weight act as the pulling power. The invariable method in Phataria I have
watched, after the body breaks, is: The arm walks away; the pyloric ceca are pulled out of the arm,
sometimes for 83 mm., and kept very tense; then a break occurs in the tube connecting the pyloric
ceca with the stomach; the pyloric czeca are withdrawn into the arm and the small portion of the tube
into the body and the arm is free. The parting of the soft tissues is not abrupt, is accomplished by
sundering strands thread by thread. The pyloric czeca and tube are wonderfully elastic. Sometimes
the tube is stretched till it appears as the finest thread.

After breaking the pyloric czca coil up, and as they are retracted to their ordinary position an occa-
ional spasmodic jerk is seen in them. Sometimes they are withdrawn immediately, but frequently an
arm crawls around for hours, or days, trailing the cecum till it is withdrawn or falls away.

Probably if the ceeca are not taken in within a few hours the muscles lose their elasticity and the
ceeca decay and drop off. Occasionally these organs are left on the disk, and sometimes the tube is broken
in such a manner as to leave them hanging like two coiled trails. After freeing itself the ray crawls up

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 247

the glass and remains so for a long time. This is the attitude of a single ray till all, or nearly all, the
pyloric cecum assumes its normal position.

After one arm is separated, or even when it is going, another may break away. I saw one animal
try to break the disk apart after having lost two rays. It took many positions, strained the tissues, and
worked for hours, but died before the disk was broken.

The results of my observations are: The breaking is automatic and is effected by pulling apart or
fracture without strain; there is coordination of parts in producing the separation; the tissues relax at
the plane of rupture; this plane may be near the disk or a variable distance from it on the ray; the
pyloric cxca are always pulled out and much stretched; the break of the caecum occurs at the
tube connecting the stomach and glandular portion; the pyloric cecum is generally taken back
into the arm; the severed ray may live more than a week, even under adverse circumstances, without
signs of regeneration; and, from my experiments extending over a period of three years, I have found
that rays cut at various distances from the disk make disks, mouths, and new rays in about six months.

Family ASTEROPID# 2 Fisher, 1908.

Goniasteride PERRIER (part), Révision des Stellérides, 1875, p. 280.

Goniasteridx (part) and Asterinidx (part) Vieurer, Squelette des Stellérides, 1879.

Gymnasteriade Perrier, Mém. sur les Etoiles de Mer, etc., 1884, pp. 165, 229.

Gymnasteriide SLADEN, Challenger Asteroidea, 1889, p. 355.—Authors generally except Perrier.

Gymnasteriide + Poraniide Perrier, Expéd. sci. du Travailleur et du Talisman, 1894, pp.
168, 327.

Asteropide Fisner, Smiths. Misc. Coll. (Quarterly), vol. 52 (No. 1799), 1908, p. 90; Zool. Anz.,
vol. 32, 1908, p. 358.

Phanerozonia with the test covered by a skin or membrane, either smooth,
granulose, or beset with spinelets; marginal plates prominent, more or less overlap-
ping, smooth or with a single spine or several marginal spinelets; abactinal skeleton
loosely tessellate or reticulate; actinal plates fairly large, in chevrons or in isolated
serial arrangement; papule usually in areas never actinal but sometimes inter-
marginal; pedicellariz when present bivalved or pincer-shaped.

KEY TO THE KNOWN GENERA OF ASTEROPID®.

a', Marginal and abactinal plates devoid of spines.
b!. Abactinal plates slightly lobed or irregular, tessellated, not forming a reticulum; adambulacral
armature: furrow series consisting of two spines; actinal spines one or two; a pair of specially
localized pedicellarize at base of rays on abactinal surface.................. Petricia® Gray.

a The change of name is necessitated by the fact that Asterope supersedes Gymnasteria. Although
Poraniide is in the field it has been used only by Perrier in a restricted sense for Porania, Tylaster, Mar-
ginaster, and Poraniomorpha, which he regards as belonging to the Spinulosa. To avoid confusion it is
better to retain Asterope as the type of the family.

6 Petricia as here used replaces Asteropsis of authors, but not of Miiller and Troschel. The latter,
which was published in 1840, and notin the System der Asteriden, 1842, as invariably quoted, has exactly
the same signification as Gray’s Gymnasteria (Dec., 1840), but is in turn antedated by Asterope Miiller
and Troschel, as explained below (next footnote). The Archiv fiir Naturgeschichte appeared in three
parts toeach volume. The article Ueber die Gattungen der Asterien (Auszug aus dem Monatsb. preuss.
Akad. Wiss. Berlin, April, 1840), in which Asteropsis is described, was very probably in the beginning
of the third part, which would make the date about September, 1840. The genus is monotypic, and
the name cannot therefore be shifted to Asteropsis vernicina (Lamarck), as has been done by Perrier
(Révision des Stellérides, 1875, p. 282) and authors since. Aséeropsis is a synonym of Asterope, along
with Gymnasteria. Asteropsis vernicina (Lamarck) Perrier becomes Petricia vernicina. The genus
Petricia Gray (Proc. Zool. Soc., pt. 15, 1847, p. 81) has for type P. punctata Gray, which equals
Asterias vernicina Lamarck,

248 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

b?, Abactinal plates forming a composite reticulated meshwork, inclosing large papular areas; adam-
bulacral armature: furrows series consisting of one spine with fleshy knob, and one smaller actinal
spine, the latter hidden in a continuous membranous fold parallel to furrow; marginal plates
obliquely placed, strongly imbricating, more or less pear-shaped; no localized pedicellarie,
but sometimes granuliform pedicellariz on papular areas. ....-...-- Dermasterias Perrier, p. 248.

a?, Marginal plates with spines or spinelets.

b!. Abactinal plates rounded, forming fairly regular longitudinal series; numerous actinal intermediate
plates; all plates covered and mostly obscured by thick skin; a medioradial and superomarginal
series of spines; superomarginal plates forming the margin and bearing large isolated spines;
abactinal plates in a definite medioradial and parallel series; abactinal plates forming also
regular transverse series, the medioradial elongated longitudinally, the dorsolaterals elongated
transversely (not rounded as stated by Sladen)...-....--.----- Asterope® Miiller and Troschel.

b?. Abactinal plates not forming regular longitudinal series; no very prominent median radial series
as in Asterope.

cl. Abactinal area devoid of plating; actinal intermediate plates forming isolated bands in trans-
Verde SOLICS sok Sec cacictos seniors See eine eee ite anteeieetetome Tylaster Danielssen and Koren.

c, Abactinal area plated; actinal area with large intermediate plates.
d@. Plates covered with thick skin, usually devoid of spines, or only rare isolated ones present.
Porania Gray.
@. Plates usually spinous, or with the whole membrane closely crowded with miliary spinelets.
e!. Inferomarginal plates with a flattened comb of spinelets which form a broad marginal

PRIN O CL ae emis oo ic ela bitinie cmiercl see rie eles RNC ee eee Marginaster Perrier.
é. Spines when borne on the marginal plates grouped and inconspicuous, not forming a mar-
ginal fringe; papule isolated or grouped ....-..- Poraniomorpha» Danielsson and Koren.

Genus DERMASTERIAS Perrier.

Dermasterias PERRIER, Révision des Stellérides, 1875, p. 282. Type, Dermasterias inermis Per-

rier=Asteropsis imbricata Grube.

Asteropsis of AUTHORS, previous to 1875, and A. AGassiz, North American Starfishes, 1877, p. 106

(not Asteropsis Miller and Troschel).

Diagnosis.—Stellate, with moderately large, obliquely placed, imbricating
marginal plates; whole test covered with a thick leathery skin obscuring the plates;
no spines except on adambulacral plates; abactinal plates forming a meshwork
inclosing large papular areas; papule numerous; adambulacral plates with a single
furrow spine, and one actinal, the former webbed in a continuous series along
furrow, the latter immersed in a continuous parallel fold of thick skin; granuliform
abactinal, and valvate adambulacral, pedicellarize sometimes present.

Only one species known.

@ Within the year 1840 Asterias carinifera Lamarck was made the type of three genera: Asterope
Miller and Troschel, Asteropsis Miiller and Troschel, and Gymnasteria Gray. The first was described
in April (Monatsb. preuss. Akad. Wiss. Berlin, p. 104), the second in September (Archiv f. Naturgesch.,
p. 322), and the last in December (Ann. and Mag. Nat. Hist., vol. 6, p. 278). Miller and Troschel
thought that their Asterope was invalidated by Asterope Philippi (for a crustacean), published in June
of the same year in Archiv f. Naturg., Heft 2, p. 186, so that they changed the name to Asteropsis in
their article Ueber die Gattungen der Asterien, published about September in the same journal, p. 322.
Investigation proves, however, that Philippi’s name was published fully two months after that of
Miiller and Troschel, the evidence being a reprint of Philippi’s article in the Annals and Magazine
of Natural History, vol. 6, September, 1840, p. 89. This translation states that it is reprinted from
Wiegmann’s Archiv, Heft 2, June, 1840, thus fixing the date. There is no reason why Asterope should
not replace Gymnasteria. Sladen was in error.

, bIncludes Rhegaster Sladen and Lasiaster Sladen. See Grieg, Bergens Museums Aarbog, 1906,
No. 13, p. 41.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 249

DERMASTERIAS IMBRICATA (Grube).
Pl. 49, figs. 1-2; pl. 56, fig. 7.

Asteropsis imbricata Grose, Archiv f. Naturgesch., 1857, p. 340; Nova Acta Acad. Cees. Leop.
Carol. Nat. Cur., vol. 27, 1857, p. 6, pl. 1, fig. 5 (Sitka).—Verrm, Trans. Conn. Acad., vol.
1, pt. 2, 1867, p. 324.—A. Acassiz, Mem. Mus. Comp. Zodl., vol. 5, No. 1, 1877, p. 106, pl. 15.

Dermasterias inermis Perrier, Révision des Stellérides, 1875, p. 282.

Dermasterias imbricata SuapEN, Challenger Asteroidea, 1889 (Vancouver Island to San Fran-
cisco).—Ciark, Proc. Boston Soc. Nat. Hist., vol. 29, No. 15, 1901, p. 325, pl. 1; pl. 4, figs.
2-4 (Puget Sound).

“Dermaster” imbricatus J. F. Warreaves, Trans. Roy. Soc. Canada, vol. 4, sec. 4, 1886 (1887),
p. 117 (Melaspina Inlet, British Columbia and Johnstone Strait, British Columbia).

Diagnosis —Rays five. R=120 mm.; r=45 mm.; R—2.6+r. Breadth of
ray at base, 50 mm. A small specimen: R=55 mm.; r=22 mm.; r=2.5 mm.
Breadth of ray at base, 29 mm. Rays rather evenly tapered from broad base to
blunt extremity; disk usually fairly high and abactinal surface more or less inflated
with interradial depressions; actinal surface flat. Whole animal covered with a
thick leathery membrane wholly obscuring the underlying plates, unless dried.
No spines except along furrow margin. Membrane contains flattened lime spicules
or deposits in the shape of rods or perforated plates. Papule in areas, six or eight
longitudinal rows on each ray, and upward to twenty-two papule to an area;
large specimens usually with several (one to nine) small pedicellariw on each area,
consisting of two, three, or four granuliform jaws; pedicellarie absent from small
and some large specimens. Marginal plates smooth, imbricated, numerous, nor-
mally wholly obscured by skin. Adambulacral plates with a stout blunt furrow
spine tipped with a soft fleshy knob, the whole series along furrow margin being
connected by tough skin for about half the length of spines; back of these and
separated by a distinct furrow is a parallel ridge of tough skin not quite so high as
the furrow spines, containing, for each plate, a flattened ovate or elliptical spinelet ;
the tips of these spinelets give the. ridge a scalloped appearance. In some large
specimens, in this furrow, between the two series of spinules, at base of ray are a
number of pedicellariz with one, two, or three flattened ovate or elliptical jaws,
much larger than those of abactinal surface. Tube feet stout with large sucking
disks, no deposits in wall. Madreporic body flat, sunken in the abactinal skin;
strix fine, branching, radiating from an eccentric point.

Description Grube (1857*), Agassiz (1877), and Clark (1901) have figured
and described this species more or less completely. It will only be necessary, there-
fore, to enlarge upon certain features, as current descriptions need modification in
the light of more abundant material.

The number of papulz to an area is, of course, very variable, depending chiefly
upon the size of the example. The areas are not always in very definite or regular
series, but usually are. The pedicellariw are present in considerable numbers on
the abactinal surface of all (seven) the large northern specimens, and also on two
medium-sized ones (Naha Bay). Two fairly large examples (Barclay Sound) have
only a very few pedicellarizw, and one (Barclay Sound), with R 73 mm., has none.
Small examples usually lack pedicellarie. Specimens from the vicinity of Mon-
terey Bay (one large and numerous medium-sized and small) lack pedicellarie
altogether. The northern specimens, therefore, differ from the southern in usually

250 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

having, in full-grown examples, numerous abactinal pedicellarix. The jaws are
low, rounded, and granuliform; two and three jaws are commoner than four. The
anal opening is surrounded by numerous eranuliform spinelets. Barring the small
pedicellariz the surface of the starfish is slick and smooth. The actinal interme-
diate areas are smooth, therefore, and are seamed with minute lines or wrinkles
which have an interradial direction.

The adambulacral plates are very small, wider than long, and no longer than
the width of the spine at base. The calcareous part of the furrow spine is tapering
and bluntly pointed. The spines are joined by a continuous web and capped by a
prominent fleshy knob. The actinal spinelet is shorter (especially near base of ray),
flattened, broadly lanceolate, and often appressed to the adjacent actinal interme-
diate plate, so that, on account of the heavy skin, there appears to be, in such cases, no
actual spinelet. ‘Toward extremity of ray the spinelets are upright and form a scal-
loped edge to the investing membrane. Adambulacral pedicellariz are not of constant
occurrence in northern examples, and have not been found in southern specimens.

The abactinal skeleton is an open meshwork, consisting of rather widely sepa-
rated scalloped primary plates connected by numerous slightly overlapping elon-
gated ossicles forming narrow bands between the primary plates, as shown by the
figure. Agassiz’s figure 3, plate 15, American Starfishes, shows this feature well.
In his figure 2 the ossicles appear more loosely joined than normally, but the clear
description makes the arrangement plain. The spaces between the meshes of the
skeleton, occupied chiefly by the papulz, contain calcareous granules, which on the
disk sometimes form short spurs extending from the main trabecule. (See Pl. 49,
fig. 2, right-hand photograph.)

Anatomical notes.—The intestinal ccecum is a sac with five irregularly lobulated
short saccular divisions; aboral division of stomach large; hepatic cceca large,
reaching far along ray. Gonads in a thick tuft attached to the calcareous arch of
the interradial septa; from the point of attachment a tube (gonoduct) passes
upward for a short distance to the aboral surface. No superambulacral plates.

Type-locality.— Sitka (Grube).

Distribution.—Sitka, Alaska, to Monterey Bay, California, low tide and shallow
water, rocks.

Specimens examined.—Twenty-two, from the following localities:

Specimens of Dermasterias imbricata examined.

Locality. Depth. | Collector. Number. Collection.
|
Tongass, southeastern Alaska...........--.- Shore. .... | Lieut.-Commander Nichols. ....-.. il | U.S. Nat. Mus.
Revillagidedo Island, Alaska.................- eee T. H. Streets. 1 | Do.
Kasaan Bay, Prince of Wales Island, / -do. . ae see 1 | Do.
Naha Bay, Behm Canal, Alaska... ...- seeeeises donee | Albatross, 1905 2 Bur. of Fisheries.
Barclay Sound, British Columbia... ....... bai een | A 10 U.S. Nat. Mus.
Puget Sound, Washington..................|--d0....... |) Chas. /MoDrake® = so-so eee cee 3 | Do.
‘Tacoma. Weshington+ c-c---24- see eee ees eAry Mrs.W.) Beelare: 2-2. eee 1 | Do.
Kildare, Oregon....... ...| Oscar C. Sturges a4 Bil Do.
eee DS) Jordanses4 2... ee 1 | Do.

Monterey Bay, Californ

Also numerous specimens in the Stanford collection from Half Moon Bay and
Monterey Bay, California.

Remarks.—No other species has been described in this genus, and the genus
itself is sharply marked off from its nearest allies, Petricia and Asterope.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 251

Order SPINULOSA Perrier.

Pedicellarie rare, never forcipiform; abactinal skeleton reticulate or imbri-
cated, sometimes absent; abactinal spines always present and usually numerous,
isolated, in groups, or forming regular fascicules and pseudopaxille#; marginal plates
usually inconspicuous; papule dorsal only or also intramarginal and actinal;
ambulacral plates not crowded and compressed; actinostome with adambulacral
plates prominent; tube feet with well developed sucking disk, and usually biserial;
mouth plates medium-sized or large.

This order is poorly defined as regards its limits toward the Phanerozonia, but
is sharply separated from the Forcipulata by the absence of the characteristic
forcipiform pedicellariz of the latter. Certain of the Asteropide have been
regarded at different times by Perrier as a part of the Asterinide, or as constituting
a separate family, the Poraniide. The entire order Velata of Perrier has been
merged with the Spinulosa.

The Asteropide contain forms which are not very different from the Gonias-
teride and other genera which show decided relationships with the Spinulosa.
The family is therefore in a measure transitional. The Ganeriide are also some-
what intermediate, but appear to be nearer the Asterinide and Echinasteride,
although the marginal plates are phanerozonid.

The Cryptozonia of Sladen includes, besides the families here assigned to the
Spinulosa, all the Forcipulata and the Linckiide, and lacks the Ganeriide and
Asterinide, which were relegated to the Phanerozonia. The Linckiide are nearer
the Goniasteride than are the Asterinide, but it will probably long be a matter
upon which opinion will differ as to whether the Asteropide is a natural family.
Sladen’s arrangement of his cryptozoniate families is certainly unnatural. The
placing of the Zoroasteride and the Stichasteride next to the Linckiide and far
removed from the Asteriide has nothing to recommend it. The Forcipulata is
probably the best defined of all the orders, as its families have less in common with
the Spinulosa than have those of the Phanerozonia.

Tf the families constituting the Phanerozonia are first removed, then the For-
cipulata segregated, the remainder will be the assemblage constituting the Spinu-
losa—an order not difficult to recognize but decidedly difficult to adequately

characterize.
KEY TO THE KNOWN FAMILIES OF SPINULOSA.

a‘. Mouth plates rather small, not spade-shaped or plowshare-shaped; ambulacral furrows narrow.

b'. Marginal plates large and phanerozonid in character; abactinal skeleton imbricated or reticulated
actinal plates regularly arranged in rows extending from adambulacrals to marginals and bear-
ing one or two large spines or small groups of spinelets................---+-+-.+++0+- Ganeriidx 4

b?. Marginal plates inconspicuous.

c'. Abactinal skeleton well developed, not wholly aborted.

d', Abactinal skeleton formed of closely imbricating plates bearing small spines; actinal skeleton
formed of imbricating plates bearing a tuft or fan of spinelets...........Asterinidx, p. 253.

a Comprises Ganeria, Lebrunaster, Cycethra, Radiaster, and Scotiaster.

BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

bo
n
bo

©. Abactinal skeleton formed of plates disposed in longitudinal and transverse series forming
an irregular or regular mesh work, more or less open, the plates bearing isolated spines
or groups of spinelets, but never fans of spinelets.
e!, Ampulle single; no pedicellarize; interbrachial septa weak, more or less rudimentary;
abactinal reticulated skeleton irregular, bearing single spinelets, or groups of spinelets,
or langer spines. ....-..-------6--eecee cence eee e eee ce ces ce een cnee Echinasteridx, p. 258.
eé. Ampulle double.

f}. No pedicellariz; disk very small; no interbrachial septa, skeletal reticulations roughly
in the form of triangles grouped in hexagons; skin, and the large obtuse spines beset
with rough scales or spinelets. ..-. OE iaeieie aslo eee ea Oana eae ee erect Mithrodiidx.@

/?. Small two-jawed upright, or large low bivalved pedicellarie; disk large, interbrachial
partitions well developed, with plates; no superficial covering of scales or spinelets over
the skin and spines.

g. Rays numerous; numerous madreporic bodies; pedicellariz consisting of two upright
tapering jaws arising from a small independent plate; no calcareous rosette in suckers
of tube feet; skeleton open reticulate, armed with large isolated spines covered
with membrane beset with calcareous granules .......-.----------- Acanthasteridx.

g®. Rays five; one madreporic body; large low bivalved pedicellarize on superomarginal
plates, in addition to a few very small upright two-jawed pedicellariz on the abactinal
surface and adambulacral plates; a calcareous reticulum forming a rosette in suckers
of tube feet; abactinal plates regularly arranged, leaving small triangular papular
spaces, and bearing small isolated spines; actinal intermediate plates bearing one

tojseveral! large flat; spimeletse. con scene etele sere see ie eee Sepe-cee Valvasteridx. ¢
c. Abactinal skeleton entirely aborted, only the ambulacral and adambulacral systems remaining-
Integument covered with small membrane-enveloped spines .........-..------ Cryasteride.@

a*, Mouth plates large and shaped like a pointed spade or like a plowshare, with conspicuous marginal
and usually also suboral spines; furrows wide, adambulacral armature pectinate.

b'. Marginal plates paxilliform, in either one or two series; abactinal paxillee never with long spine-
lets united by membrane but fascicular or penicillate; mouth plates spade-shaped. Actinal
intermediate plates present; adambulacral armature in two series at right angles; interbrachial
BODE werner eT sein ot io Stee ee eee sie = int et eae oat te atete etokele aerators Solasteridx, p. 305.

b?. Marginal plates either hidden by a supradorsal membrane, or if visible not conspicuous and
paxilliform; abactinal plates with spines more or less united by membrane; mouth plates
plowshare-shaped (of the type of Pteraster); actinal intermediate plates absent or indistinguish-
able; adambulacral armature in one or two series; interbrachial septa present or rudimentary.

c!. No supradorsal membrane forming a special nidamental cavity; no segmental pores or papillz.
d', Interbrachial septa partly calcified; adambulacral armature forming with that of inferomar-
ginal plate a single transverse series, never webbed; abactinal skeleton composed of round-

ish plates, or lobed plates forming a wide mesh.......-...-------- Korethrasteridx, p. 340.

@ Mithrodiinze Viguier, 1879, p. 128; Mithrodine Perrier, 1884, p. 164; Mithrodide Perrier, 1894,
p. 4; Mithrodiide Fisher, 1906, p. 1094. This family contains only one genus, Mithrodia Gray, and
differs from the Echinasteride in having double ampullx, in lacking interbrachial septa, and in having
the entire body, including spines, overlaid with rough granules and squamiform tubercles or spinelets.

+ This is equivalent to the Acanthasterine Sladen, 1889. Acanthaster is a very isolated type, the
sole survivor of perhaps a once extensive group. The family of which it is the type differs from the
Echinasteride, restricted, in having double ampulle, well developed interbrachial septa, numerous
madreporic bodies, upright, two-jawed pedicellariz, and many rays.

¢ This family is equivalent to the Valvasterine Viguier, 1878. It differs from the Echinasteridz
in having double ampulle, bivalved and two-jawed upright pedicellariz, a regular abactinal skeleton,
conspicuous marginals, strong interbrachial septa, and a rosette of calcareous reticulum in the suckers
of the tube feet. Koehler also considers this group of family rank (‘‘ Valvastéridés,’’ Koehler, 1910,
p. 176).

@“*Cryastéridées” Koehler, Expédition antarctique frangaise, Echinodermes, Dec., 1906, p. 24,
Genera Cryaster Koehler and Magdalenaster Keebler.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 253

@. Interbrachial septa rudimentary, sometimes partly calcified;¢ adambulacral spines delicate,
webbed, not associated with inferomarginal spines; abactinal and lateral plates very deli-
cate, imbedded in the integument, bearing groups of delicate short or long spinelets

united by a web or imbedded in a sacculus ........-..........-.-----.-.-- Myzasteridg.»
c’. A complete supradorsal membrane present; segmental pores and papille; interbrachial septa
not calcified ;¢ actinolateral spines present ...............-.--------++-- Pterasteridx, p. 343.

Family ASTERINTDAE Gray 1840, emended.

Asterinide Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 228 (includes Palmipes, Porania,
Asterina, Patiria, Socomia).—Prrrier, Révision des Stellérides, 1875, p. 289; Mission scientif.
Cap Horn, 1891, p. 107 (includes Porania, Ganeria, and allies); Exp. sci. du Travailleur et du
Talisman, 1894, p. 141.—StapeEn, Challenger Asteroidea, 1889 (includes Ganeriinw).—FisHer,
1906 (excludes Anseropodinse+Tremasterins).

Spinulosa with dorsal skeleton formed of imbricating plates bearing small
spinelets or granules singly or in groups or tufts; with regular imbricated actinal
intermediate plates bearing a few spinelets or granules, sometimes in combs; mar-
ginals minute, usually defining ambitus.

KEY TO THE SUBFAMILIES AND KNOWN GENERA OF ASTERINIDE.

a‘. Papuls distributed widely over disk and radial regions; no aperture in each interradius perforating
the disk as in keyhole sand-dollars (Mellita); abactinal plates externally crescentiform, thick,
bearing granules or short spinelets; plates outside papular areas not thin and scale-like, with
ER GED A OG CB feat ais era e ele eo )aeteaale oe os mie rer ciate Se ele g iee et et Asterining Sladen.

b!. Abactinal plates not imbricated; covered with spines..................-..-- Parasterina 4 Fisher
6?. Abactinal plates imbricated throughout the abactinal area, or only in definite regions.
c!. Rays cylindrical, more or less elongate; disk small; with a definitely and sharply defined band
of crescentiform plates along the median areas of the rays............-.-.- Nepanthia Gray.
c?. Rays and body flattened; disk large with sharply angular edges.
d'. Abactinal plates imbricated by three or four fairly regular internal processes or lobes and
bearing on exposed more or less crescentiform surface, or on free edge, granules or short

spinelets; all abactinal plates subsimilar.................-.------- Asterina Nardo, p. 254.
d*. Abactinal plates imbricating only near the margin, the others spaced apart; covered with
Tiss OOS Kat eee ae ieee eee cieete d Seek ce eie sauce ob ieee maccnue Disasterina Perrier.

a’. Papule confined to comparatively narrow radial band; plates of, papular area strongly stellate.
imbricating, the others thin, scale-like, strongly imbricating, the internal edge pointed and pro-
longed downward into a process which meets a similar process from a ventral plate and forms a
dorso-ventral supporting pillar for the very thin disk; general form of dorsolateral plates broadly
IAN COOIREG Jose eon Stine cule ere a cae ie Sats eeiaea e mh ahs ae Inia ain ee Anseropodinw¢ Fisher.

4 Sladen (1889, p. xlii) and Bell (1892, p. 4) imply that septa are absent. They are present but
do not extend far toward center of disk, and contain calcareous plates.

6 This family includes the Pythonasterine with one genus, Pythonaster Sladen, and Myxasterinse
with two genera, Myzaster Perrier, and the even more remarkable Asthenactis Fisher.

cThe Pterasteride are said to lack interbrachial septa. These are present in several species of
Pteraster (as, for example, jordani and tesselatus) and in Hymenaster kahleri, but are not calcified.
The interbrachial septa of. 7. quadrispinosus have small thin embryonic plates, which are not visible
unless the membrane is examined under a microscope. Such a condition will not be readily confused
with an ordinary calcareous or partly calcified septum. See Sladen 1889, p. xlii.

@This genus is equal to Patiria of authors, not of Gray. Gray’s Patiria is a synonym of Asterina.
Fisher, Smiths. Misc. Coll. (Quarterly), vol. 52, no, 1799, 1908, p. 90; Zool. Anz., 18 Aug., 1908, p.358.

e Bull. U. S. Fish Comm. for 1908, pt. 3, 1906, p. 1087, footnote. I now think it is better to keep
the three subfamilies in one family, and segregate Cycethra, Ganeria, and Kadiaster as Perrier has done

(1894).

254 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

b!. Abactinal plates bearing tuits of spinelets; no membrane or granules; papule in a single row on
either side of median radial line.......-.--------s222222--0-ere sss ee- == Anseropoda Nardo.

}2, Abactinal plates covered with membrane, bearing granules; no tufts of spinelets; papule in
several rows on each side of the median radial line. (Furrow spinelets two or three, united along

ray by granulous membrane; no actinal adambulacral spinelets)........-.-.- Stegnaster Sladen,

«3, With five interradially placed slits, perforating disk (opening on both surfaces); general form that of
Anseropoda but thicker; abactinal plates imbricated, with free margin toward center of disk, and
forming a somewhat rhombic squamiform tessellation ; papulee in a petaloid radial area, not reach-
ing center of disk......--------------- +22 eee seen cere eee e eee rere Tremasterine © Sladen.
Interradial slits about as long as width of plate, situated one-fourth r from center, guarded by seven
or eight spinelets on each side; ventrally apertures open just behind broad mouth plates; anal
opening prominent, guarded by spinelets; on papular areas the adcentral free curved edge of
plates provided with four to ten sharp appressed spines covering four or more papulze and appar-
ently protecting them; papule emerging singly under edge of plates; elsewhere abactinal plates
without marginal spines but roughened by low granule-like protuberances; actinal intermediate
plates armed with slightly flattened spines similar to those of Valvaster, about one to a plate;
adambulacral plates with a minute spine in furrow (sometimes two), and then ina transverse line
two conspicuous spines like those of actinal intermediate plates >........-------- Tremaster Verrill.

Genus ASTERINA Nardo.

Asterina Narvo, De Asteriis, Oken’s Isis, 1834, p. 716. Type, Asterias minuta Olivi=Asterias

gibbosa Pennant.

Asteriscus Miter and TroscHen (part), Monatsber. preuss. Akad. Wiss. Berlin, 1840, p. 140.

Patiria Gray (not Perrier, Sladen, and authors), Ann. and Mag. Nat. Hist., vol. 6, Dec., 1840,

p. 290. Type Patina (Asterina) coccinea Gray.

Diagnosis.—General form broadly stellate to pentagonal, with large, more
or less elevated disk, and sharp angular margins defined by minute regular marginal
plates; abactinal plates externally crescentiform, interspersed with eranuliform
secondary plates, and bearing granules or short spinelets; internally three or four
lobed and imbricating; papule in areas or single, widely distributed; actinal and
adambulacral plates with combs of spinelets sometimes united by web.

This genus, which is in need of a thorough revision, contains about thirty species,
some of which have a wide distribution. The genus as a whole is confined to the
littoral zone, never occurring at any considerable depth, and is nearly cosmopolitan
in tropical and temperate seas.

ASTERINA MINIATA (Brandt).
Pl. 56, figs. 8, 8a.; pl. 61, figs. 1-4; pl. 62, figs. 1, 2.

Asterias miniata Branpt, Prodromus, 1835, p. 68 (Sitka).
Asteriscus miniatus Stimpson, Boston Journ. Nat. Hist., vol. 6, 1857, p. 530.
Asterina miniata PERRIER.—SLADEN, Challenger Asteroidea, 1889, p. 774.

Diagnosis.—Rays five (exceptionally six or seven). General form stellate with
short broad rays and high, inflated disk; R=67 mm.; r=40 mm.; R=1.7 r;
breadth of ray at base, 44 mm.; occasional specimens with R=2 r; interbrachial
ares rounded and wide; tip of rays blunt. Primary abactinal plates crescent-shaped
(the concave margin toward center of disk), interspersed with numerous small

@ Challenger Asteroidea, 1889, p. xxxiy.

bI examined the type at the Peabody Museum, New Haven, and another specimen in the U. 8.
National Museum in Washington.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 255

granuliform ossicles on the intervening papular areas; toward margin of disk and
rays papular areas and the secondary ossicles disappear, the primary plates becoming
elliptical, then roundish; primary plates arranged in fairly regular series parallel
with the median radial, and closely beset with numerous (upwards to forty) small
truncate or round-tipped granules which when covered with the natural epidermis
are about as high as broad; the secondary ossicles scattered on the papular areas
(upward to twenty-five on large areas of rays) bear a paxilliform group of smaller
granules (five to twelve); in life and alcoholic specimens the primary plates and
intervening ossicles with their granules are closely packed together, leaving room
only for the papule (upward to twenty to an area on rays) which diminish in num-
ber toward margin, disappearing entirely before the marginal plates are reached.
Marginal plates about equal, the inferomarginals defining ambitus; superomar-
ginals slightly larger than adjacent abactinals. Actinal intermediate plates in
very regular chevrons bearing each a webbed comb of three to five more or less
flattened, sometimes chisel-tipped spinelets, somewhat longer than the base line of the
series; usually a large comb just back of mouth plates. Adambulacral plates with
three partly webbed tapering, membrane-invested furrow spinelets, two longer
than near-by actinal spinelets and the adoral member of series usually markedly
shorter than rest; on actinal surface (which is raised above furrow margin) is an
oblique (sometimes almost transverse) series of three (two to four) stouter chisel-
tipped spinelets (webbed at base). Mouth plates large (free margin equal to three
succeeding adambulacrals), bearing five marginal partly webbed spinelets graduated
to an inner actinal spine, about as large as the tooth. Madreporic body fairly
large, with irregular, branching centrifugal stri.

Color in life-—Color extremely variable; some shade of red, clear or mottled
with other tints, is common for the abactinal surface. Some of the variations
found in specimens from Monterey Bay are listed.

1. Red predominating.

a. Clear vermilion above; below, cadmium yellow or orange near ambitus,
shading into light straw yellow; tube feet raw sienna.

b. Carmine above and below.

c. Very dark dull burnt carmine above, Naples yellow below.

d. Above, very dark purplish Indian red, the same actinally.

e. Above, dark purplish Indian red, mottled with orange or with light yellow
ochre; light Naples yellow below.

jf. Very dark red, shading to grayish green on radii and center of disk.

2. Yellow predominating.

a. Dull light yellow ochre above, sparsely spotted with very dull dark purple;
very pale yellow below.

b. Dull fawn color with a few spots of deep red; the same below. Sometimes
a single large spot in center of disk.

c. Light yellow ochre, spotted with orange and dark red.

d. Orange above, mottled with slightly darker orange brown; orange encroach-
ing on yellow of actinal surface.

3. Brown and dull greens predominating.

256 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

a. Dull dark olive brown above, with dark reddish spots at center of disk; below
orange around edge, rest light yellow.

b. Dark reddish brown shading to dark grayish green along radii and center
of disk (similar to 1f, but browner).

c. Above, pale dull grayish green; the primary plates are greenish white or
gray, the papular areas dull grayish green; sparsely spotted with very dark green;
papule when extended give the surface a brownish cast, leaving the plates in
sharp relief. Below, pale straw yellow, darker at edge of disk.

4. Miscellaneous.

a. Above, pale greenish, deep cadmium, and orange, mingled with large
irregular spots of very dark olive green on radii and center of disk.

b. Same, but green spotting all over. Even other combinations may be found
at Monterey Bay.

Anatomical notes —Externally the abactinal plates have a crescent form rounded
at either end, and the surface is marked by a slight elevation, also of crescent form,
which bears the granules. The papular areas, outside the central region of disk,
have a convex margin toward edge of ray (corresponding to the concavity of the
plate) and the opposite side is sometimes concave. These areas are everywhere
broken up by small intermediate ossicles, which encroach onto the border of the
primary plates, and form an incomplete irregular network on the areas. Between
the secondary ossicles emerge the papule. Sometimes two or more of these granuli-
form ossicles coalesce to form an irregular rod. In the center of the disk the papular
areas are much larger than elsewhere, containing more papulz and more secondary
plates, and the primary plates are smaller, except in the case of the ‘‘primary
apicals,”’ some of which form a pentagonal ring surrounding the apical area. The
photograph will show the arrangement and form of the plates, which have been
treated with caustic potash and the spines removed.

Viewed from within, the dorso-lateral plates have four lobes by which they
imbricate, the carinals are transversely elongated and nearly oblong, and the adra-
dials are elongated transversely but are triangular. (See pl. 56, fig. 8a.)

The actinal intermediate plates are very numerous, imbricated, and marked by a
slight spiniferous elevation. The inner edge of this elevation or convexity (except
on the small peripheral plates) is marked by two to four slight serrations, trans-
verse to the long axis of plate. The serrations are fairly large, but in very low
bas-relief (fig. 8, pl. 56), and can be seen only when the plates are thoroughly
cleaned. The spinelets are inserted just back of the serrations. The surface of the
adambulacral plates has three similar markings, of which the median is largest.

Variations.—Rays usually five, but occasionally six, and rarely seven; one
seven-rayed specimen (Monterey Bay) has three madreporic bodies. Sometimes
the tip of a ray is bifureate, each branch having a normal furrow and plating. This
species is stellate, but the length of the major radius varies from 1.7 to 2 r; it is
apparently never pentagonal, as is commonly the case with A. pectinifera (or a
variety). The amount of webbing on the actinal spines is variable, as is the apparent
thickness of the epidermis (variability perhaps due to fixation). Some combs are
free nearly to base, others webbed nearly (but not quite) to tips; all have more or

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 257

less conspicuous wings or membranous vanes above the webbing. The points of the
spinelets are frequently less obviously chisel-shaped than in the figures.

Variations in color, already mentioned, are wide.

Type-locality.—Sitka, Alaska.

Distribution.—Sitka, Alaska, all the way to San Diego, California, in suitable
localities, thence to Gulf of California, north of La Paz, Lower California; low
tide to 165 fathoms, on rocks, shells, gravel, and hard sand.

Specimens examined.—Kighty, from the following localities:

Specimens of Asterina miniata examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
2008 ...... Santa Barbara Channel, California. . 31 gray sand, broken shells... U.S. Nat. Mus.
sha sae Cortes Banik, off San Diego, Cali- 26 broken shells.............. Do.
fornia (long. 119° 3’ W.).
2040...... San Pedro Bay, California (lat. 20] Bealcene a ODse ace Do.
33° 36’ N.).
2962...... Off Santa Barbara, California...... 165 sand, stones, coral........ Do.
2965 ...... eee Os ee aac as eee 27 fine gray sand, rocks...... } Do.
? eae Off Anacapa Island, California. -.... 36 gravel, broken shells... ... Do.
Laas North of La Paz, Gulf of California. 33 fine gray sand, broken Do.
shells.
hol eee Monterey Bay, California........... 48 brown mud, rocks ........ Do.
Drake’s Bay, California ............ SSDS ba cee ace so ioe eee mee Do.
San Francisco, California........... Shores Sees sc sesee coon oer U.S. Nat. Mus.,D.S. Jordan.
Seuet OO tere eee renee Nee ca Sane san ae | anew esceeee re acccsameresen el U. S. Nat. Mus., Dr. J. Hor-
nung.
Monterey Bay, California .......... (ee |Saccnaeiean. ca secscu cere nae U. S. Nat. Mus., Albatross,
Dall, Jordan.
San Simeon Bay, California (lat. GUONUN I cose sea cceste ata eenc seas | U.S. Nat. Mus.
35° 38’ N.).

to
4
nm

. Nat. Mus., Rosa Smith.
. Nat. Mus.,C, R. Orcutt.

oy
wm
q
mn

San Pedro, California... Shore 1 U.S. Nat. Mus., H. N. Lowe,
Santa Catalina Island, California...| Shore. 2 U. S. Nat. Mus., W. H. Dall.
Gorda, California.............-....: Shore. 2 Do.
4563 ..<..- Near Santa Cruz, California... ..... 8 1 Albatross, 1904.
San Diego, California............... (%) 3 Do.
Clayoquot Sound, British Columbia.| Shore. 3 Stanford, Dr. C. H. Gilbert.
@ Shore and 8 fathoms. » Beacon 2.

Remarks.—This is one of the commonest low-tide species of the California
coast, being abundant at San Diego, Monterey Bay, and north of San Francisco
Bay. The most favorable situations are sheltered coves where the bottom is
covered with fair-sized bowlders and smaller rocks, overgrown with sponges,
Phyllospadiz, coralline, and bryozoans, and where the interspaces are floored with
coarse sand, broken shells, and coralline. At low tide the animals are found
clinging to the sheltered sides of the rocks, or among Phyllospadix and alge. They
are also found abundantly on hard, sandy bottom, as at San Diego.

This species is closely related to A. pectinifera (Miller and Troschel) from
Japan. The general type of armature is strikingly alike, and the differences are
mainly in details of the skeleton and in color. A. pectinifera seems to vary more

57444°—Bull. 76—11——_17

258

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

widely than does miniata, judging by the specimens in the Stanford University
collection. The principal differences are as follows, large specimens being used for

comparison :

miniata,.

General form stellate with broad rays and curved
interbrachial arcs; R=1.7 to 2 r.

Papular areas large, with about fifteen to twenty
pores on proximal radial regions.

Primary abactinal plates rather well spaced with
fifteen to twenty-five secondary plates on the in-
tervening meshes of proximal radial region.

Medioradial (carinal) series of abactinal plates
and armature usually fairly regular and conspicu-
ous proximally, the crescents being much larger
than adradial series.

Primary groups of abactinal granules of a wide,
crescent form, except toward margins, and with
many granules (about forty on proximal carinals).

Spinelets of actinal combs longer than the base
line of a comb of four spinelets (much longer than
one of three); usually three or four spinelets to a
comb.

Serrate scars on actinal plates larger (difference
apparent on comparison).

Three furrow and two or three actinal adambu-
lacral spinelets; longest furrow and actinal spine-
lets equal to or exceeding two adambulacral plates
with suture.

pectinifera.

General form stellato-pentagonal, to pentagonal
with nearly straight sides (Mororan, Hokkaido); -
R=1.4rtol.dr.

Papular areas small; less than fifteen papule on
proximal radial regions.

Primary abactinal plates more closely placed
with fewer (five to fifteen) and relatively larger
secondary plates on the papular areas of proximal
radial region.

Carinal series of abactinal plates and their gran-
ules not conspicuously larger than adradials, and
not forming a noticeable series (viewed internally
they do, however).

The primary groups of abactinal granules only
exceptionally of the wide crescent form (then disk
is pentagonal) but generally more elliptical, or
bowed elliptical, with fewer granules (about twelve
to twenty on proximal carinals).

Spinelets of actinal combs not longer than base
line of comb of four. Difference in size of actinal
spinelets in the two species more evident on com-
parison of specimens. Usually five or four, @ rarely
three spinelets to a comb.

Serrate scars with fine teeth or bas-relief serra-
tions.

Frequently four furrow and four or five actinal
adambulacral spinelets; longest actinal and furrow
spinelets less than two adambulacral plates with
suture.

Family ECHINASTERID Verrill, restricted.

Echinasteride Verri1, Trans. Conn. Acad., vol. 1, pt. 2, 1867, p. 343.—PERRIER, 1875, 1884,
1894.—ViauiER, 1878.—SLApDEN, Challenger, Asteroidea, 1889, p. 535.

Spinulosa with a reticulate, more or less irregular skeleton formed of small
imbricating plates bearing isolated or grouped spines; with biserial tube feet and
single ampulle; suckers of tube feet without rosette of calcareous reticulum; no
pedicellarix. Interbrachial septa weak, more or less rudimentary.

Remarks.—This family, as originally established by Verrill, comprised the
genera Echinaster, Acanthaster, Mithrodia, and Ferdina. Perrier subsequently (1875)
added Solaster, Henricia, and Valvaster, and placed Ferdina in the Linckiide. Viguier
divided the family, in 1878, into four tribes, the Echinasterine, Mithrodiine, Valvas-
terine, and Solasterine. In 1884 Perrier established the family Solasteride for
Solaster and placed Mithrodia in a distinct family, which, however, he called ‘‘Mithro-
dine.’ Sladen, 1889, maintained the family Solasteride, and placed Mithrodia

@ These numbers refer to specimens 40 mm. or more major radius.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 259

back in the Echinasteride, maintaining the subfamilies constituted by Viguier, and
establishing a new subfamily, the Acanthasterinw, for Acanthaster.

In this report the genera Acanthaster, Mithrodia, and Valvaster have each been
made the type of a distinct family. They have no more in common with the
Echinasterin of Sladen than have some of the genera of the Solasteridw; in fact, are
more trenchantly separated, I think, than the latter. It is highly probable that the
three genera enumerated above are surviving types of once larger groups.

The family Echinasteride as limited in this report is therefore coéxtensive with
the Echinasterine of Sladen’s Challenger Asteroidea, with the addition of genera
since described. The component genera are all included in the subjoined key.
They differ from Acanthaster, Mithrodia, and Valvaster in having (so far as known)
but one ampulla to each tube foot. The other differences are noted in the key to the
families of Spinulosa and accompanying footnotes.

Henricia and Poraniopsis are the only genera which occur in the region covered
by this report. Clark has recorded Echinaster tenuispinus from Puget Sound,* but
I do not believe his specimens came from that locality. The presence of such forms
in his collection as Arbacia stellata, Diadema mexicanum, Echinometra oblonga, Toxop-
neustes semituberculatus, and Clypeaster rotundus would at once arouse strong sus-
picions that two separate lots had in some way become mixed. This is especially
the case since the labeling was evidently poor. Echinaster or rather Othilia tenuis pina
is from La Paz, Lower California.

KEY TO THE KNOWN GENERA OF ECHINASTERID2.

a’. Abactinal and lateral skeleton composed of lobed plates connected by intermediate ossicles
forming a more or less irregular mesh work.
b!. Adambulacral spines or spinelets disposed in transverse series on each plate, or without
regularity; not arranged in a regular double longitudinal series.

c!. Abactinal plates bearing small spinelets, usually in groups, without a mamelon; usually
the skin hiding the plates is not thick and fleshy; the spinelets are never large and
isolated; frequently so small as to be granuliform.

d'. Actinal papulz (i. e. between marginal and adambulacral plates); one or more spine-

lets on furrow face of adambulacral plates..................- Henricia Gray, p. 266.

d@*. No papulz between marginal and adambulacral plates (that is, none on actinal surface).

e'. A sabre shaped spinelet on furrow face of adambulacrals........ Cribraster > Perrier.

GLEN OSEDIDELEb AM ULTO W's ralstewees ore Soe nae te see nn ae oe noe ee Perknaster Sladen.

c?. Abactinal and marginal plates covered with a rather thick skin and bearing simple iso-
lated spines.

aProc. Bost. Soc. Nat. Hist., vol. 29, 1901, p. 329. (See antea, p. 8.)

>I have never seen specimens of either Cribraster Perrier or Perknaster Sladen and include them
here only on the authority of their describers. The generic diagnosis given by Perrier (Stellérides,
Cap Horn, 1891, p. 104) would apply to Henricia sanguinolenta. In the description of C. sladeni how-
ever, it is stated that “‘face ventrale depourvue de papilles respiratoires."’ The genus is founded upon
a single specimen.

Perknaster has been challenged by Bell who calls it ‘‘only an aberrant Echinaster.”’ (The Echino-
derma found off the coast of South Africa. Pt. 2, Asteroidea, p. 253.)

It is worth noting that the distinction used by Perrier in his key to the genera of Echinasteride
(1894, p. 142), of disk small, Henricia, and disk rather large, Cribraster, Perknaster, will not hold, as Henricia
sanguinolenta eschrichtii is sometimes nearly pentagonal. Neither does his statement that in Henricia
the ventro-lateral and marginal plates form series at the same time longitudinal and transverse, always
apply, for Henricia sanguinolenta constitutes an exception. In this species the character is variable.

2°60 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

d', Papular areas very large containing upward to sixty papule; spines large, isolated;
no small spine in furrow; madreporic body ‘‘euplacote,’’ an independent skeletal
piece; skin covering the wide meshed skeleton, thick; two adambulacral spines;
rays short and swollen .....--------------------------- Poraniopsis Perrier, p. 260.

@,. A small spine in furrow; madreporic body ‘‘pseudoplacote” (=the primary inter-
radial or basal plate); meshes of skeleton smaller, and rays frequently long and

slender.
e!. No intermarginal or actinal papule....-...-.-.----. Echinaster Miller and Troschel.
e?. Intermarginal and sometimes also actinal papule.......-.--.--.------ Othilia Gray.

b?. Adambulacral spines arranged in two longitudinal series, but no spinelet deep in furrow;
abactinal plates forming wide-meshed network.

c!, Adambulacral spines in two longitudinal combs of four or five each, disposed in a double
longitudinal series along ray. Abactinal plates bearing compact groups of short equal
upright spinelets imbedded in a membranous mass......-.-------. Plectaster Sladen.

c?, Adambulacral spinelets not in combs.

d'. One furrow, one actinal spine; actinal intermediate spines; inferomarginal spines
single; abactinal spines generally single and large... .--- Poraniopsis Perrier, p. 260.

@. Twoor three furrow spines; one much larger sometimes bi- or trilobed ligulate spine just
behind them; inferomarginal spines three to five, borne on a horizontal lamina;
abactinal plates bearing single spines or groups of two or three. ......-......--..--
Dictyaster Alcock and Wood-Mason.

a2. Abactinal and lateral skeleton composed of quadrilateral plates united at their angles by
imbrication. Four adambulacral spinelets along aboral edge of plate; rudimentary spine-
let within furrow; actinal intermediate and inferomarginal plates with three or four aboral
spines; dorsal surface with very small, short, blunt spines forming irregular longitudinal
and transverse rows borne upon edges of plates; ventral and lateral plates connected by
an intermediate ossicle; intermediate ossicles occasionally occur between the other plates
olskeleton: ...2¢ s2 emia << ee niaeelsaectse eee te Shape COAG PARR eae Tee Thyraster Ives.4

Genus PORANIOPSIS Perrier.

Poraniopsis Perrier, Compt. rend., vol. 106, 1888, p. 763 (nomen nudum); Mission scientif.
Cap Horn, IIT, Stellérides, 1891, p. 105, pl. 10, fig. 2a, 2b. Type, P. echinaster Perrier (erro-
neously P. echinasteroides on plate, Bay of Nassau, 95 meters).—FisHER, Zool. Anz., vol. 35,
March 29, 1910, p. 568.

Lahillea pe Lorton, Notes pour servir 4 1’étude des Echinodermes, sér. 2, fasc. 2, 1904, p. 32, pl. 3,
fig. 1, la-lg. Type, L. mira P. de Loriol (Gulf of San Mathias, Argentina, low tide).

Alexandraster Lupwie, Mem. Mus. Comp. Zodl., vol. 32, July 17, 1905, p. 210, pl. 15, figs. 79, 80;
pl. 31, figs. 181-183; pl. 32, fig. 184. Type, A. mirus Ludwig (Gulf of Panama, 458 fathoms).

Ortmannia de Lorton, Revue critique de paléozoologie, Paris, vol. 10, Jan. 1906, p. 78. Proposed
to replace Lahillea preoccupied.—Batuer, Revue critique de paléozoologie, Paris, vol. 10,
1906, p. 131.

Lahillia de Lortou, Revue critique de paléozoologie, Paris, vol. 10, 1906, p. 77.

Diagnosis.—Echinasteridx with wide-meshed, reticulate skeleton bearing at the
nodes, upright prominent single spines; with large abactinal and intermarginal
papular areas (no actinal areas); with two adambulacral spines, but none in furrow
(as in Henricia and Echinaster); whole body overlaid with rather thick skin, mount-
ing the bases of spines, and sometimes investing them; no pedicellarie; madreporic
body an independent plate (‘‘euplacote,”’ Ludwig); tube feet with strong sucking
disks and single ampulle.

« Thyraster Ives, type Echinaster serpentarius Miller and Troschel. Echinoderms from the northern

coast of Yucatan and the harbor of Vera Cruz. Proc. Acad. Nat. Sci. Phila., 1890, p. 329, pl. 8, figs. 19,
20, 21.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 261
KEY TO THE SPECIES OF PORANIOPSIS HEREIN DESCRIBED. —

Adambulacral spines with a groove; rays rigid and skeleton stout; adradial row of intermediate ossicles

ATAU BINONEN Ueeen eerioe See t SaL eE tS. 3), Meee eee ive kee hae Cae ee inflata.
Adambulacral spines tapering without a groove; rays weak and fle exible; adradial series of intermediate
FeSO PSNONIO WINES A OBEN bas. jee nchw lease c aa 0 eam oer duc RRO Moan cen acacia been flexilis.

Remarks.—This genus differs from Echinaster in lacking a spinelet on the furrow
face of the adambulacral, and in having the madreporic body an independent plate.
The skeleton is very much more open than in either Echinaster or Othilia, the reticu-
lations being comparatively few and large. Ludwig (1905, p. 210) mentions as
another difference from Echinaster the presence of intermarginal papular areas, but
this will not hold unless Othilia is recognized as independent, since Echinaster sentus,
spinosus, and spinulosus have intermarginal papule, and spinulosus has actinal
papule. The type of Othilia is Asterias spinosa Retzius, 1805 (Echinaster spinosus
Miller and Troschel, 1842) Brazil.

I think a comparison of descriptions and figures will convince any one that
Perrier’s Poraniopsis, de Loriol’s Lahillea, and Ludwi ig’s Alexandraster are the same
genus. At any rate there are absolutely no differences of generic importance,
Alexandraster mirus being even specifically close to Poraniopsis echinaster. Perrier’s
comparison of Poraniopsis with Porania is certainly misleading for the resemblance
is most superficial. The name, Poraniopsis mira (de Loriol, 1904) invalidates
P. mira (Ludwig, 1905). The latter is doubtfully distinct from my flerilis.

I did not see de Loriol’s paper until after I described “‘ Alexandraster’’ inflatus
(June 1906) when the identity of the two genera became evident at once. De
Loriol published several good figures. A further search brought to light Poraniopsis,
which so far as I know is the oldest name.

PORANIOPSIS INFLATA (Fisher).

Pl. 58, figs. 7, 7a; pl. 63, figs. 1, 2; pl. 112, fig. 1

Alexandraster inflatus Fisner, Zool. Anz., vol. 30, June 19, 1906, p. 300.

Poraniopsis inflata FisHer, Zool. Anz., vol. 35, March 29, 1910, p. 569.

Diagnosis.—Rays short, robust, evenly tapered or slightly inflated at middle,
bluntly pointed; disk capable of considerable inflation; whole animal appearing
inflated and turgid. Type, R=60 mm.; r=23 mm.; R=2.6 r; breadth of ray at
base, 27mm. Skeleton forming large squarish or polygonal areas (occupied by very
numerous papule), the whole covered by a rather thick skin mounting the base of
the upright, usually sharp, prominent spines which stand at the corners of the
papular areas; counting around the ray from inferomarginal to inferomarginal
series inclusive there are seven more or less regular longitudinal rows of spines in
type but as many as thirteen in a few large specimens, the extra spines being shorter
and forming rows interpolated between the primary series; adambulacral plates
with two prominent grooved spines in a transverse series; actinal intermediate
spines, often grooved; no actinal papule.

Description.—Abactinal and lateral skeleton composed of three or four lobed
primary plates connected by several elliptical oblong overlapping ossicles forming a
large mesh, inclosing spacious subquadrate and irregular papular spaces, which

262 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

contain on the disk of large specimens about fifty to sixty-five conspicuous papule,
these decreasing gradually on the rays to about thirty, then fifteen, then ten;
smaller specimens with fewer papuls. No papule on actinal surface. Skin rather
thick, obscuring the outlines of plates which form narrow raised ridges, the isolated
spines occurring on the larger raised primary plates, usually at a point where three
or four ridges join; in other words at corners of papular areas. The spines are
borne on a slightly raised boss of plate, boss and spine together being 3 to 6 mm. in
length. Spines stout, tapering, usually rather sharp, in small specimens being
arranged in a mid-radial, superomarginal and inferomarginal series, with a row of
scattered adradials and one of actinal intermediate smaller spines. There is also a
large primary basal spine in each interradius and a large central spine. The primary
“radials” and ‘‘basals’’ are connected by a ridge so as to form a pentagon, the
corners radial. Within this apical area is another more stellate pentagon with the
primary basals for corners and the infra-basals for interadii. Connecting the infra-
basals with the central plate are five straight ridges, which also form interradial lines
to the inner pentagon. The infra-basals have short spines and occasionally one
of the ridges radiating from the central plate is missing. These ridges are of course
radial with reference to ray, since the points of the inner pentagon are interradial.
In mature specimens this regular arrangement is more or less disturbed, although the
two pentagons, one within the other, can be readily recognized in the apical region.
In large specimens the radial series of spines is usually irregular. There is some-
times an intermediate series of shorter spines between the median radial and adradial,
and one between adradial and superomarginal; thus three laterodorsal series, of
which the middle (the adradial of young specimens) is more regular and has longer
spines; one or two irregular intermarginal series, shorter than marginal spines;
inferomarginal spines slightly grooved along one side at tip, or simply flattened and
truncate; occasionally not perceptibly flattened but blunter than upper series.
Actinal intermediate spines often entirely wanting, even in adult specimens, and
when present are arranged in one to four very irregular and often incomplete series
which do not extend beyond middle of ray; sometimes only a few irregularly
scattered spines are present. Fine creases or lines in the integument proceed out-
ward between the spines, from the adambulacral to marginal plates. Lowest papular
areas are the longitudinal intermarginal series, containing upward to twenty-five
papulz to the area.

Preparation of the body wall shows that the papular areas contain many small
scattered calcareous grains, most numerous near central portion of each area. In
some areas these are arranged in lines and appear to be embryonic plates, while in
other areas a few have enlarged sufficiently to form isolated chains of small plates or
have even formed an attachment with the nearest trabecula.

Adambulacral plates considerably wider than long, rather small and crowded.
Armature consists of (1) a single truncate or round-tipped untapered or tapering
furrow spine, 3.5 to 4.25 mm. long, usually flattened and grooved along the upper
(or outer) side; (2) on the actinal surface of plate, a similar but generally longer
upright spine, usually very conspicuously scoop-shaped at tip. Both spines are
covered with membrane, as is also the surface of plates. Some specimens have
outer furrow spines longer and slenderer than others or relatively longer than the

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 263

furrow spines, or more tapered and less conspicuously grooved. The inner or furrow
spines vary in the extent and depth of the groove.

Mouth plates with about four marginal spines (rarely five) decreasing in length
toward the first adambulacral, the inner or largest being about the same size as a
furrow spine and all similarly grooved; one spine on the actinal surface of plate of
same size as corresponding adambulacral spines.

Anal opening surrounded by numerous slender papilliform spinelets.

Madreporic body situated at about middle of minor radius, prominent, with
more or less sinuous centrifugal striations.

Color in life, general tint cream color, abactinal surface lighter, but apical area
buff yellow. In some specimens buffy spots are scattered here and there over
surface of body.

Anatomical notes.—Stomach spacious; hepatic or radial cceca large, reaching
nearly to end of ray; intestinal ceca large, with numerous small lobes, forming a
mass not unlike an ovary in appearance. In the example dissected there are
eight or nine of these diverticula lying over the hepatic ceca. Centrally they open
into a rather spacious chamber separated from the stomach by a small orifice;
dorsally the chamber opens by the anus; the diverticula are about one-third or
one-fourth R in length. Retractors of stomach large and strong. One Polian
vesicle in each interradius, except that of the madreporic canal; canal slender,
sinuous, with easily distinguishable rings of calcareous matter in its walls, reminding
one of the trachea of a bird; tube feet with strong sucking disks, and single ampulle
(not two-parted or divided as described by Ludwig in Alexandraster mirus; this is
a mistake, I think). The ampulle are rather large and form a double series along
either side of the ambulacral ridge, but there is only a single ampulla to each orifice;
possibly this double series gave the erroneous impression that the ampulle are
double in Alexandraster mirus. Interradial septa membranous, the gonads being
attached to the dorsal surface on either side of it, and in the interradial angle.
First ambulacral ossicle forked at bottom and heavier than the succeeding ones.

Variations.—This is a variable species in respect to the length, number, and
sharpness of the abactinal and lateral spines, those of the type being stouter and less
numerous than in others from Monterey Bay and San Diego, California. At the
same time there are specimens from the latter locality which are nearly or quite
typical and match examples from Monterey Bay. The type has the grooving of
the adambulacral and actinal spines more pronounced than in some other speci-
mens from both Monterey Bay and off San Diego. In these the spines are slenderer
and the groove instead of running to the base runs only part way down the spine,
or is restricted to the tip, and the depth of the groove varies, being very shallow
in a few examples. In the type the two adambulacral spines are about equal in
length, near the base of ray, the inner shorter distally. More than half of the
specimens have the inner considerably shorter all along the furrow; as, for instance,
a large specimen from station 4370 has outer spine 5.50 mm, and corresponding fur-
row spine 4.25 mm. The thickness of the skin is slightly variable, and some exam-
ples have the rays more inflated and rigid than others. Certain other variations
have been considered in the description and need not be repeated.

264 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Young.—The smallest specimen (station 4361) has R = 8.5 mm.,r = 4 mm,,
and is easy to identify when one is acquainted with the adult. There are three
longitudinal rows of abactinal tubercles, in place of spines, and a superomarginal
row of tubercles. The inferomarginal plates bear a short stout spine or two.
Intermarginal papul two to an area; abactinal, one to four. Adambulacral spines
prominent, without grooves. Marginal plates prominent, especially the inferomar-
ginals which encroach conspicuously onto the actinal surface; the creature is
strongly phanerozoniate. Only one or two actinal interradial spines, and only a
few plates. An example from 2893 with R = 12 mm. is much like the above.
The characters of the species appear first in a specimen with R = 15 mm. and are
pronounced in one with R = 25 mm.

Parasites.—A specimen from 4471 contained a Dendrogaster, perhaps of the
same species as that figured for Hippasteria californica. The starfish has weaker
rays and a generally flabbier appearance than two other specimens from the same
station. The Dendrogaster was found in the body cavity, its branches extending
into two rays. The gonads of the starfish are almost rudimentary, although the
major radius measures 60 mm.

Type.—No. 309, Stanford University collection (invertebrate series).

Type-locality.—Monterey Bay, near Pacific Grove, California, off shore in
about 60 to 70 fathoms, possibly considerably shallower. The type was caught
by Chinese fishermen.

Distribution.—Oregon to San Diego, California (and undoubtedly southward), 26
to 159 fathoms, apparently gradually changing in deeper water to the following race.

Specimens examined.—Twenty-seven, from the following stations:

Specimens of Poraniopsis inflata examined.

Station. | Locality. Depth. Nature of bottom. Number. Collection.

Fathoms. |
S86c5~-~ | Off Heceta Bank, Oregon ...........-..-- 50 LOCK Yio 'e on see eee 1 U.S. Nat. Mus.
2893......| Santa Barbara Channel, California.......- 145 fine gray sand, mud..-.. -=--| 2 Do.
2946. ....- | Off Santa Cruz Island, California ......-. | 150 coarse gray sand...........-- 2 Do.
ASTI = | Near San Diego, California................ 110-143 | green mud, fine sand-.......- 1 Albatross, 1904.
4361 a 97-91 | gray sand, mud............- | 1 Do.
a3622 2 == |aeae CO ala ieee ae ee | 100-159 | fine gray sand.- | 3 Do.
AS TO Rees =o ls ORS eee ce a ee set ae are Ae 99-147 | gray sand, rocks. | 2 Do.
ATi pater secee ORS csenc cess secede eae eee 127 green mud, sand...........- | 1 Do.
44712. -.52 Monterey Bay, California...........-..... | 144-65 | gray sand, rocks...........-. 3 Do.
71-54 very hard sand...........--- 1 Do.
56-46") \TOCKS-< oe eee ee eens 1 Do.
83-66 | green mud, rocks.........-.- 2 Do.
74-65 | rocks..........-... Seal 1 Do.
60-81 | green mud. .| 1 Do.
66-69 |... do. PP. ae Do.
47-26 TOCKS: seed .| 1 Do.
Seeeniee seen see eee eee eta | 2 Stanford University.

Remarks.—I1 have examined a large Poraniopsis from off Honshu, Japan, 182
fathoms, which differs in a few details from inflata and probably represents a
different though closely related species. The rays are shorter and the disk larger

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 265

than in inflata, the rays being decidedly broader at the base; R=2 r. The
abactinal spines are numerous, sharp and not arranged in definite rows (except
marginals), though on one ray three longitudinal series can be made out between
the carinal and superomarginal series (as in extreme examples of inflata); no inter-
marginal spines; adambulacral spines two, round or truncate at tip, flattened and
grooved. In proportion to the size of the animal they are longer than in inflata.
Inflata differs, therefore, from the unnamed Japanese species in having a smaller
disk, narrower rays, more or less constricted at the base, more regular and usually
fewer abactinal spines, intermarginal spines fairly well developed, shorter adam-
bulacral spines, and numerous grains in the integument of papular areas. The
two species are close, in fact look more alike than inflata and flerilis, and may be
found to intergrade over the intervening territory, although it is not probable,
since all the work in Alaska has failed to bring any specimens to light.

From P. echinaster Perrier (equal-sized) specimens of inflata differ in having
longer narrower rays of a different form (echinaster resembles the Japanese species),
grooved adambulacral spines, a single (not double) inferomarginal spine to each
plate, and longer and more regularly disposed abactinal spines.

Poraniopsis mira (de Loriol)—formerly Lahillea mira and Ortmannia mira—
has the grooved furrow spines of inflata, but the abactinal spines have a flattened
often capitate top and are quite short and tubercular; and the ventral surface, in
addition to the prominent spines, is covered with myriads of minute prickly spine-
lets. As usual, de Loriol’s excellent figures of specific characters render accurate
comparison very easy.

PORANIOPSIS INFLATA FLEXILIS Fisher.

Pl. 58, fig: 8; pl. 64; pl. 112, fig. 2.
Poraniopsis inflata flevilis Visner, Zool. Anz., vol. 35, March 29, 1910, p. 568.

Diagnosis.—Rays longer and slenderer than in inflata, very flexible and weak;
R=72 mm., r=26 mm., R=2.8 r. Breadth of ray at base, 30 mm. Abactinal
surface very weak owing to the slenderer skeletal elements and the larger papular
areas which frequently extend from the mid-radial line to the superomarginal plates
without a break; meshes of actinal skeleton very wide. Spines slender, long, and
sharp; adambulacral spines also slender, tapering, and pointed and without grooves.
No calcareous grains and tiny plates in the integument of papular areas.

Description.—The rays are notably slenderer than in inflata, especially at the
tip; and the integument instead of being rigid is flexible. This is owing to the loose
skeleton and the immense papular areas. In the type the adradial line of spines
and primary plates is present, but the longitudinal series of intermediate ossicles
is wanting so. that the two series of dorsolateral papular areas is merged into one.
In a specimen from 3104, which if not typical seems to be near this race, there
are two dorsolateral rows of papular areas so that this character may not be constant.
The spines are weaker than in inflata and in seven irregular longitudinal series. On
account of the slenderness of the rays distally the actinal intermediate plates do
not extend so far along the ray as in inflata and the meshes of the skeleton are very
large and the trabecule slender and weak. Actinal spines few, slender, scattered.

266 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

There are only a few intermarginal spines near the base of ray where the inter-
marginal papular areas are very wide.

Papulsw very numerous to anarea. There are no small plates or grains embedded
in the integument, as in inflata.

The two adambulacral spines are slender and tapering, some pointed, some
blunt. The outer is the longer as a rule, although both spines are not at all uniform
in length. There is no trace of a groove. Mouth plates with four marginal and
one actinal spine, therefore the same as inflata except in the form of the spines.

The internal anatomy, madreporic body, and anal papille are similar to inflata.
The ambulacral ossicles are slenderer than in inflata. Ampulle single.

Type.—Cat. No. 27778, U.S.N.M.

Type-locality.— Albatross station 4407, off Santa Catalina Island, California,
334-600 fathoms, rocks, shells, fine gray sand.

Distribution.—Central and southern California, in 334 to 600 fathoms; possibly
ranging south to the Galapagos Islands and including Poraniopsis mira (Ludwig)
untenabre.

Specimens examined.—Two, the type and a smaller example from 3104, off
Half Moon Bay, California (lat. 37° 23’ N.) 391 fathoms, coral; this not quite
typical.

Remarks.—The differences which separate this form from inflata have been
noted above. It seems, from a study and comparison of material, that this is a
deep water race of the variable inflata. Some specimens of the latter approach
flevilis. It is very likely that there is a continuous distribution of inflata from shallow
water into deep water and a consequent change of characters. I’lexilis seems to be
variable also, and the presence of a less strongly differentiated specimen from off
central California, north of Monterey Bay, would lend color to the intergradation
hypothesis.

Genus HENRICIA Gray.

Linckia Forses (not Nardo, 1834), Mem. Wernerian Soc., vol. 8, 1839, p. 120.

Henricia GRAY, Ann. and Mag. Nat. Hist., vol 6, 1840, p. 184. Type H. oculata Gray= Asterias
sanguinolenta O. F. Miiller—Bertt, Ann. and Mag. Nat. Hist., ser. 6, vol. 6, 1890, p. 472.
Since then Bell, Grieg, Scott, Fisher, Clark, and others.

Cribella Fores (not Cribrella Agassiz, 1835), British Starfishes, 1841, p. 100.

Echinaster MULLER and Troscuet, System der Asteriden, 1842, p. 22 (part). For some reason
Echinaster has persisted in the writings of Storm 1878, 1879, Stuxberg 1878, 1886, Hoffmann
1882, Jarzynsky 1885, Aurivillius 1886, Brunchorst 1891, Appell6f 1892 (see Ludwig, Fauna
Arctica, p. 473).

Cribrella LiivKen, Vid. Medd. for 1856, 1857, p. 93; Grénlands Echinodermata, 1857, p. 31.
Most authors since then except Gray 1866, Bell 1890, 1892, Grieg 1896, 1898, Scott 1897, Fisher
1906, 1910, Clark 1909.

Diagnosis.—Kchinasteride: with a more or less close meshed skeleton bearing
numerous small spinelets either in groups or scattered along the ridges composing
the skeleton; marginal plates more or less distinguishable; adambulacral plates with
one or more spinelets placed deep within the furrow.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 267
KEY TO NORTH PACIFIC FORMS OF HENRICIA.@

a‘. One furrow spinelet, except sometimes on the distal portion of ray, where two may be present on the
furrow face of the adambulacrals.

b'. All the pseudopaxille or groups of spinelets small, rather close set, the abactinal skeleton not form-
ing an open mesh work; pseudopaxillz with comparatively few (one to fifteen) short spinelets;
papule few (one to three, less often four or five) to an area.»

c'!. Rays moderately long and slender; marginal plates typically inconspicuous, and not normally
forming regular series as in eschrichtii (but in some varieties marginal plates are present, when

Ris considerably greater than 3r and size is frequently larger). ......... sanguinolenta, p. 271.
c*. Rays short, stout, and often very tumid; marginal plates with few spinelets, forming regular
REMEGse een Sense she oe Solace es desea nce bite ae we eee eae eschrichtii, p. 276.

6*. Abactinal pseudopaxille moderately to very large, close-set, with numerous small delicate, some-
times granuliform spinelets; papular areas small, typically smaller than the pseudopaxille;
marginal and actinal intermediate plates forming three very regular, conspicuous, usually
juxtaposed series.

c!. Abactinal pseudopaxille typically with small granuliform or clavate spinelets; adambulacral
spinelets few in about two transverse series. Typically one spinelet in furrow except at very
tip of ray where two may be present.

d'. Papular areas smaller than adjacent pseudopaxille which have usually more than twenty
spinelets; marginal plates squarish, large, not much wider than long, not separated by
prominent transverse grooves (as wide as the elevated ridge of plate); actinal intermediate
series reaching quite or nearly to tip of ray ..............-..-.---+---+-- leviuscula, p. 280.

d*, Papular areas usually larger than pseudopaxille, much sunken, the latter elongated with
fewer than twenty spinelets arranged on the tabulate elevation in two or three rows;
marginal plates much compressed, the thin spine-bearing tabula separated by fairly
deep grooves; actinal intermediate plates not distinguishable much beyond middle of
TAY~= -- -- ener ee ee eee cence nee cence mec e este cece nceneee annectens, p. 291.

c?. Abactinal pseudopaxille with numerous slender often glassy spinelets ending in three to several
long sharp awns or points; adambulacral spinelets numerous (twenty-five to fifty); typically
with two spinelets in furrow beyond middle of ray but only one proximally; actinal inter-
mediate plates extending one-half to three-fourths length of ray .......... multispina, p. 286.

bd’. Abactinal skeleton more open, with large often sunken papular areas containing numerous papule.

c!. Spinelets scattered along the ridges surrounding papular areas, often quite small; not in
definite clusters or pseudopaxillee; adambulacral spinelets in a single, sometimes zigzag,
transverse series.

d'. Rays slender, long, and rigid, the skeleton forming an open mesh work with large sunken
papular areas; abactinal spinelets minute, generally immersed in the integument and
arranged along the trabecular ridges in one to three irregular series; adambulacral spinelets
in zigzag transverse series, few, thick, stubby, and immersed in membrane; marginals
regular with large intermarginal papular spaces. . . aspera, p. 293.

d*. Rays weak; spinelets comparatively large (as lori 2 as the height of ridge bearing them) well
spaced, in a single series along the skeletal ridges. Adambulacral spinelets slender,
long, in a transverse series, webbed basally, but not impeded by a thick investment.
MaYoIni alk NOL dL Wihys TOR LIAL ae seater ds fa docbet Gunes as ene sede e mas asthenactis, p. 297.

EEE

a Owing to the variability of the species of this genus identification by a key is unreliable. The
accompanying ‘‘attempt”’ will serve to place a specimen near its proper place in the system when recourse
must be had to the figures and descriptions. There is a small percentage of specimens that are very
difficult to classify even with ample material for comparison; such are usually intergrades, ‘ freaks,’’ or
hybrids.

+ For California specimens which seem to belong here, see also //. leviuscula dyscrita and young or
dwarfed leviuscula.

268 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

c?, Spinelets in definite pseudopaxille or in fascicular groups.
d', Adambulacral spinelets less than twenty.
el. General appearance resembling leviuscula with small pseudopaxille, papular areas not
unusually large, often slightly larger than plates; spinelets not in spaced fascicular
groups, but in definite pseudopaxille; no intermarginals.

/. General form intermediate between aspera and leviuscula; papular areas usually small;
pseudopaxillee oblong with two or three rows of upward to twenty very short stubby
spinelets; prominent wide and short (compressed) marginals; no intermarginals;
adambulacral spinelets short, usually in two transverse series; only one spinelet in
furrow allalongirayo=.s--e eee eae ee eee eee eee eee annectens, p. 291.

f?. General form intermediate between leviuscula and multispina, but nearer leviuscula;
abactinal pseudopaxillee small, often smaller than papular areas, and spinelets
numerous, delicate, ending in several minute points; actinal intermediate plates
ending near but not at tip of ray; two furrow spinelets near tip of ray; marginal plates
not conspicuously compressed, with wide sutures between ...... dyscrita, p. 289.

e?. General appearance not at all like leviuscula; papular areas fairly to very large; spinelets
fairly to unusually long, in spaced fascicular groups; a long series of intermarginals;
marginals not wider than long, armed with a cluster of comparatively long spinelets;
adambulacral spinelets unusually long in a zigzag transverse series.

J‘. Papular areas not unusually large, about one to two times as wide as the intervening
skeletal pieces; actinal interradial areas very small; secondary plates or calcareous
grains in papular areas, when present, few.......-.------------ longispina, p. 299.

J/?. Papular areas unusually large, the skeletal pieces slender and forming a wide mesh
bearing spaced fascicular groups of spinelets; papular areas containing numerous small
calcareous grains, often with a single spinelet; large actinal interradial areas containing
numerous plates and papule; large intermarginal papular areas. ..-. aleutica, p. 300.

@, Adambulacral spinelets more than twenty; abactinal plates rounded or crescentic (with

very numerous, delicate spines) and much less crowded than in leviuscula; adambulacral

spineletsi twenty-five tothirty.o.- 2. -ceseeee eee eee eee eee spiculifera, p. 295.

a. Regularly two or more spinelets on the furrow face of the adambulacral plates all along ray; adambu-
lacral spinelets numerous; rays flexible.

b!. Abactinal and lateral pseudopaxille small, evenly spaced; papule single, arranged in fairly regu-
lar oblique transverse series on either side of the median radial line (at least when viewed
internally) .22iishas Seis d= Seek laa! hare ee ee ee ee ee ees polyacantha, p. 302.

b?. Plates small forming a sinuous irregular mesh work inclosing smaller, lower plates also more or less
joined in a close mesh between which emerge single papule, which do not form transverse
Ob]igue tows: susie \aiedd ole vel ote shane aa eRe Se ee ee clarki, p. 303.

Remarks.—A little over one thousand specimens have been examined in prepar-
ing the following account of North Pacific species of Henricia. Probably no better
collection exists, at least for a region of similar extent. Although well acquainted
with the variability of starfishes, I have never before met with such an extreme
example as the species of this genus present. All attempts at classification are to be
regarded as provisional, but in the case of Henricia especial emphasis should be
directed to this fact. The more diverse individual and other variations become, the
more difficult it is to limit precisely the species of a genus, and the more must per-
sonal equation and fallible judgment figure in the final decision. For this reason it
is probable that another worker, with the same collection would differ considerably
in his results. The problem resolves itself into observation and interpretation of
not very obvious evidence—evidence in which many important gaps are present
and evidence which may be construed in several different ways.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 269

So great and numerous are the variations in most of the species recorded below
that each is to be regarded more as a center of variation, deviations from the type
proceeding in many directions until they meet and often merge with aberrant mem-
bers of nearly related forms. This will very likely hold true in many other genera
when enough specimens from numerous localities are examined. Another way of
stating the idea (which is not new) is this: So-called species of starfishes intergrade
in various ways (geographically, bathymetrically, by individual variation, and per-
haps also by hybridization) with neighboring and sometimes also with accompanying
forms. Many species seem well marked only when we have a few specimens.
Furthermore, wide differences in outward structure form no barriers to such inter-
gradation, which may be accomplished from one species to another through a series
of intermediate species or races.

With deliberate inconsistency I have called the well-marked types species, dis-
regarding certain indications of intergradation between sanguinolenta and leviuscula,
and between leviuscula and aspera; and I have called subspecies certain small species
which are obviously closely related to and were found to intergrade with sanguino-
lenta and leviuscula. Intergradation seems to take place between sanguinolenta and
eschrichtit and leviuscula; between eschrichtii and leviuscula; between leviuscula and
multispina; between multispina and eschrichtii (not so evident); between leviuseula
and annectens; between annectens and aspera. Certain of these cases I feel sure are
due to hybridism When two or more species range together, ‘‘freaks” and aberrant
specimens immediately make their appearance. It is sometimes impossible to
classify these; they are very baffling. Since hybridism is possible and in some cases
probable, it has seemed better to keep such distinct types as leviuscula and aspera
separate from sanguinolenta, and to rank under each the races which are evidently
most nearly related to them. Thus we have Henricia leviuscula annectens and
Henricia leviuscula multispina rather than Henricia sanguinolenta annectens and
Henricia sanguinolenta multispina, which would obtain if the custom of ornithologists
were followed. I am not convinced that the terrestrial subspecies has an exact
counterpart in the sea, nor the extreme and baffling variation of the Asteroidea
a counterpart in any class of vertebrates.

If the supposed relationships were to be fully expressed in the name we should
be obliged to denominate the first six species somewhat as follows: (1) Henricia
sanguinolenta; (2) Henricia sanguinolenta eschrichtii; (3) Henricia sanguinolenta
leviuscula or Henricia sanguinolenta eschrichtii leviuscula; (4) Henricia sanguinolenta
leviuscula multispina or Henricia sanguinolenta eschrichtii leviuscula multispina; (5)
Henricia sanguinolenta leviuscula annectens; (6) Henricia sanguinolenta leviuscula
annectens aspera. This will indicate why for practical purposes it is better to keep
the well-marked types separate, even if evidence of intergradation is present. The
subspecies, races, or varieties are in all cases simply deviations from the type. Hen-
ricia sanguinolenta annectens is strictly what should be written, but if the form in
question has obviously nothing to do with sanguinolenta and is as obviously related
to leviuscula, I believe the course followed to be the best, even if it is ultimately
proved beyond peradventure that sanguinolenta and leviuscula intergrade without
hybridization. Similarly I think it better, if leviuscula, aspera, and sanguinolenta
constitute well marked and well differentiated types to keep them separate rather

270 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

than to subordinate them by a trinomial designation. The study of this collection
of Henricia has strongly suggested the possibility that all the species of the genus
are connected by intergrades, not serially, but by numerous, often anastomosing,
lines. The possibilities are somewhat indicated by the puzzling interrelationships
of the North Pacific forms.

If the number of species and races recognized seems large, it should be noted that
there are probably still more awaiting a name. Three problematical forms were
united with sanguinolenta, because the evidence of their distinctness was not suff-
cient or because they seemed scarcely more than extremes of individual variation
(possibly mutants or hybrid offspring); several local varieties or variant forme
of leviuscula and multispina are unnamed; H. asthenactis is divisible into slightly
differentiated northern and southern races, which seem too insecure to name at
present; and H. clarki is possibly composed of two very closely related species. It
may be added that the course followed was the best according to our lights. The
difficulties which each species presented are noted under that species.

Perhaps the sum of this rather lengthened apology for what follows is this: That a
system of nomenclature perfected for a rather limited set of animals (the higher
vertebrates) may not so well meet the requirements of a different class of creatures
which have evolved under very different conditions, and have been subjected to
possibly fewer, perhaps more numerous modifying factors. If this system of
nomenclature, therefore, is not rigorously adhered to, but is just a trifle altered in
order to make it less obviously the awkward instrument that it is, no serious
critcism should follow.

Three species of Henricia have been described from the region between San Diego,
California, and the equator. H. minuta® (Bell), H. gracilis’ (Ludwig), H. nana‘
(Ludwig).

The first is from Ecuador and probably a young specimen; nana is from the
Galapagos, 53 fathoms, and gracilis from oft the Tres Marias Islands, Mexico, and
from near Chatham, Galapagos, 146 to 680 fathoms. The small size of two of these
makes it difficult to compare them with the northern forms, especially as the figures
of the entire animal are not sufficiently enlarged to render much assistance. None
of them fall under the second division of the appended key, in which two or more
spinelets are present on the furrow face of the adambulacral plates. H. gracilis
resembles in the formation of the skeleton H. asthenactis, but its spinulation is of the
type of H. leviuscula duscrita. It is apparently distinct (at least subspecifically)
from this form. Without specimens for comparison a consideration of the rela-
tionships would be of little value, on account of the variability of the species of this
genus. HH. nana, however, seems to be nearest H. asthenactis or H. aspera. The
variety of H. clarki recorded from off the Revillagigedo Islands, in the following
pages, is entirely distinct from the species described by Ludwig, although I do not
feel very certain that the form will not ultimately be found to be a distinct species,
when more and larger specimens are obtained.

@Proc. Zool. Soc., 1882, p. 122, pl. 6, fig. 3.
bMem. Mus. Comp. Zo6l., vol: 32, 1905, p. 204, pl. 13, figs. 67, 88; pl. 30, fig. 178; pl. 31, fig. 179.
cIdem, p. 207, pl. 13, figs. 65, 66; pl. 31, fig. 180.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. Qi

Henricia spatulifera Verrill (1909b, p. 554, fig. 6) from southeast Alaska has not
yet been described, so that it is not included in this report. I have not recognized
it in the collections, although the figure of the adambulacral and marginal plates is
not alone sufficient for purposes of identification. Both H. leviuscula and H.
leviusula multispina may have spatulate adambulacral spinelets.

HENRICIA SANGUINOLENTA (O. F. Miiller).
Pl. 65, figs. 1, 2; pl. 66, figs. 1-5; pl. 68, fig. 3.

Asterias sanguinolenta O. F. Miitier, Zoologisee Danicee Prodromus, 1776, p. 234.
Asterias pertusa O. F. Miter, Zoologize Danicwe Prodromus, 1776, p. 235.
Asterias oculata PENNANT, Brit. Zool., vol. 4, 1777, p. 52, pl. 30, fig. 56.
Asterias spongiosa Fasnicius, Fauna Greenlandica, 1780, p. 368 (perhaps refers to eschrichtii but
is practically a nomen nudum).
Asterias seposita Retzius, Nya Hand]. Kongl. Svenska Vet.-Akad., vol. 4, 1783, p. 237.
Linckia oculata Forpes, Mem. Wernerian Soc., Edinburgh, vol. 8, 1839, p. 120, pl. 3, fig. 5.
Henricia oculata Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 184.
Cribella oculata Forses, A History of British Starfishes, etc., 1841, pp. 100, 106, fig. p. 100.
Echinaster oculatus MUtrer and TroscHeL, System der Asteriden, 1842, p. 24.
Echinaster sanguinolentus M. Sars, Archiv f. Naturg., Jahrg. X, vol. 1, 1844, p. 169; Fauna
Litt. Norvegi, 1 Heft, 1846, pp. 47-56, pl. 8, fig. 1-37.
Echinaster sarsti MULLER and Troscuet, Archiv f. Naturg., Jahrg. X, vol. 1, 1844, p. 178.
Linckia pertusa Stimpson, Invert. Grand Manan, 1853, p. 14 (possibly H. pectinata (Verrill)).
Cribrella sanguinolenta LUTKEN, Vid. Medd., 1857, p. 93.—Lupwic, Fauna Arctica, 1900, p. 472
(extensive synonymy and list of localities; includes eschrichtii under sanguinolenta).
Cribrella oculata P. Fiscuer, Actes soc. linnéene Bordeaux, vol. 27, 1869, p. 366.—Perrnrier,
Révision des Stellérides, 1875, p. 109.—Stapen, Challenger Asteroidea, 1889, pp. 542-543,
and numerous other authors.
Cribrella oculata, var. murina Perrier, Révision des Stellérides, 1875, p. 112 (Greenland).
Cribrella oculata, var. cylindrella SuaDEN, Trans. Roy. Soc. Edinburgh, vol. 27, 1883, p. 160, pl.
26, fig. 8.
Henricia sanguinolenta GrieG, Bergens Mus. Aarbog, 1894-95, No. 12, 1896, pp. 8, 12.—Fisner,
Zool. Anz., vol. 35, March 29, 1910, p. 569.
Nore.—The above is not intended to show the complete synonymy of the species, but only the new
names and new combinations. For an extensive list of citations consult Ludwig, Arktische Seesterne,
Fauna Arctica, vol. 1, 1900, p. 472.

This widely distributed and variable species is represented in the region of Bering
Sea by two well-marked forms which it has been deemed advisable to separate under
different names. One of these is the common Henricia of the Aleutian Islands, the
shores of Bering Sea and Arctic Ocean to the east, and of the Kamchatkan-Okhotsk
region. It is a short-rayed thick-set form with conspicuous marginals and has been
called Henricia sanguinolenta eschrichtii (Miller and Troschel). It differs markedly
from typical sanguinolenta of Norway. The other form is perhaps more variable
than the first, and is found usually in deeper water. It has longer, slenderer rays,
more numerous adambulacral spinelets, and an inconstant arrangement of the ven-
tral plates in regular series; sometimes there is no sign of regular marginals; some-
times only a series of actinal intermediates is present, all the plates above being
irregularly arranged. In an aberrant variety, possibly the result of hybridization,
the marginals are large and conspicuous (station 4784).

Diagnosis.—Rays rather slender and long (R=5 r to 7 r) and size often fairly
large; extremely variable, due possibly to crossing with nearly related species (as

272 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

for example H. aspera, H. le viuscula, and H. leviuscula multispina). Abactinal
pseudopaxille small, close-set, with few papilliform spinelets to a group, these
groups arranged so as to form a close-meshed network; papular areas thus small
with one to three or four papule; marginal plates typically indistinct and small, or
not distinguishable from the other plates; when distinguishable they are small and
two series separated for part of length of ray by a series of small intermarginals;
adamulacrals with one spinelet in furrow, and on the surface upwards to fifteen
or, in very large examples, even more, spinelets; of these, two or three large ones stand
on the furrow margin, forming an angular group and behind them the others very
abruptly decrease in length, those on outer part of plate being in a coordinate group.
(Typical Scandinavian specimens have fewer spinelets than these.) A single series
of actinal intermediate plates, except at base of ray, in large examples, where a
second row may be present. ;

Variations.—The variations are legion and are bewildering. Some of the most
baffling varieties (so far as classification is concerned) I believe are due to hybridi-
zation of what is here called sanguinolenta with aspera and multispina and perhaps
also with leviuscula. One who has not been over the numerous specimens in the
collection can form no idea of the number of small and large deviations from the
type. 1t has not seemed feasible to name these variations because the specimens
are not constant enough to warrant it, and because it is more than probable that
some specimens are hybrids. A few of the principal varieties found in the collection
are listed below.

Variety A.—Large size, nearly typical as compared with Scandinavian examples;
abactinal groups of spinelets small, with few spinelets to a group; marginal plates
not distinguishable except at tip of ray; adambulacral spinelets ten, those on furrow
margin heavy and slightly compressed at tip (stations 3561, 4803, and others). A
slight variation from this form has the spinelets with rather thick skin and the
marginal plates forming very inconspicuous and not very regular series among the
other plates. They are noticeable by their larger size and more numerous spinelets
(station 4281 and others). (Pl. 65, fig. 1.)

Variety B.—Abactinal plates larger (more like small plated leviuscula) with
twelve to fifteen spinelets; marginal plates easily distinguishable, a long series of
intermarginals being present (3214). A freakish looking variety possibly due to
hybridization with multispina; perhaps only an extreme of the next.

Variety C—Often very large; abactinal plates variable, either quite small and
close set with about twelve to fifteen slender spinelets, and two or three papule, or
larger with about twenty shorter spinelets. Yet the plates appear small on account
of the large size of the specimens. The marginal plates are wide and conspicuous,
in regular series, a series of small intermarginals extending far along the ray; a
regular series of actinal intermediate plates for about half length of ray, a second
for about one-sixth, and sometimes a third on disk; marginal plates with numerous
spinelets, and the plates are more or less strongly compressed, so as to be much
wider than long, the spine-bearing portion forming a prominent keel; adambulacral-
plates with upwards to twenty-five or thirty spinelets. A ventral view reminds one
strongly of a very large leviuscula with compressed marginals, while an abactinal view

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 278

suggests sometimes sanguinolenta, and sometimes leviuscula with unusually small
plates. (Pl. 65, fig. 2; Pl. 66, figs. 1, 3-5.)

The specimens indicated above have given no end of trouble; in fact, their
classification here is tentative. The examples are for the most part large, and
two are giants compared with our ideas of Henricia. One of these has R=235 mm.,
r=32 mm.; breadth of ray at base 34 mm.; diameter of madreporie body, 6 mm.
The other is about the same size. Unless I err, these are much larger than any
heretofore recorded Henricia (stations 4778, 4784). Two other specimens from
4784 (R=122 mm. and 116 mm.) are similar to the above in general, but one has
the abactinal plates larger, almost large enough to constitute an aberrant leviuscula.
A suspicious circumstance connected with these aberrant and variable specimens is
that with few exceptions they are found along with two or more other species.
Of course there is no way to prove that they are hybrids, but the extreme instability
of their characters (or infidelity to type) and the intermediate nature of the same
strongly suggests such a view. They “act” like hybrids, to use a colloquialism;
and, as if the organism were in some way upset, we have giantism. Whatever may
be the true explanation of these remarkable variations, it does not seem advisable
at present to give them a name.

The characters of this variety are such that they might be produced in different
localities by different sets of factors. The specimens from 3602 and 3500 look like
simple intermediates between sanguinolenta and leviuscula, the leviuscula appearance
dominating on account of the prominent regular ventral rows of plates. Means of
discriminating species of Henricia are so crude that it is not possible to determine
whether all the specimens relegated to this variety really belong there. Some may
be simply extreme variations of leviuscula and multispina.

One is almost forced to conclude that nearly all the so-called species of this
genus are simply centers of variability which intergrade (both geographically
and bathymetrically) with their associated and neighboring forms.

The following stations at which variety C was taken also yielded other species:
4777, H. aspera, H. leviuscula multispina; 4778, multispina; 4779, the same; 4784,
typical sanguinolenta, aspera, multispina, leviuscula; 3223, aspera, multispina,
typical sanguinolenta; 3602, aspera, multispina, var.

Variety D.—Six specimens, alike in general features but differing in a few minor
details, remain more or less enigmatical. They evidently belong close to sanguino-
lenta, but the abactinal skeleton is much more open than is usual in this species.
The spinelets are very small and in spaced clusters of three to six along the top of
the ridges which inclose in their meshes incomplete secondary ridges, and about five
to eight papule. The latter are more or less segregated by the secondary ridges.
Actinally the specimens resemble more nearly typical sanguinolenta in the adambu-
lacral armature and general appearance of the ventral plates. Two series of mar-
ginal plates, an intermarginal series and an actinal intermediate series (all small),
are sometimes distinguishable, although the last may be all that is easily seen. It
is of course possible that this is a closely related species, but in view of three fairly
intermediate examples it has seemed safer to follow the present course. Stations
3331, 3548, 4784, 4792.

57444°—Bull. 7¢—11——18

274 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Variety E—This variety intergrades with eschrichtii (4779), with variety D,
and variety C. It might be considered as a variety of eschrichtii, owing to the
usually thick rays and small marginal plates. Two specimens from station 3224 are
about half way between the prevalent form of variety E and variety D, but are a
little nearer the latter abactinally and nearer variety E actinally.

This variety has thick arms and the abactinal skeleton forms a very definite
mesh work along which the rather short stubby spinelets are arranged in groups.
These groups are slightly spaced and have five to fifteen spinelets in each. The
papular areas formed by the meshes are subdivided irregularly by plates which are
often more or less isolated and in the minor intervals the papule protrude (see pl. 68,
fig. 3). Marginal plates are relatively much smaller than in eschrichtii, with fewer
spinelets. Proximally there are sometimes three rows of actinal intermediate plates,
soon becoming one which ends in distal third of ray; proximally sometimes as many
as three to five intermarginal series of small, irregularly arranged plates, two series
of which persist to the middle of ray, and one sometimes extends to tip, but ordi-
narily not so far; adambulacral spinelets, five to ten, in one zigzag or two irregular
transverse series, similar to typical eschrichtit.

Like many other varieties this form at first sight looks like a distinct species, but
it intergrades with eschrichtii, almost typical specimens of which occur at station
4779. The open character of the skeleton suggests aspera, as does also the spinula-
tion of the actinal surface. Again hybridism is probable, for some of the specimens
strikingly resemble a mosaic of sanguinolenta or eschrichtw and aspera.

This variety was taken at station 3213 (yielding also eschrichtii); 3224 (possi-
bly nearer variety D; multispina from this station); 4777 (from this station also
sanguinolenta variety C, multispina, and aspera); 4779 (from here an apparent inter-
grade with eschrichtii also, and sanguinolenta variety C, aspera, multispina, and an
aberrant leviuscula); 4792 (aspera and sanguinolenta from here).

Type-locality.—Denmark or Norway.

Distribution.—Cape Hatteras to Labrador; Greenland (north to lat. 70° 30’ N.);
Iceland and eastward; vicinity of Jan Mayen; Spitzbergen, north to lat. 81° 20’N.;
south along the Scandinavian coast; North Sea, the coasts of Great Britain and
Treland and Faroe Islands; and south to the Bay of Biscay and Azores, to lat. 38°
34’N. From Finmark eastward and northward to the Murman coast, the White Sea,
Barents Sea, Kara Sea, to East Cape; Bering Sea, south to vicinity of Kuril Islands
(and northern Japan?), and on the North American coast to Washington. The
distribution, therefore, is cireumpolar, and south on the continental shores to about
lat. 35° to 45° N. The greatest depth at which the species has been taken in
the North Pacific is 229 fathoms, but Sladen records a depth as great as 1,350
fathoms. These very deep records must be viewed with some doubt.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 275

Specimens examined.—Seventy-three, from the following stations:

Station. Locality. | Depth. Nature of bottom. Number. Collection.
Fathoms. | | -
PRADA INeariUnalaahkn = 2-2cs~s2aqscuck ewes BOs i Deb blet: csi: ca sutiet o- ax hee sawetae 1 | U.S. Nat, Mus.
Penge Near Afognak Island, Alaska ........ 68 _| shells, black specks............... ee,
MO sata Leis as a Ce cee amine ss gerne syn ome 51 broken shells, gray sand........... 1 Do.
$214: ..... Near Unimak, Alaska ............... 38 gray sand, gravel.................. 1 Do.
hyo INGarUDAMSK RC cnitot ccm acne esnasece: 56 black pebblesoia 72's sec stgeseas 1 Do.
3480...... Amukta Pass, Aleutians.............. 283 black sand, coral, rocks........... 1 Do.
Bass nese =e |eoses COs ncn eas Seer eres Satins see 85 Diack sands aiccc soos ei vc ones 1 | Do.
2 | Do.
1 Do.
1 Do.
1 Albatross, 1903.
1 Do.
1 Do.
3 Albatross, 1906.
1 Do.
Islands.
4792......| Near Bering Island, Commander 72 neh lei 0 5h. es tas eas 3 Do.
Islands. |
4803...... Off Simushir, Kuril Islands. ......... | 229 black sand, coarse pebbles......... 1 Do.
BBA eoweclon. oe! Cllatencitae Sale Cac noaeeeoee eos | eee) beaten OGug ac eects eeccwscn coon eer en 3 Do.
Sanborn: Harbor, Nagal; /Shumagins: «|o22 < 2. <2]: | zac ce coco sce Sec cee cde ences sees 1 U.S. Nat. Mus.
| i
VARIETY C.
(Intermediate specimens; some nearer to H. leviuscula.)
;
OAD cant Near Unimak Island, Aleutians....... | AA OTRO). en ene ene eee 1 U.S. Nat. Mus.
Neel iereramectine ma as ata arcana ae oe | 41 IAG BANG wae ona aancnas nes ae ene 1 Do.
3223...... Near Unalaska Island...........-.... 56 black pebblesi..22s20-2aceceeaees 5 Do.
G222 oo =n South of Unimak Island. 50 Bile sean ei 205s she Sas 1 Do.
3258...... Near Unalaska Island. ............... 7 black sand, gravel................. 1 Do.
35000... . South of Pribilof Islands.............. 121 fine gray sand, gravel............. 6 Do.
3602 0..... Southwest of Pribilof Islands. ........ 81 Rrcer wis sncen ses crsck sce ace ade 2 Do.
RTI2 eee Bowers Bank, Bering Sea............ 344 green mud, brown sand........... 3 Albatross, 1906,
RTT ea Near Semisopochnol Island, Aleutians| 52 Arie BT OWE. aes nese nace ka Bal. 3
Miao eben pecan GOs eee cece nace eee 43 fine black gravel.................. 1 Do.
4784¢..... Near Attu Island....................- 135 coarse pebbles............--...---- erg Do.

VARIETY D.
(Found at 4784 (1) and 4792 (1). See list of stations above, and also at the following, where typical sanguinolenta was not taken.)

Bashers... Near iUnalaskn oes ne. csctt ee anaes 350 SU en sows ecw anhs seeses~cacwesva 1 U Nat. Mus.

1 oR
GUsa cena North of Unalaska....... 0-5... 42 es 2a 91 IAG SRI kc Shine ae nanan | Do.

I

w

$213...... Near Unimak Island, Aleutians...... 1 U.S. Nat. Mus,
WE exten Near Semfsopochnol Island, Aleutians 1 Albatross, 1905.
Miroeeercaisesoc Cree aesidita a decane setae sae say ae 41 | Do,
RNG deeb slew ROD eae Rete ae den Ste caeew eae vs x a | Do.
4702....5 Near Bering Island.............-..++- 1 | Do.

« Intermediate in general appearance between sanguinolenta and leviuscula.

» Probably extremes of leviuscula, or hybrids.

¢ Some specimens nearer sanguinolenta abactinally, some nearer leviuscula.

d Near eschrichtii.

€ Stomach full of larval gastropod shells. Two of these specimens close to exchrichtil,

276 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Remarks.—The identification of this species, in view of the latitude of variation
of which it is capable, is a matter of difficulty and some uncertainty. I am unable
to decide just what to make of the several very aberrant forms listed as varieties.
If they are really the results of hybridism of course there is nothing gained by
naming them. The species seems to be in a very unstable condition, and the extreme
variants might, in the opinion of some naturalists, constitute mutations (which is per-
haps only another way of saying that hybridization has taken place more or less
recently). Whatever the cause of these variations, the fact that they exist is pain-
fully evident to a worker with ample material.

HENRICIA SANGUINOLENTA ESCHRICHTII (Miiller and Troschel.)
Pl. 67, figs. 1-3; pl. 68, figs. 1-2.

Echinaster eschrichtii MijLLbeER and TroscHEL, System der Asteriden, 1842, p. 25 (Greenland).—
Branpt, in Mrippenporrr’s Reise in den iiussersten Norden and Osten Sibiriens, vol. 2,
Theil 1, 1851, pp. 32, 34 (Okhotsk Sea; var. microdiscus, p. 33; var. macrodiscus, p. 34).

Cribella eschrichtii DusJARDIN and Hur#, Hist. nat. zoophytes Echinodermes, 1862, p. 349.

Cribrella sanguinolenta Murpocu, Ray’s Report International Polar Exp. to Point Barrow, 1885,
p. 159 (Point Franklin).

Cribrella oculata Lupwia, Zool. Jahrb., Abth. f. Syst., vol. 1, 1886, p. 289 (Bering Sea; Metschig-
men Bay).

Henricia tumida Verriti, Amer. Nat., vol. 43, Sept., 1909, fig. 5 (p. 555, name only).

Diagnosis.—Rays short and thick, rather tumid, disk large, often greatly arched
when animal is brooding eggs. In general appearance this species resembles a
short-, thick-rayed sanguinolenta abactinally, and actinally a short-, thick-rayed
leviuscula with small marginal and actinal intermediate plates and few adambulacral
spinelets. Abactinal and lateral pseudopaxille small close-set with relatively few
spinelets, and few papule to an area as in sanguinolenta; marginal plates in two
fairly regular series with few spinelets, the inferomarginal plates the widest; proxi-
mally a series of smaller intermarginal plates; a series of small pseudopaxille
adjacent to adambulacrals; adambulacrals with one furrow spinelet, and on actinal
surface five to seven short stubby clavate spinelets in a single zigzag series or
occasionally in two, decreasing in size as they recede from furrow.

Description.—Rays five, rarely six. The proportions are rather variable, espe-
cially the width of the ray at base. An average specimen from Bering Island: R=
32 mm., r=12 mm., R=2.5r; breadth of ray at base 15 mm.; an unusually slender
armed example: R=29 mm., r=8 mm., R=38.6r; breadth of ray at base 9 mm.;
a remarkably thick-set specimen from Attu is nearly pentagonal; R=14 mm.,
r=10 mm., R=1.4 mm.; breadth of ray at base, 11-12.5 mm.; height of disk, 11
mm. (Pl. 68, fig. 2.) The disk is nearly always marked by a sulcus or crease in
each interradius, and the rays are variously inflated or the disk is greatly arched.
These purely ‘“‘mechanical” differences often lend a deceptive appearance to many
specimens. The abactinal surface may be very hard and firm or less often soft;
extremity of ray blunt.

The abactinal skeleton is a close mesh work as in sanguinolenta, beset with very
slightly spaced clusters of short stubby spinelets which average a trifle shorter than
in Scandanavian specimens of sanguinolenta. Some examples, especially from
deeper water, have the spinelets longer and more delicate than the shore forms,

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 277

and more pointed at the tip; length of spinelets variable. ‘The number of spinelets
to each cluster is variable, but few (five to fifteen). Some examples have the general
average nearer the larger number, some nearer the smaller. Papule one to five to an
area, typically one to three. The lateral pseudopaxille are generally a trifle smaller
than the abactinal.

The marginal plates are clearly defined, forming with the actinal intermediate
series three parallel series, the upper of which diverges at base of ray to make room
for a short intermarginal series. The actinal intermediate plates do not reach the
end of ray. They are small, as arule about half as wide as the inferomarginals which
in turn are wider than the superomarginals. The latter series is often more or less
irregular and not well defined. As compared with leviuscula these three series have
much fewer spinelets and are less conspicuous, being more compressed. The infero-
marginals, for instance, are shorter in proportion to width than in leviuseula, with
about a dozen (frequently fewer) spinelets in two uneven transverse series. There
is a’series of single intermarginal papule and one inframarginal (between the infero-
marginals and actinal intermediate plates).

Adambulacral armature sparse, consisting of five to eight stubby spinelets, of
which the inner is the largest, arranged in a single zigzag series or more rarely in two
on the outer part of the small plate. The furrow spinelet is short, papilliform, and
situated deep in furrow. Mouth plates small, with few spinelets.

Madreporic body variable in size, circular, with coarse irregular radiating
ridges bearing spinelets.

Color in life probably variable; a specimen from Attu Island, 1906, is labeled
“rich creamy white” (Prof. J. O. Snyder); others show a reddish tinge in alcohol,
which, however, may be entirely due to the effect of the preservative. Brandt says
specimens vary from ‘‘ Himmelblau bis zum Cyanblau” upper side, and ‘‘Zinnober-
farben zum Fleischfarben”’ beneath.

Variation.—There is variability in the thickness of the arms and their relative
length as already noted. The abactinal pseudopaxille or clusters of spinelets are
also subject to considerable variation. Specimens from Bering Island have the
marginals less regular than typical examples, but the latter also occur among the
same lot, so the difference is probably purely individual. Examples from the nearer
Aleutians occasionally approach leviuscula in having larger than typical abactinal
pseudopaxille, but I have never found any specimens that seemed to prove inter-
gradation. Perfectly typical eschrichtii occurs at Kadiak. Examples were taken
from a jar containing leviuscula possibly showing that the two forms range together.

Care of the young.—This form broods the eggs and takes care of the young for
some time after they have metamorphosed. The brood pouch is the region about
the mouth, formed by arching the disk and bringing the bases of rays together.
Into this cavity the eggs are tightly packed, and are of course protected by the
body of the mother. Specimens collected at Agattu, June 8, and at Attu, June 10
and 11, 1906, are with eggs, as also a specimen from Nikolski, Bering Island, June 14.
An example collected by the Corwin has the brood pouch filled with young about
4mm. in diameter and with eight pairs of tube feet to a ray. This specimen is
figured. (PI. 68, fig. 1.)

278 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

A small variety of leviuscula occurring among the rocks at low tide in the vicinity
of Monterey, California, and which may be related to eschrichti, also broods its eggs
in December and January. (See under H. leviuscula.)

Ty pe-locality —Greenland.

Distribution.—Greenland, and from the Arctic Ocean north of Bering Strait
south to Kadiak; westward along the Alaskan Peninsula and Aleutian Islands to
Bering Island, Kamchatka, to Sea of Okhotsk (Brandt), and south to Simushir,
Kuril Islands. Low tide to 53 fathoms.

Specimens examined.—With one exception all the specimens (253) listed below
are from the National Museum collection and that of the U.S. Bureau of Fisheries
steamer Albatross, 1906. Many of the specimens were collected by Dr. W. H. Dall,
between 1871 and 1880, and by others whose names are given in the appended list.

Specimens of Henricia sanguinolenta eschrichtti examined.

| |
Station. | Locality. | Depth. Nature of bottom. Nun Collection.
| |
| Fathoms.
$2135... -3 | South of Alaskan Peninsula (Sannak 41 blackicand noe e eee eee sen 5 U.S. Nat. Mus.
Islands).
3244...... Near Hagemneister Island, Alaska. .......- 4 fine gray sand............--- 2 Do.
S251 eee | Bristol Bay, Alaska.............-..--..-- B75 da |e GOR ata ke eo 1 Do.
3257... ..| Bering Sea, north of Unimak Island....... 81 gray sand, gravel.......-.- 1 Do
S216 shoes | North of Alaskan Peninsula (vicinity of 18 gravel, sand, rocks.........- 2 Do.
| Point Moller).
47 ll Do.
41 10 Do.
36 ll Do.
53 fine sand, green mud.....--. 12 Do.
3285.....- | North of Alaskan Peninsula (vicinity of 35 gray sand, black specks. .-.. 3 Do.
Cape Kutuzof).
DOU saseeea| Sees CO teen areren ae seen See eee are 33 placksand 2. sae anes 3 Do.
B1975-~ 3. South of Pribilof Islands..........-.....-- 86 gray sand'2s..2ss20s< 22 eeoee 1 Do.
|(CapeTiaburne;"Allaskar® - (2-5 - too sects |peemioeniess | seen seem eae eee a teens 2 U.S. Nat. Mus., Henry
D. Woolfe.
Bering Straitics.22 as ccacs occ eee aosutseacs [Deans sence | pear ee ceh secs dee eee aee eee eeete 1 U_ S. Nat. Mus., Dr.
| Robt. White.
Mose WO Es Sat oo besos eee eee Soke se estat eae ce ees | eee nee Sees eer eee eee 2 U.S. Nat. Mus., Lieut.
Geo. M. Stoney.
Port Clarence, Alaska.--- <5... .-~-e- ces Bs S| re ae ee petal SaSboianee nae [ima Stanford University.
Cape Etolin, Nunivak Island, Alaska... - 8 Sand eeeee== Weise Sep ee 2 U.S. Nat. Mus., W. H.
| Dall.
Wagemeister Island, Alaska .............-|-.-------. {eco i SS SUS eee 2 Do.
Hagemeister Strait, Alaska.......... --| 8-15 gravel....... ene eee 9 Do.
| St. Paul Island, Pribilofs, Alaska... et 6-9 GaN... 22. Rew eee ee cs 8 Do.
Kadiak: (Alaska 02. so. scdecc cuneseee eee Shhoreiy || 4 9825 eh. e as eee Fee 7 U.S. Nat. Mus., Alba-
tross.
Kadiak (Chajafka Cove), Alaska ......... Shore:ss|=22- esse See eee 2 U.S. Nat. Mus., W. H.
Dall.
Tliuliuk Harbor, Unalaska .........-...-- | 10 shingle Yo). h essen 2 Do.
| Popoff Strait, Shumagin Islands, Alaska. | LOWeStil5 <3. 22s acewc cncenssccseee snared 4 Do.
water. |
, Sanborn Harbor, Nagai, Shumagins.....- Lowest Me ceca set tee ee 25 Do.
| water.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 279

Specimens of Henrica sanguinolenta eschrichtii examined—Continued.

| |
Station. Locality. | Depth. Nature of bottom. a Collection,
ae ses ir 4 |—_——_—|
Fathoms.
Sanborn Harbor, Nagai, Shumagins.....|..........)......2.0..0eeee0e0- Naaneae ees 6 U. 8. Nat. Mus., Capt.
| | G. W. Hall.
Round Island, Coal Harbor, Unga, Shu- |eceeeeeeee leads avec eetews ean eeeee ek 4 U.S. Nat. Mus., W. H.
magins. | | Dall.
nialankes-s- s- ts Saee ioe chats faced feeeea es do alivaiel an dauacaceege ets ket 6 Do.
Captains Bay, Unalaska ...............2.|.-.....-.- jxqsensdaneatepwe «amp atweapaare 3 Do.
Captains Harbor, Unalaska 9 | 1 Do.
PATTAYA OLIMAR ERS ia waisla ne onda a cnim ms = = [an ain nine ne ee ene 1 Do.
Port Levashef, Unalaska................. | 20-30 |... el ba Do.
Nazan Bay, Atka Island, Aleutians....... [somal Eafid ede i Se-rde eek Ab. al 1 Do.
Adak Island, Aleutians ................-- epee knee See ere 15 | U.8. Nat. Mus., Alba-
| | tross, 1893.
‘Constantine Harbor, “Amchitkm Island, |... 2.2.0.3. -<-ascccncccccccueveveveccas ll U.S. Nat. Mus., W. H.
Aleutians. Dall.
Kyska Harbor, Kyska Island, Aleutians. .| Low wa- |.............-...--..---2--00- 4 Do.
ter.
Sewes WO Geenewcates sweet scset eeeeecenset| SOMIS: | MENOY NU scscercesccccescl PO) "f Do.
Agattu Island, Aleutians -| Shore. | rocks... 6 | Albatross, 1906.
Attu Island, Aleutians ...............--- Shorey, ||, ate-2=e0 9 | Do
Nikolski, Bering Island, Commander Is- | Shore. |............-......----------- 19 | Do.
lands.
(Berne Islantcecuessns2cnssece cece cacee aes Bhhore. © |SSa. coos Se aes beta } 2 U. 8. Nat. Mus., L.
| Stejneger.
res 0 seep ep eee antes enn en tetae eee orate One toes e see ee ae Pn U. 8. Nat. Mus., N.
| Grebnitsky.
Medni Island, Commander Islands. .....- . WrBhoresn| |S 2856522 ee sc dekak acoee a 8 U.S. Nat. Mus., Alba-
| tross, 1906.
Simushir, Kuril Islands (Milne Bay)...... HOM occa cee ee at mec esaaeacer 6 Do.
Humboldt Bay, Shumagins, Alaska ..... | Shore. | “sand bar’”................. ey U.S. Nat. Mus., Alba-
tross.
BAGS NOL ORO: Seer ene setae ts aes ae ea eretocs aT U. 8S. Nat. Mus., Cor-
win, Capt. M. A.
} Healy, 1884.
MOBY ona Near Petropavlovsk, Kamchatka ........ | 648 | green sand, pebbles......... 1 Albatross, 1906.

Remarks.—Miiller and Troschel’s description of Echinaster eschrichtii agrees in
the main with this common shore form of Bering Sea. Although there are no
records between Point Franklin and Greenland (which is not surprising), I believe
I do not err in applying the name eschrichtii to the species under consideration. It
has seemed desirable to keep this short-rayed, thick-set form with conspicuous
marginal plates distinct from typical sanguinolenta. Whether it is a distinct
species with a distribution nearly coéxtensive with sanguinolenta or only a variety
of the latter, must for the present remain undecided. I have, however, found
several intergrades, and for that reason and the obviously close kinship of eschrichtii
to sanguinolenta have used a trinomial designation.

280 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
HENRICIA LEVIUSCULA (Stimpson).
Pl. 69, figs. 1, 2; pl. 70, figs. 1, 2; pl. 71, figs. 2, 3; pl. 111, fig. 6.
Linckia leviuscula Stimpson, Boston Journ. Nat. Hist., vol. 6 (No. 4), 1857, p. 529 (Puget Sound,
Washington).

Chetaster californicus GruBrE, 42 Jahresber. d. schles. Ges. f. vaterliand. Cultur, 1865, p. 52
(possibly refers to leviuscula).

Cribrella leviuscula Wuirraves, Canadian Nat., vol. 8, Dec. 20, 1878, p. 466; Trans. Roy. Soc.
Canada, vol. 4, sec. 4, 1886 (1887), p. 117.—SuapeEn, Challenger Asteroidea, 1889.—pE Lorto1,
Mem. soc. phys. et d’ hist. nat., Genéve, vol. 32, pt. 2, No. 9, 1897, p. 14, pl. 2 (17), fig. 2—
CxiarKk, Proc. Boston Soc. Nat. Hist., vol. 29, No. 15, 1901, p. 327.

Cribrella leviuscula crassa CLARK, Proc. Boston Soc. Nat. Hist., vol. 29, 1901, p. 327.

Cribrella leviuscula attenuata CLARK, Proc. Boston Soc. Nat. Hist, vol. 29, 1901, p. 328.

Henricia leviuscula Fisner, Zool. Anz., vol. 35, March 29, 1910, p. 570.

Diagnosis.—Rays five (rarely six). R=3 to 7 r; R=upward to 100 mm.
General form variable, rays slender, more or less elongated or short and thickened,
cylindrical, tapering, blunt; abactinal pseudopaxille or groups of spinelets close-
set, typically larger than intervening papular spaces; spinelets very short, numerous,
and close-set, having as a rule a granuliform appearance under low magnification;
adjacent to adambulacral plates three very regular series of juxtaposed plates, the
median largest (inferomarginal), but the relative sizes variable; near base of ray
the uppermost (superomarginal) series bends upward in the interradial angle, and
the triangular area thus formed between the infero- and superomarginals is filled
with a variable number of smaller plates. Adambulacral plates wider than long,
typically with eight to twelve stubby spinelets in two transverse series, all of these
larger than adjacent actinal spinelets and increasing in size toward the furrow,
where two are considerably bigger than the rest.

Description.—As the variations in this species are many, a composite description
would be unintelligible. For description, what is believed to be the typical form
from Puget Sound will be taken. Then some of the principal variations of this type
will be noted. R=89 mm.; r=13 mm.; R=6.8 r; R=49 mm.; r=11 mm.;
R=4.5 r. Searcely any two have exactly the same proportions. Breadth of ray
at base is slightly greater than r. The abactinal pseudopaxille or groups of spine-
lets are largest on the median radial area of disk and diminish gradually on either
side toward the superomarginal plates which are usually abruptly larger. The
plates are larger than the intervening papular areas which contain two to five
papule on the dorsal surface, one or two on the side, and one on the actinal surface.
The plates are covered with minute short close-set spinelets, typically 30 to 60 on
the larger plates. There are two quite distinct variations even in otherwise nearly
similar examples. (1) The spinelets end in a clear, blunt, or more or less pointed,
glassy tip, below which are minute denticulations, more or less hidden by the tip,
variety A. (2) The spinelets are slenderer, and, as a consequence, less crowded,
ending in numerous minute denticulations, there being no sign of-the special glassy
tip, variety B. In the first variety the dorsal spinulation has a decidedly granuliform
appearance, especially as the glassy tips are more globular in some examples.
Under a glass the spinelets appear pearly or bead like (when dried). This form of
spine may be due to greater age, or to the action of waves. On the ventral surface

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 281

the glassy tip is sharper and less conspicuous. The dorso-lateral plates are fre-
quently in fairly regular vertical rows, each corresponding to a superomarginal;
in the interspace between them are smaller secondary plates. These rows are more
regular in the individuals with glassy spinelets, and the presence of the smaller
intermediate plates may be another indication of advanced age.

There are three regular rows of conspicuously large, regular plates adjacent to
the adambulacrals. The middle row (the inferomarginal) has the largest plates.
Usually the upper or superomarginal series has slightly smaller plates, while those
of the actinal series, adjacent to the adambulacrals, are smallest. A regular row
of mostly single papulze runs between these series, although that between the actinal
and adambulacral series does not extend far along the ray except in old examples.
The intra-marginal plates at base of ray do not extend far along ray (about two-
thirds of r, measured from interbrachial line).

Adambulacral spinelets variable in number but usually eight to twelve, in the
form with glassy spinelets, and twelve to eighteen in the other variety. In the
former there are typically two transverse rows, with one or two enlarged furrow
spinelets (two in an oblique series, or one median and two in a longitudinal row
just back of it). In the variety with slender, multidenticulate spinelets there are
usually three transverse rows of spinelets to each adambulacral plate, and two or
three furrow spinelets. Intermediate specimens, in respect to the adambulacral
armature, are several of the second variety with the armature characteristic of the
first. As usual in Henricia, there is a curved spinelet on the furrow face of each
adambulacral plate. Madreporic body covered with spinelets and with coarse
irregular radiating striz.

Color in life, bright brick red, orange on actinal surface; specimens collected
by Dr. J. C. Brown at Port Renfrew, British Columbia, were as above; some small
thick-armed examples had slate-gray to lilac or drab disks with the same colors
running out upon arms. In other specimens there was a yellowish brown color
instead of red as a background, with here and there invasions of brick red on the
disk and arms. Specimens from Monterey Bay vary greatly in color, for which
see below, ‘‘Specimens from Monterey Bay.”

Variations.—The littoral form from Puget Sound with very regular marginal
plates, eight to fifteen adambulacral spinelets, and the abactinal spinelets ending
in numerous sharp denticulations rather than in a solid glassy tip, is assumed to be
the typical one. This conclusion is based on an examination of a specimen evi-
dently labeled by Stimpson. The type is therefore one of the various forms of
variety B, among those listed herewith.

Variety A.—This form varies in the length and breadth of the rays. The
thick-armed examples are usually small, while the medium and slender-armed ones
are much in the majority. I take Doctor Clark’s Henricia leviuscula crassa and
H. 1. attenuata to be variations in proportions of this and the following form, as no
differences other than the shape of rays and size of disk are given. Monterey Bay
to Kadiak Island. There is less variability in the spinulation of this than of any
other variety. (Pl. 69, fig. 1; Pl. 70, figs. 1, 2a.)

Variety B.—This is the second variety noted in the foregoing description.
The principal differences are in the slenderer and multidenticulate spinelets, more

282 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

numerous adambulacral spinelets, the absence of any great regularity of dorso-
lateral plates above the superomarginals (and also the intervening vertical series
of spaced smaller plates). Specimens of this variety also vary in respect to propor-
tions from R=3rto7r. (Pl. 70, figs. 2b, 2c.)

Specimens occur which are intermediate between variety A and variety B,
or which are modifications of A with a slender glassy tip to the granuliform spinelets
(station 4205). Such are scarcely separable from variety D.

Variety C—A fairly typical form, as regards arrangement of plates, has the
spinelets stout as in A, but more or less clavate and without a sign of the glassy
tip. Adambulacral spinelets 15 to 25, and frequently three on furrow margin
(station 4209). From this haul also variety B and a specimen near A.

Variety D.—Some large specimens from offshore are similar to A but have
fewer spinelets to the abactinal plates and the glassy tip to the granuliform spinelets
is very slender, sharp, and inconspicuous, being sometimes worn off. In these
specimens the papular areas sometimes equal or slightly exceed the plates in size
and there may be some secondary smaller abactinal plates scattered among the
larger. This variety is only a trifle different from the typical form. ‘The marginal
plates are regular and abruptly larger than the lateral plates above them (stations
2874, 2886).

Variety E—A ‘‘freak” variety from Monterey Bay (shore) has small abactinal
plates with short spinelets of B type, seven or eight adambulacral spinelets which
are heavy and spatulate near the furrow, very irregular marginals, the supero-
marginal series being difficult to follow at middle of ray. (No. 387, Stanford coll.,
two specimens; station 4431, one specimen.)

Specimens from vicinity of Monterey, California (Variety F).—In addition to
variety A, which occurs offshore at Pacific Grove, Point Pinos, Monterey, and
near Santa Cruz on rocky bottom, the variety B is even more common, and a
dwarf form which far outnumbers both of the above. This small variety at first
sight, owing to its different coloration, appears to be very distinct from leviuscula.
I am unable to find constant differences between it and what appear to be really
young, or immature, leviuscula from Puget Sound and Kadiak. The size of a
starfish is not a sure index to its age, which probably partly accounts for the varia-
bility of these specimens. The marginal plates, for instance, vary from the pro-
portions of those of eschrichtii to those of leviuscula.

The abactinal plates are variable in size but of the type of leviuscula, the plates
being usually larger than the papular areas. Spinelets vary just as in the large
specimens, except that the beady or glassy-tipped spinelets are not present. Some
specimens have the spinelets conspicuously slenderer, approaching dyscrita in form.
The adambulacral plates are short, with five to eight spinelets in two transverse
rows, and a single more or less compressed furrow spinelet much larger than the
rest. There are three longitudinal series of regular plates external to the adambu-
lacrals, of which the two outer (marginals) are the largest. The marginal plates
are more compressed, less squarish than in large leviuscula, and have two to
four irregular transverse rows of spinelets pointed outward.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 283

These small Henricia are sexually mature at R=12 mm. A good-sized indi-
vidual measures about twice that. R equals about 3 r, but the thickness of the
ray is variable. They brood their eggs in January, forming for that purpose a
concavity about the mouth by arching the disk. The eggs are orange yellow and
are hidden by the mother, which in spite of its bright colors, harmonizes fairly well
with the coralline on the rocks. However, when brooding the eggs this species is
usually found under rocks or hidden between them, in darkness.

So far as I know, no large leviuscula has been found brooding eggs, a habit of
both H. sanguinolenta and H. eschrichtii. A knowledge of this would be of great
value in determining the status of the form under consideration.

I have specimens of this small variety from Marin County, San Francisco, and
Crescent City, California, and from the Straits of Fuca, and Puget Sound. A
young leviuscula collected along with many large specimens of variety A at Kadiak
is not different in any important respects from some Monterey Bay specimens of
the dwarf variety, especially as regards the compressed marginals. Dr. J. C.
Brown found the dwarf form at Port Renfrew, British Columbia, on the Straits of
Fuca, and noted the difference in color, the large examples being uniform brick red
above (as at Monterey Bay). I am unable to find any differences between the
Monterey and Straits of Fuca specimens except the slightly larger abactinal plates
of the latter, these being absolutely typical of leviuscula (as in fact are those of
some of the Monterey specimens).

I am rather against my will forced to range these small specimens under H.
leviuscula. In view of the great variability of these specimens and of the lack of
any tangible structural difference, it would be hazardous to separate them on the
strength of our ignorance of the egg-laying habits of typical leviuscula.

The color variations are so decided that the most important are classified
below.

1. Variations of brown predominating.

a. Ground color of abactinal surface drab gray mottled with light red, burnt
sienna, Van Dyke brown; mottling on disk, or disk and arms; ground color extend-
ing over marginals. Actinal surface below marginals yellowish.

b. Ground color dull light brown (tan) variously mottled with brown madder
on rays and disk, or disk only; ground color varying to light yellowish brown, and
mottling of different intensity. *

c. Dull brown, arms tipped with orange (rare).

2. Orange, yellow, or red predominating.

a. Bright orange above with irregular bands of grayish at base of rays; light
straw yellow below.

b. Dull vermilion above, yellowish white below.

c. Deep dull yellow (several shades) sparsely mottled with dark reddish brown.

d. Brick red, sparsely mottled with drab.

e. Light reddish (brick) yellow mottled sparsely with brown.

f. Light brown, dull orange, and brown madder of several intensities, marbled
together.

284 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

3. Lavender and purplish or brown.

a. Ground color lavender, more or less pinkish, mottled on disk and rays, or
on rays only with reddish or purplish brown, or more rarely with yellowish brown.
Intermediates in ground color between 1 and 3 occur.

4. Various shades of pink or pale lavender (all small specimens).

a. Very pale rose pink, nearly white below.

b. Very pale lavender and lilac.

c. Shades intermediate between these and light yellowish brown.

Type.—An authentic specimen, labeled by Stimpson, in the U. S. National
Museum is perhaps the type; if not, it is at least a topotype.

Ty pe-locality —Puget Sound, Washington.

Distribution —From Aleutian Islands, Alaska, to Monterey Bay, California,
in a typical form, and thence south to the Santa Barbara Channel, and San Diego,
California, the southern forms not being typical. The typical form is found along
shore at low tide, or in comparatively shallow water. Bathymetrical range, low
tide to 53 fathoms.

Specimens examined.—Three hundred and forty-two, from the following
stations. (The letters after the number of specimens indicate the varieties as
detailed under variations.)

Specimens of Henricia leviuscula examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
48 mud, sand, broken shells....| 8 B U.S. Nat. Mus.
40 pebbles 26-2 oo eee pee 1B - Do.
27 rocks and shells............- 6D Do.
40 Saree (0 an AE AA 26BCD Do.
50 rockysetitese- eee eee 3BD Do.
: 42 coral and pebbles....... 1 Do.
| Cortes Bank off San Diego... 5 26 broken shells..........: 1 Do.
201 Sense | Off Santa Cruz Island...............2. 48 fine gray sand.......... 1 Do.
48 coral, broken shells, rocky 2B Do.
64 apes On otro se a eecaeneoeaiene 1B Do.
53 TOCKY. cam an eee wan eneceenaee) 1B Do
46 coral and pebbles....... 12 Do
4G eeace Owersesaee eee ae 3B Do.
80 green mud, pebbles......... 2B Do
37 LOCK. 2 acon deceneenscescane 1 Do
49 rocky, gravel, sand.........- | 1B Do.
26-15 rocks shellstaec.- eee emeame ae IAB Albatross, 1903.
~..| 25-24 rocky, coarse sand, shells... -| 3B Do.
4431 ¢..... Off Santa Cruz Island, California......| 30-41 mud, gray sand.......-...-- | 1E Albatross, 1904.
4779 ¢..... Near Semisopochnoi, Aleutian Islands. 54 broken shells, pebbles, sand.| 1 Albatross, 1906.
Monterey, California...............:.- B12 oaeee coe ene see eee ase eee 2 U.S. Nat. Mus., W. H.
Dall.
Reem GO! sco sescostctsac sczcre sence ecge tee sect Coc eeeee a ee ee ee RT U. S. Nat. Mus., H.
| | femphill.
: oon Eines, «Monterey Bay, /Call- | Lowi} Sa eae ane | 132 ABEF) Stanford University.
fornia. | tide. |j |
Capitola, Monterey Bay............... { roy | ews aaa emcee oeeeceaeeas 1A | Do.
| tide. | | !

a Some not at all typical.

B More or less aberrant variety. Dorsal and marginal plates somewhat smaller than in typical leviuscula, and spinelets
a trifle more delicate. Some examples with either one or both of the furrow spinelets spatulate and truncate. The specimens
are variable.

¢ Aberrant variety.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 285

Specimens of Henricia levinscula examined—Continued.

ee

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
San Francisco Bay, California ........|.......... [oth enpaccdcacusaoa sth sanpp aus 2B U. 8. Nat. Mus., Dr.
} John Hornung.
Marin County, California.............. BONCS | hn is oat cman mela ametetae WAF University of Califor-
nia.
Bolinas, California.................... Shore. 1A Do.
Crescent City, California --| Shore. IF Stanford University.
Neah Bay, Washington............... Weimeaemeate 4B U. S. Nat. Mus., J. 6.
| Swan.
Straits of Juan de Fuca, Washington..|..........|....csceceesseeeeceeeeeeeeees 5AB | U.S. Nat. Mus, D.8.
| Jordan.
Fort Steilacoom, Puget Sound, Wash- |..........]........scceecceececcecccences 1B U.S. Nat. Mus., Dr.
ington. Geo. Suckley.
(NPG t MOUNT <a ce tonh was eaweseulon ssn ot oul ebay pdec awe sscpes dadag 4AB U.S. Nat. Mus., D. 8.
| Jordan.
eee DO eee ween an eee een ene | ene eae eee a eeie aan a coees pes adsl Oe U.S. Nat. Mus., Chas.
M. Drake.
SPACOMIA Wy OANINP UO oon cece cv anicttcenl-ae denon ce | seus tee vat etedenthsscewenee joes U.S. Nat. Mus., Mrs.
| W. B. Hare.
PU pet SOUNC ence cceasenecndan-eeenen|ececen == |-necceseeeereeeeeecereeeeegees| BA Stanford University.
Port Renfrew, Vancouver Island, Low © z
British Columbia (on Straits of | tide, fc et nee he {conan | Sere Dicey,
| John C. Brown.
Fuca). rocks.
Clayoquot Sound, British Columbia..) Shore. |...............-.-....-..2+05- | 1B Stanford University,
| C,H. Gilbert.
Massett, iQueen. Charlotte , Island) |c.occc. ccc [icon sanccewscocsccnsnucsccvens |} 2B U. S. Nat. Mus., J. G.
British Columbia. Swan.
Barclay Sound, British Columbia. ....|..........|....02.sceescssencccescenccees} SAB U. 8. Nat. Mus.
Naha Bay, Alaska..................-. Shore. 1B Albatross, 1905.
Site pAISAKie ead atecdon esos as 15 1B U. S. Nat. Mus., W.
| | H. Dall.
Paes nee er a acl ern ese ccitwruceeceane] OD” |OUR NM Mle, ren
| dinand Bischoff.
PATA KULC Sas aan akan einnas ce ues -cese Bae eee Pag tess chek vaes te tat Jdoseeee 1 ALB | U.S. Nat. Mus., Dr.
| | W. Hf. Jones,
Fort Wrangel, Alaska.................|....-..... Weaiot date Cit oe ec a Searecon essere 2B | U.S. Nat. Mus., W. EH.
| | Dall.
Karluk, Kadiak Island, Alaska....... Bire Alp eee Ss Ossett is 0A | Cloudsley Rutter, 1903
(U.8. B. F.).
(Radiak Island estes tat cae coscee et ttn ea carne Peep renee a Sent arte | 2 U. 8. Nat. Mus.
Middleton Island, Alaska............. 10-12 | gravel and sand............. ra | U.S. Nat. Mus., W. H.
| Dall.

Remarks.—This species is the form ordinarily encountered along shore from
Monterey Bay to Kadiak, the largest examples coming from the more northern
localities. Typical specimens are very easily distinguished by the numerous short
almost granuliform appearing spinelets of the large dorsal pseudopaxille and by
the three regular series of large plates adjacent to the adambulacral series. The
adambulacral spinelets are few, short, and robust.

At the south end of Monterey Bay the small variety is the only one at all
common. This is found under rocks at low tide usually in more or less sheltered
coves where there is a profuse growth of alge, Phyllospadiz, and coralline. A few
are found on the exposed surface of rocks and on alge. At Port Renfrew, Van-
couver Island, on the Straits of Juan de Fuca, Dr. J. C. Brown, who kindly collected

286 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

for me, found typical large individuals common at low tide along the rocky shores,
where they occurred in pot holes attached to rocks, or kelp, and nearly always
submerged. With them were numerous small specimens comparable to the dwarf
forms so common at Monterey Bay.

HENRICIA LEVIUSCULA MULTISPINA Fisher.
Pl. 72, figs. 1-4; pl. 73, figs. 1, 2.
Henricia leviuscula multispina FisuEer, Zool. Anz., vol. 35, March 29, 1910, p. 571.

Diagnosis.—General form very similar to that of H. leviuscula but the abactinal
plates more compactly placed, usually smaller and more numerous; the papular
areas smaller; the spinelets very numerous, delicate, longer than in leviuscula
and ending typically in three or four (sometimes more) slender, very sharp points
or awns; adambulacral plates with numerous (twenty-five to fifty) spinelets, and
typically with two spinelets on the furrow face beyond the middle of ray; proximally
only one; actinal intermediate series of plates extending one-half to three-fourths
length of ray, not whole length as in leviuscula.

Description —Rays five. R=88 mm.; r=15 mm.; R=5.8r. Breadth of ray
at base, 18mm. Rays as arule rather slender, long, gradually tapering. Abactinal
plates rather variable in size, but typically smaller and more compactly placed
than in H. leviuscula. The spine-bearing surface of plate raised and convex as in
leviuscula; the spinelets are numerous, close-set and bristling, overhanging more
or less the papular areas. They are very delicate and glassy when dried, and
terminate in three to several sharp awns or points. The spinelets form a very fine
close velvety nap all over the surface of the body. It is not possible to count the
spinelets with much accuracy, but there are forty to sixty (frequently more) on the
median abactinal plates. Dorsally there are two to five papule to an area, laterally
and actinally three to one.

Superomarginal, inferomarginal, and actinal intermediate plates form three
regular series as in leviuscula and have also the same relative proportions, but the
intermediate series extends only one-half to three-fourths the length of ray. As a
rule the superomarginals are abruptly larger than the dorsolateral plates just above.
Inferomarginals are conspicuously elongated transversely, usually about twice as
wide as the actinal intermediate plates (those adjacent to adambulacrals). The
intermarginal plates are variable; sometimes they are confined to a small triangular
area at base of ray and sometimes extend halfway along ray as a regular series of
small roundish pseudopaxille; occasionally a second row is present extending nearly
or quite as far as the first. Therefore the superomarginal series is bent upward
proximally at the interbrachial angle, while the inferomarginals (as in leviuscula)
run parallel to the actinal intermediates to the mouth angle. Between these
regular series are series of papular areas (except between adambulacrals and actinal
intermediate rows where papule are present only on disk) with commonly two or
three papule to an area in large examples. The spinelets of the marginal and
actinal plates are similar to those of the abactinal surface, and the plates are com-
pactly placed.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 287

Adambulacral plates with typically twenty-five to forty (sometimes as many
as fifty or as few as twenty according to size of specimen) multidenticulate spinelets,
those on outer part of plate similar to the actinal intermediate spinelets in size,
and increasing in length and thickness toward furrow, on the margin of which are
about three or four stout often compressed spinelets. The spinelets are in four to
six not very regular transverse series on outer part of plate, and three or four on
inner. A characteristic of this form is the presence of two furrow spinelets (one
above the other) on the furrow face of the plate, on outer third, half or two-thirds
of ray; proximally there is but a single spinelet, as in H. leviuscula. In some
intermediate or aberrant examples the diplacanthid plates are confined to the tip
of ray. No typical examples lack them, however.

Madreporic body slightly convex, circular, of very variable sculpture, but
covered with small spinelets similar to those of the adjacent plates.

Color in life-—A specimen from station 4791 was colored in life as follows:
Central aboral region pale lilac, fading toward and disappearing near middle of
arms; other parts creamy, more yellow toward tips of arms (J. O. Snyder).

Variations.—The typical examples of this species or race (whichever it is) are
so different from leviuscula that they may be told at a glance. But the deviations
from the type are both varied and numerous, so that it is sometimes impossible to
decide how a given specimen should be classified. Whatever may be the explana-
tion of the multitude of variations it is certain that the Alaskan specimens present
the extreme of difference from leviuscula, while the Californian race, dyscrita, forms
a fairly perfect connecting link. Dyscrita has much fewer than the typical number
of adambulacral spinelets (usually fifteen or less), fewer and shorter spinelets
to the abactinal plates, and two spinelets on the furrow face of the adambulacrals
only at tip of ray and sometimes not at all. Certain extremes of leviuscula from
tide pools of Monterey Bay are so close to extremes of this southern representative
of multispina that it is well-nigh impossible to separate them. One is forced to
conclude that the two forms really intergrade. It is possible that multispina is a
deeper water variety or race of leviuscula, and there is also evidence that the more
delicately spined forms are from soft bottom while the short-spined anomalies, so
difficult to classify satisfactorily, are the same species or race from hard bottom.
However, multispina is found along shore, for there are several specimens from
Kadiak Island which are nearly like the typical form. From here also are a number
of specimens which depart from the type and approach leviuscula, and a very few
which are intermediate. They were evidently collected along with typical leviuscula
and presumably occur together (see Kadiak variety below). Still another possi-
bility is that multispina is a distinct but variable species which encroaches upon
the domain of leviuscula (which occurs from low tide to moderate depths) and that
whenever this occurs hybridization takes place more or less frequently. In this
manner some of the freaks might have originated. The probability of hybridism
is very strong in the case of aspera and leviuscula. When two well-marked forms
range together it often happens that specimens are found which are more or less
mosaics of these. Such aberrant forms are bewilderingly variable. Some of the
more important variations are noted below.

288 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Alaskan specimens.—The specimens which present the extremes of difference
from leviuscula are from off the Aleutian Islands and Commander Islands. These
and other typical examples of the race differ among themselves in the size of the abac-
tinal plates (as shown by the four photographs on plate 72) and in the thickness of
the rays, as well as in the extent of the intermarginal plating. The slenderer-armed
examples usually have very few intermarginal plates on the rays (only a small
triangle in the interbrachial region) while the stouter-rayed specimens have one
or two series extending half the length of ray between infero- and superomarginals.
The presence of intermarginals is not strictly correlated with thickness of ray, for a
specimen with moderately slender rays has the intermarginals well developed
(station 4777). Specimens with smaller abactinal plates have finer spinelets.

A well-marked variety occurs at Kadiak Island (Karluk), Herendeen Bay
and Port Clarence, Alaska. This has rather thick rays (R=4 to 4.5 r) more or less
swollen at the base, and one or two series of intermarginal plates extending a varying
distance along ray. The spinelets are shorter than in typical forms, and the
adambulacral spinelets are fewer (about twenty). There are usually two spinelets
deep in the furrow, on the outer part of ray. This variety passes into specimens
which are difficult to separate from extremes of variety B of leviuscula. They
may be told usually by the presence of two furrow spinelets near the end of ray
and by the actinal intermediate plates not reaching the end of ray. I regard them
as intergrades, either by extremes of individual variation or by hybridization.
Of course occurring as these intermediates do, all at a single station, the case is
different from the intergradation of terrestrial subspecies.

Type.—Cat. No. 27779, U.S.N.M.

Type-locality — Albatross station 4779, near Semisopochnoi Island, Aleutians,
54 fathoms, broken shells, pebbles, sand.

Distribution.—From Oregon to Bering Sea, Bering Strait, the Aleutians, and
south to Simushir, Kuril Islands, on the Asiatic side; typical form in moderate
depths, a variety more or less intermediate with Jeviuscula occurring at low tide at
Kadiak, the Aleutians, and northward. Bathymetrical range to 238 fathoms.

Specimens examined.—One hundred and seven from the following localities:

Specimens of Henricia leviuscula multispina examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.

Near Unalaska /< os. 2.22 <cncssccce~a= ses 56 DEDDIES sees caeae ee eee eee 1 U.S. Nat. Mus.
Near Shumagin Islands, Alaska ........ 69 preen mud. 9. s.--c-seo-=~ 2 Do.
Near Afognak Island, Alaska........... 51 gray sand, gravel....... = 1 Do.
.| Off Vancouver Island, British Columbia. 238 gray sand and pebbles......- 1 Do.
Admiralty Inlet, Washington........... 40 pebbles. 2.- ccc oe ese cee 2 Do.
Off Cape Flattery, Washington ........ 27 rocks and shells. .......--.... 1 Do.
ee CQO! Sentcn nies as seen aiee eee eae 59 black sand and mud.......- 2 Do.
Heceta Bank, Oregon ................-- 64 coral, broken shells, rocky... 1 Do.
.| South of Alaskan Peninsula (Sannak 41 black'sanid==2c—--2o-=< esos 5 Do.

Islands).

+\eNear(Unalaskn< -o.cec sna teacsened 56 black pebbles. .....-......-. | 1 Do.
KS 12 stones 6 Do.
10 stones, pebbles.............- 14 Do.
5 sand, pebbles=<--c-----—--.= 6 Do.

a These specimens are all more or less Intermediate.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 289

Specimens of Henricia leviuscula multispina examined—Continued.

Station. Locality. Depth. Nature of bottom. Number. | Collection.

Fathoms. |

3262...... Near Unimak Island, Aleutians........ 43 black stones, rocks.......... 4 | U. 8. Nat. Mus.
= OSes | Straits of Fuca, Washington .......... 125 TOOK Ss. cdvcbevbaebckaaerce: 1 Do.

3546. ..... ean Unalagka.... se-c-ns---<-4e-- ae 36 gravel, black sand.... 1 Do.

3595...... , Admiralty Inlet, Washington........... 49 rocky, gravel, sand.......... 1 | Do.

BOM cle mas aa WD eeaedee stent cede tone td aawas 67 coarse black sand............ 2 Do.

4273...... | Alitak Bay, Kadiak Island, Alaska....| 36-41 | green mud, fine sand........ 1 | Albatross, 1903.

4291¢.....| Shelikof Strait, Alaska ................. 65-48 | blue mud, sand, gravel...... 2 | Do.

BT Tintom a= | Near Semisopochnoi Island, Aleutians. . 52 TING STAVE. cnc cen wasmacmen 1 | Albatross, 1906.

4778....-- Pte: lo eeeedeeens seb eeldods= <-eee eee eees 43 fine black gravel........-... 6 Do.

MO penta RL ae en tenia as aan nina es i broken shells, pebbles, sand. 8 Do.

4784...... Near Attu Island, Aleutians ........... 135 coarse pebbles.........-....- 6 Do.

4701.2... Near Bering Island, Commander Islands! 76 TOOK Wires <Sc\de nana tea we te 1 Do.

4803...... Off Simushir, Kuril Islands....:....... | 229 black sand, coarse pebbles... 1 Do.

FINCA Ae SIS RON ace ace se nee seal S's an «sdk aelv aeniewewunsbesehecs sa sxnennass 1 U. 8. Nat. Mus., Mrs.
| } W. B. Hare.
Whales ss -- 222 o.c0-gcucen Gan cha aot sep epeee feick amine cease eas eerranw as on U. 8. Nat. Mus., W. H.

| Dall.

Sanborn Harbor, Nagai, Shumagin | lowest |...............-..--.--------- 5 i Do.

Islands. water. i
West GOBSG INOKUR AIDONMON@o0 op otoace nue |saqacaxseelwvecaswevousensancaneucewsvur | 1 U. S. Nat. Mus., Lieut.
Commander H. E.
| Nichols, U. S. Navy.

Kyska Harbor, Aleutians ®............- Os Ned an ciaeetesee eer erots 1 U.S. Nat. Mus., W. H.

Dall.

Semidi Islands (west of Kadiak)........ 128 RIM VON seins nooe a wee aaecperes 1 } Do.

Pore Ginrenioe A lASKe 8 eee on oceanic nf ce coun sane owevaseniccsacncennnseces seuss 1 | Stanford University.

Orin PHI Cte ne nase a dcacennencn| cues andre Fos eeince ce easans a miclaaeis ve ates 1 U. S. Nat. Mus.

ROMAN ANAS Oa ono we sca seeces ou |-adunancan|scecccsucunmcdsseanaseseswenea 2 U. 8. Nat. Mus., W. J.

| Fisher.

St. Paul, Alaska >............ Sneed sea lanada tects hha cccnqamsen tenn weeee Stele «ae | 1 Do.

Rendlak, Alssia tooo con o ccc cece cece caen|ooccousans|eovmstoneesrs-<e-ceee aaa Sane 9 Do.

aA probable hybrid between aspera and multispina in this haul.
> Shore specimens are never quite typical, but show a certain resemblance to H. leviuscula, with which this form ts believed

to intergrade.
c These as well as a few others may have a slight Infusion of H. sanguinolenta eschrichtii blood; the two forms range together

in places (Port Clarence, Kadiak, Aleutians).
d Nearly typical.
HENRICIA LEVIUSCULA DYSCRITA,? new subspecies.

Pl. 74, figs. 1-5.

Diagnosis.—Nearly intermediate in general appearance between H. leviuscula
and H. leviuscula multispina, but the abactinal plates smaller with fewer spinelets
than in multispina and papular areas usually relatively larger; adambulacral plates
with about fifteen spinelets (hence fewer than in multispina) and with two spinelets
in the furrow only at tip of ray, or not at all; spinelets of abactinal plates delicate,
similar to those of multispina.

Description.—This variable form seems to be a southern variety of multispina,
intermediate with Zeviuscula. A number of specimens have been classified as inter-
grades between dyscrita and leviuscula. Apparently also the transition to typical

@ Sboxpzzoc = hard to determine, referring to the status of the form.
57444°—Bull, 76—11——_19

290 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

multispina is gradual. The specimens may be considered from another point of
view as intergrades between leviuscula and multispina. I think it not improbable,
however, that dyscrita is an offshore variety or race of southern leviuscula, while
multispina bears a similar relationship to northern leviuscula. Certain intergrades
between true multispina and northern leviuscula are therefore difficult to separate
from some dyscrita. Owing to the extended distribution of the latter, occupying
as it does the shallower depths off middle and southern California, it has seemed
better to recognize it as a special race rather than as a series of intergrading forms
only. The hypothesis as to its derivation would make such a course preferable
also.

Rays of moderate thickness to slender; size rather small. R=28 mm.; r=7
mm.; R=4 r. Breadth of ray at base 9 mm. RK varies to slightly more or less
than 4 r.

Pseudopaxille typically smaller than in leviuscula, being about as large as
intervening papular spaces or even smaller, and not so compactly placed as in
multispina, in which, even if small, the plates are closely placed, leaving papular
areas decidedly smaller than the plates. A comparison of figures will demonstrate
this difference. There is less regularity in the shape of the pseudopaxille than in
typical multispina. Spinelets typically like those of multispina, rather long, slender,
ending in a few sharp points or awns and joined at the base by a slight web (often
not present, or not visible). The number of spinelets is quite variable, but usually
ranges from fifteen to twenty-five to a plate (less than in multispina). Papule
single, rather large.

There are three regular series of plates external to the adambulacral plates,
similar to those of multispina. These are less conspicuous than in leviuscula, the
middle row having the largest plates, and are a little wider, to decidedly wider
than long. <A short row of small intermarginals at base of ray.

Adambulacrals with one spinelet in the furrow except near tip of ray, where a
few plates have two. Adambulacral spinelets ten to fifteen, sometimes a few more
or a few less, the two or three largest standing on furrow margin, the others rapidly
growing smaller as they recede from furrow. There are two or three transverse
rows, and the spinelets themselves are round-tipped with numerous little prickles.
The armature is not essentially different from leviuscula variety B. The nearer to
leviuscula the specimen (as indicated by dorsal plates) the fewer the adambulacral
spinelets.

Type.—Cat. No. 27780, U.S.N.M.

Type-locality.— Albatross station 2907, near Point Conception, California, 44
fathoms, fine gray sand.

Distribution.—Middle and southern California, off shore to about 80 fathoms;
south of Monterey Bay this form predominates everywhere over leviuscula.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 291

Specimens examined.—Forty-three, from the following localities:

Specimens of Henricia leviuscula dyscrita examined.

1

Station. Locality. Depth. | Nature of bottom. Number. Collection.
Fathoms. |

2007. <= -.- Near Point Conception........... 44 1 U. 8. Nat. Mus.
2943...... Near Santa Barbara Islands...... el 2 | Do.
48 SFP Do
Steen ache ele tee ae ’ 26 1 Do.
eee een nee ee cna aaa 21 sand, stones...........- 7 Do.
Ee aase = Roane 27 fine gray sand. 4 Do.
=e aetna etek save sas eee 30 coarse mud... 1 Do,
2975¢..... Off Santa Cruz Island............ 36 gravel, broken shells. ca 1 Do.
3666...... Near Monterey Bay.............. 68 mud, sand, bowlders........ 1 Do.

44200..... Between Santa Barbara and San | 32-38 fine gray sand............... 6 Albatross, 1904.

Nicholas Islands.
Mae eo. Off Santa Cruz Island............ 30-41 mud, gray sand, rocks....... 5 Do.
44Aj'0..-..|| Monterey Bay-<...-.-.--.....-.-- 35-28 | gray sand, black mud....... 1 Do.
4532a..... 30 1 Do.
MODs occa foee <= OOSeo sae eteecdensetcace clawed 56 coarse sand, shells........... 1 Do.
abG4 Sco antewase Oe eee seccs ea b- has 60-80 | green mud, rocks............ 3 Do.
45580.....)....3 MO tS oo 8 oho co aten AD-O8. | COOKS cas Sc'fccn mana as 2a 2 Do.
Off Wilmington, California....... Wt, Dicdeetassateceanet mean ae aaeee 1 Do.
Off San Pedro, California......... 26-28 | green mud, sand...........- 3 Univ. of California.

a These specimens are more or less aberrant, being apparently intermediate with leviuscula. One from 4420 is aberrant and
not at all intermediate. See pl. 74, figs. 3, 4, 5.

HENRICIA LEVIUSCULA ANNECTENS Fisher.
Pl. 70, fig. 2d; pl. 71, figs. 1, 3.
Henricia leviuscula annectens Fisner, Zool. Anz., vol. 35, March 29, 1910, p. 572.

Diagnosis.—Intermediate in general appearance between H. leviuscula and
H. aspera, but superficially more like aspera; rays slender, disk small; abactinal
skeleton less open than in aspera, with smaller, deeply sunken papular areas; ridges
inclosing papular areas not undivided as in aspera, but subdivided into spine-bear-
ing tabule, or pseudopaxille; latter with comparatively few (five to twenty) short
tapering stubby spinelets in two or three rows; one to three papular pores to an
area; marginal plates shorter and much more compressed than in leviuscula, with
comparatively few spinelets, the spine-bearing ridges of the two series being some-
times confluent; successive ridges separated by a prominent transverse groove; no
intermarginals beyond first two or three marginals; actinal intermediate series
extending only part way along ray (one-half to two-thirds) as in aspera, not the
whole length as in leviuscula; adambulacral plates with ten to twelve spinelets in
two transverse rows, and deep in the furrow one spinelet, except near tip of ray,
where there are two.

Description and variations.—It is doubtful if a detailed description of this form
would be of any value. The figures wiil show the characters sufficiently well. There
are numerous variations from the typical form, especially in the size of the abactinal
spinelet groups. In some examples the pseudopaxille approach the size of those
of leviuscula; in others they are nearly as inconspicuous as in the southern specimens

of aspera.

292 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

The fact that this variety is intermediate in many of its characters between
leviuscula and aspera, that it is extremely variable, and is found frequently with
both aspera and leviuscula, or multispina, naturally suggests that it is a hybrid
between the forms mentioned. There is much to support this idea, for the speci-
mens vacillate between the two species, never quite reaching either; nevertheless
there are examples which it is difficult to classify, as they might be considered either
as aberrant specimens of aspera (or sometimes leviuscula) or as extreme variants
of annectens. The ‘illogical’? distribution of the variety makes one somewhat
suspicious also.

The chief variation is in the size of the abactinal pseudopaxille, as mentioned
above, these being narrow and with two rows of spinelets, or more oval with three.
The marginal plates are usually abruptly narrower than the abactinal directly above
them, the latter being sometimes in irregular transverse lines (as in leviuscula) with
small secondary plates here and there between the lines. The marginal plates in
appearance are about halfway between the thin, sparsely armed aspera (with large
intermarginal papular areas) and the conspicuous many-spined plates of leviuscula.
A character of aspera is the incomplete actinal intermediate series, which does not
run the whole length of ray. The adambulacral plates are nearer leviuscula than
aspera, there being two transverse series of spines, while in aspera there is but one.
The double-furrow spinelet far along ray is about halfway between the two forms;
in aspera the double spinelet may occur on most of the plates of the distal two-thirds
of ray, or only on a few distal plates; in leviuseula it occurs only on a few distal
plates, or not at all.

Madreporic body circular, slightly raised, similar to that of aspera, traversed
by a few radiating ridges bearing spinelets.

Type.—Cat. No. 27781, U.S.N.M.

Type-locality.— Albatross station, 3095, off Oregon, 42 fathoms, rocks, stones,
shells.

Distribution —From Washington to Santa Barbara Channel, California, 21 to
73 fathoms. é

Specimens examined.—Fifty-seven, as follows:

Specimens of Henricia leviuscula annectens examined.

Station. | Locality. Depth. | Nature of bottom. Number. Collection.
z | ai : |
Fathoms.
Biase Off Cape Flattery, Washington.-...... 27 rocks and shells.........-.-------- 19 U.S. Nat. Mus.
DBO Se estes ac OO Sevane ches one tance see ane eee 40 ee) Do.
Ab emerweloeiae GO sSesecect eas roe tenee ate e 59 1 Do.
2044.2... Near Santa Barbara Islands, California 30 1 Do.
Dob enssee | pees DO ease ae ee eee 48 fine\pray sands see eee | 2 Do.
2064s e.--2 Off Santa Barbara, California......... 21 Sand, StONeSso 2. «cnc sac ce ee cee race 1 Do.
2065...-.: I Seek do sn ay Ee Tiga ee davis) Ares 562 ea 2 Do.
3095......; Off Orford Reef, Oregon. 42 rocks, stones, shells. 2 Do.
3120......| Monterey Bay, California............. 54 green mud, sand...............-:- 1 Do.
3595. ....- 49 rocky, gravel, sand........-......- 4 Do.
3597 67 coarse black sand. .....-.-....---- 2 Do.
4441 35-28 | gray sand, black-mud..........-.- 1 Albatross, 1904.
4552......|-.c<« do 73-66 | green mud, rocks..........-...--+- Lied A Do.
|

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 2938

Remarks.—This form may be distinguished from aspera by the smaller papular
areas, the raised groups of spinelets (in aspera the spinelets are scattered along the
ridges surrounding papular areas and never form pseudopaxille), by the coarser
spinelets, larger marginal plates, and more numerous adambulacral spinelets,
. the outer of which are never immersed in membrane; there are two rows of adam-
bulacral spinelets instead of one. From Jeviuscula the present variety is separated
by the more open abactinal skeleton, smaller and more spaced pseudopaxille,
thinner marginal plates, and the incomplete actinal intermediate series. A
comparison of photographs will make the distinctions more apparent.

HENRICIA ASPERA Fisher.

Pl. 75, figs. 1-5.
Henricia aspera FIsHER, Proc. Wash. Acad. Sci., vol. 8, Aug. 14, 1906, p. 127.

Diagnosis —General form slender (but not invariably so); rays long, disk
small; skeleton forming a wide, open meshwork with large papular areas, which
are rather deeply sunken between the narrow ridges; abactinally the larger meshes
have seven to ten papule; spinelets very small, sharp, generally immersed in mem-
branous covering of body and arranged along the summit of the trabecular ridges
in one or two irregular series; a regular series of superomarginals, inferomarginals,
and actinal intermediate plates (the latter extending half or three-fourths length
of ray); a series of intermarginals extending a varying distance in some specimens;
also a second series of actinal intermediate plates; adambulacral plates with six to
eight coarse spinelets in a single zigzag or a double series, the two nearest furrow
margin more or less thickened and considerably larger than the rest; deep in furrow,
one or two spinelets.

Description.—Rays five. Type, R=100 mm.; r=15 mm.; R=6.6 r; breadth
of ray at base 14 mm. Aslendererrayed form: R=80 mm.;r=11 mm.; R=7.2r.
A large thick-rayed specimen with the arms swollen at base R=107 mm.; r=20
mm.; R=5.3 r; breadth of ray at base, 23 mm. The abactinal skeleton forms an
open meshwork, much as in H. hyadesi (Perrier), the individual plates being indis-
tinguishable and spinelets very short granuliform, not arranged in evident pseudo-
paxille as in leviuseula. Meshes are roundish quadrate, or irregularly polygonal,
more open in some examples than in others, containing sometimes one or two small
secondary ossicles with a few granuliform spinelets. Meshes usually considerably
wider than inclosing trabecule, and with five to twelve papule on proximal two-
thirds of ray, five to seven distally (but fewer in small specimens). Spinelets not
crowded but spaced, short, sharp, much shorter and fewer than in leviuscula, often
reduced to mere granuliform sharp elevations on the plate and more or less obscured
by a tight thin skin; arranged along ridges irregularly, but in not over three rows,
often in only one irregular series. These rows are interrupted, dividing the spinelets
and granules into groups probably corresponding to underlying plates, although
no divisions are evident. There are commonly five to fifteen spinelets in one of these
groups, but in some specimens they are so obscured by the superficial membranes
that only the very tips of the spinelets are visible. They are invisible to the naked
eye, and are seen only with a strong glass. Division into groups more evident on
sides of ray.

294 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Marginal plates regularly arranged. Superomarginal series departing frcm
interradial angle about midway between dorsal center of disk and inner angle of
jaw-plates; occasionally rather irregular near interbrachial angle; plates sometimes
transversely elongated, with ten to twelve spinelets. Inferomarginals slightly
larger or exactly equal to superomarginals; one or two rows of intermarginal plates
on basal fifth of ray; also one or two rows of actinal intermediate plates, two extend-
ing about one fifth length of ray, and one series for one half length, beyond which
point inferomarginals and adambulacrals are in contact. Intermarginal and
inframarginal papule; one to sixinan area. Marginal plates also form fairly regular
transverse series with adambulacrals, although latter are more numerous than
former.

Adambulacral plates with one small spine deep in furrow; on some plates,
especially in large specimens, a second may be present just above it and in line.
On actinal surface two blunt, stubby spines stand in an oblique transverse series
on furrow margin (frequently a group of three); and behind them three or four
much shorter graduated spinelets in a single zigzag series, all more or less united
by membrane. Armature varies greatly, sometimes two transverse series of spines
being present, and the spines themselves vary in shape from slender, cylindrical
and tapering to thick, clavate, and blunt. Armature generally has appearance of
being in a single series and rather sparse. The outer spinelets of some specimens
(those which have very minute spinelets generally) are buried in membrane and all
but invisible.

Madreporic body variable, usually subtubercular, roundish, with coarse
striations.

Color in life—Abactinal surface deep chrome yellow; papular areas deep
saffron yellow; actinal surface pale Indian yellow.

Type.—Cat. No. 21930, U.S.N.M.

Type-locality.— Albatross station 3052, Heceta Bank, Oregon, 48 fathoms, on
coral and broken shells.

Distribution—From Bering Sea (Bering Island, Pribilofs, Aleutian Islands)
south along the coast to the Santa Barbara Islands, California. Bathymetrical
range, 26 to 313 fathoms. On muddy, sandy, pebbly, shelly, and rocky bottom.

Specimens examined.—Seventy-six, as follows:

Specimens of Henricia aspera examined.

Station. Locality. | Depth. Nature of bottom. Number. Collection.
——_ 2 a
Fathoms.
RAD ecm: | Near Unalaska, Alaska...........---- | 72 DED DIOS sete sete enetsin sane ikea 2 U.S. Nat. Mus.
2857 4..... Near Afognak Island, Alaska | 68 shells, black specks........--.----- 2 Do.
PEG Leese Queen Charlotte Sound, British Co- ZOLS OE | Sn sate isiciees Sete so ia cee ee 2 Do.
lumbia.
ZEGE Sas on Admiralty Inlet, Washington 48 mud, broken shells, sand........-- 2 Do.
2800 case sa| esate CO hese tnetracents sen eee ao a 40 PED leSancasenes esate se eee 2 Do.
28754..... Off Cape Flattery, Washington 40 rocks'and|shells-~ <s-55.22--ces=p=- 3 Do.
Biliran | Sao WO ence een ge een See 59 black sand, mud...........-..---- 3 Do.
ealiemce=n Heceta Bank, Oregon 42 coral and pebbles... 1 Do.
2958¢..... Near Santa Barbara Islands, California 26 gray sand........... 2 Do.
DOB cs ces Off Santa Barbara........ccscsceceees 27 fine} Pray SANG <a eeneacieas messes 1 Do.

a Not typical.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 295

Specimens of Henricia aspera eramined—Continued.

Station. Locality. Depth. Nature of bottom. Number. Collection,
Fathoms.|
80 7 | |foosrawmudSjaaccees eee eden. 1 | U.S. Nat. Mus
59 coarse broken shells 1 | Do.
48 coral, broken shells 2 Do.
Ch lees COs zeae eae 1 | Do.
cine: Off Alseya River, Oregon........... 46 coral, pebbles 6 Do.
3095. ..... Off Orford Reef, Oregon ............ 42 rocks, stones, shells................ 1 Do.
OLE cas. Monterey Bay, California . ae o4 1 Do.
Bags o2. 2. Neat Unglaskas...s2-scc0se-.ceeeccote]|) 66 nae) Da,
$450...... Straits of Fuca, Washington....-..... 123 BIOY RANGA, soc note eeetsk te trees Vaal Do.
3462...... Admiralty Inlet, Washington......... 92 dark sand, rocky...... 1 Do.
3486...... Bering Sea, west of Pribilof Islands -. 150 green mud, fine sand.. 1 Do.
Oiaeccsse|saune! COs arc een eee satan twescoeecs 81 green mud, sand...... 6 Do.
4239...... Junction Clarence Strait and Behm | 206-248 | coarse sand, rocky................. 1 Albatross, 1903.
Canal], Alaska.
Rear Funter Bay, Lynn Canal, Alaska] 300-313 | mud.....................--------- 1 Do.
.| Shelikof Strait, Alaska ............... 65-48 | blue mud, sand, gravel............ 1 | Do.
eoken 30-41 mud, gray sand, rocks............. 1 | Albatross, 1904.
30 gray sand, rocks....... 1 Do.
86-76 | hard gray sand.................... 1 Do.
Salerno teeta! se cadeasn ae SSeee ee ee Cece tia ese) Kelme Srtoct hse Stoacahee eer eee oe 2 Do.
56-46 | coarse sand, shells, rocks.......... | 1 Do.
7a preen mud, TOOKS..... 2... cesscae 3 Do.
74-65 TOURS cose een esct nesses Vaccnacsa | 1 i Do.
66-69 | green mud, rocks.... 2 Do.
45582.....| Monterey Bay 40=28> ) | )TOOKSc 3-2. sean sc uees 1 Do.
WITT cos 22s Near Semisopochnoi Island, Aleutians. 52 fine gravel 1 Albatross, 1905,
47844..... Near Attu Island, Aleutians.......... 135 coarse pebbles 12 Do.
4790...... Bering Island, Commander Islands .- 64 pebbles 1 Do.
tC ERE! Bee el peer crh et eee wees eee 72 cae OD.25 scrcms uaceamin case coeees 1 Do.

a Not typical.

Remarks.—This species is easily recognized on account of the rather slender
rigid rays, very open and sunken papular areas, the rough minute spinelets, and
simple adambulacral armature. It is, however, variable, especially in the southern
part of its range, and specimens occur which are difficult to distinguish from /.
leviuscula annectens. There are indications that this species crosses with Jeviuscula,
multispina, and eschrichtii. It is possible that some of the specimens ranked as
annectens are hybrids.

In spite of the intermediate nature of many specimens (as for instance /7,
leviuscula annectens) I have kept this well-marked type as a distinct species, for
reasons mentioned in the introductory notes to the genus.

HENRICIA SPICULIFERA (Clark).
Cribrella spiculifera CLARK, Proc. Bost. Soc. Nat. Hist., vol. 29, No. 15, May, 1901, p. 328, pl. 2;
pl. 3, fig. 1.

Diagnosis —R =5 r; rays somewhat tumid near base; abactinal plates rather
large and not closely crowded, the plates much less crowded and the groups of
papule much more numerous than in /. leviuscula; plates irregularly rounded or
crescentic; marginal and actinal intermediate plates forming three rows all along
ray, the latter extending to tip; abactinal spinelets very numerous, longer and

296 BULLETIN 76, UNITED STATES NATIONAL “MUSEUM.

slenderer than in leviuscula; adambulacral plates with a single spinelet in furrow,
and with twenty-five to thirty spinelets on actinal surface, arranged in six or seven
series (with three or four to each) parallel to ambulacral furrow.

Description.—The following is the original description:

R:r ::64mm.: 14, or77mm.:15. R=5r. Breadth of ray near base about
i9 mm. Rays somewhat tumid near base, but tapering quite rapidly to a rather
attenuate tip; somewhat narrowed at base, and with rather deep interbrachial
sulci. Plates of abactinal surface rather large and not closely crowded, the groups
of papule being much more numerous than in H. leviuscula or sanguinolenta. The
plates are irregularly rounded, or crescentic, with the concave side toward the
center of the disk. They are irregularly scattered and show no evidence of
arrangement in rows, except along the sides of the rays, where they assume a longi-
tudinal arrangement. Near the tip of the ray, on each side, are three longitudinal
rows, the lowest of which consists of small plates lying close to the adambulacral
plates and running parallel to that series the whole length of the ray. The two
other rows consist of larger plates, and the lower of the two runs the whole length
of the ray parallel to the row of small plates. The upper one, however, diverges
from the lower about the middle of the ray and runs upward toward the abactinal
side of the disk, where it joins the corresponding row of the next ray. The narrow
triangular space between this upper row and the lower one is filled by three or four
short, irregular rows of somewhat smaller plates. Toward the tip of the ray the
plates of the three longitudinal rows are arranged in more or less regular transverse
series, but this arrangement is wanting near the base of the ray. Spinelets very
numerous on all the abactinal plates, rather longer and more slender than in
leviuscula, measuring 0.3 mm. in length. The abactinal plates are so much less
crowded and the spinelets are so much more numerous and slender than in leviuscula
that the surface does not appear granular, as it does in that species. Adambulacral
plates with a single, stout spine, 0.75 mm. long, high up in the groove, and with
twenty-five to thirty spinelets on the actinal surface. These spinelets are arranged
in six or seven series parallel to the ambulacral furrow, with three or four spinelets
in each series. The spinelets on the edge of the furrow are largest, measuring 2 mm.
in length, and each successive series consists of smaller ones, the last series bemg
very small. They are all closely crowded together. Madrepore plate single,
small (2 to 3 mm. in diameter), covered with spinelets arranged in ten to twelve
rows, radiating from the center. Color of alcoholic specimens uniform dark gray,
with a slight yellowish tinge in some lights.

Type.—Probably lost.

Type-locality—Probably Puget Sound, Washington.

Remarks.—No specimens of H. spiculifera have been recognized in the collec-
tions which I have examined, so that I can add nothing to the original description.
The two types are supposed to be at Columbia University, but are temporarily
or permanently lost. If the locality were certain H. spiculifera might possibly
be considered an extreme variant of the form I have named multispina. But in
the collection of which spiculifera formed a part there were species evidently col-
lected somewhere south of the United States, so that I am not at all convinced that
spiculifera is really from Puget Sound. :

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER.

297

In a general way, spiculifera and multispina resemble each other. Doctor
Clark examined several of the extreme variants of multispina and expressed his
opinion that none of them represented his species.

The following are the principal differences between spiculifera and multispina:

spiculifera.

Abactinal plates not closely crowded, much less
so than in leviuscula.

Abactinal plates
rounded.

Papular areas subequal to or larger than plates.

crescentic or irregularly

Actinal intermediate series of plates extends to
tip of ray.

Adambulacral plates with one spinelet in fur-
row, and with actinal spinelets (twenty-five to
thirty) in six or seven series parallel to furrow.

multispina.

Abactinal plates closely crowded, much more so
than in leviuscula.

Abactinal plates rounded or irregular,
crescentic.

Papular areas smaller than plates, and smaller
than in typical leviuscula, as the plates are more
crowded.

Actinal intermediate series of plates extends only
half to three-quarters length of ray.

Adambulacral plates with two spinelets in fur-
row on outer half to fourth of ray and with actinal
spinelets (twenty-five to fifty) not arranged in

not

series parallel to furrow.

The form of spinelet is highly characteristic in multispina and probably differs
from that of spiculifera, although it would not be safe to make comparisons without
specimens.

HENRICIA ASTHENACTIS Fisher.
Pl. 77, fig. 1; pl. 111, figs. 4, 4a.
Henricia asthenactis Fisher, Zool. Anz., vol. 35, March 29, 1910, p. 572.

Diagnosis.—General appearance somewhat like that of young aspere, but the
abactinal and lateral skeleton more open and weaker, the papular areas being pro-
portionately much larger and the spinelets scattered, few, and conspicuously longer;
adambulacral spinelets long, and in a single transverse series (four to six in number
counting the spinelet in furrow); two regular series of plates adjacent to adambu-
lacrals separated by a row of single papule.

Description.—Rays, five. Type: R=22 mm. : r=5mm; R=4.4r. Breadth
of ray at base 6mm. The rays taper from a broader base than in aspera to a sharper
point and are more depressed. The skeleton consists of a more open mesh work
even than in aspera, the ridges being narrow, and surmounted by a single row of
slender well-spaced spinelets about as long as the height of the calcareous ridges
above level of intervening integument. There appears to be only one or two
spinelets to each plate. The large meshes which constitute the papular areas are
often squarish, and frequently arranged in fairly regular longitudinal rows, so
that three or four rows may be seen when the ray is regarded from above (or six
from superomarginal across ray to superomarginal). It so happens that there is also
sometimes, but not always, a distinguishable median row of ossicles. The papular
areas, especially on disk, may contain one to three small isolated plates bearing one
or two spinelets. There are commonly upward of ten papule abactinally to each
area. While the papular areas are always large, they vary considerably in shape
and regularity and are largest on the median area of ray andon disk. It is not
possible to appreciate their extent accurately until specimen is dry.

298 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Parallel with the adambulacrals are two series of marginal plates, each of which
bears a group or comb of two to four conspicuous spinelets. A short series of actinal
intermediate plates is wedged between the inferomarginals and adambulacrals
proximally, and similarly a few intermarginal plates, bearing one or two spinelets,
occur between the two series of marginals proximally. In some specimens the
marginal plates are regular, in others irregular and difficult to follow.

The adambulacral spinelets are typically in a single transverse comb of three to
five on the actinal surface, the inner being the longest and much longer than width of
plate; the spinelet within the furrow is much shorter and slenderer than the longest
of the above—about one-third its length.

Madreporie body small and inconspicuous, with a few coarse striations, and
several spinelets on its surface.

Variations—Two much larger specimens (R=48 mm.) from station 4772, 344
to 372 fathoms, Bowers Bank, Bering Sea, have been referred with some hesitation
to this species. As compared with aspera the rays are much weaker, the skeletal
elements slenderer, the papular areas larger, and not nearly so sunken, the spinelets
more delicate, longer, and sharper, and the adambulacral spines longer and slenderer.
These specimens differ from those described above in having a series of intermarginal
plates nearly to tip of ray, and at the base two or three series. The actinal interme-
diate series is very variable in length, but does not extend to end of ray. The mar-
ginal plates are inconspicuous a2 bear only two or three spinelets.

Type.—Cat. No. 27782, U.S.N.M.

Type-locality.—_Station 4423, between Santa Barbara and San Nicholas Islands,
California, 339 to 216 fathoms, green sand shells.

Distribution.—Vicinity of Santa Barbara Island, California; Shumagin Islands,
Bering Sea (Bowers Bank and off Kamchatka), 178 to 682 fathoms. The northern
form differs more or less from the type.

Specimens examined.—Thirty-eight.

Specimens of Henricia asthenactis examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
2048255 - <0 Off Santa Cruz Island, California. . .. 266 gray sand, gravel, broken shells. . . 2 U.S. Nat. Mus.
3338......| Off Shumagin Islands, Alaska........ 625 green mud, sand.........-..- - 1 Do.
4410...... Between Santa Catalina and Santa | 178-195 | fine gray sand, rocks 7 Albatross, 1904.
Barbara Islands, California.
4416...... Between Santa Barbara and San | 448-323 | dark green mud, rocks...........- 1 Do.
Nicholas Islands, California.
aaa GO noone cccnce ned sc ecne reece ee oaxce)) 200-310)")) DIACK MUG KOCkS’.ceses sete 20 Do.
anes (OO SES aco ROSDSCE CENCE REcoREaaae 298 gray;mud,rocks:ee-- o7-ne eee aee 1 Do.
hee Ole aos ac 5. cas AACR ee 339-216 | green sand, shells...............-- 2 Do.
-| Off Santa Cruz Island, California..... 75-510) 4, preenomudes=s ose eee aaa eee 1 Do.
..| Bowers Bank, Bering Sea............ 344-372 | greenish brown sand..........---- 2 Albatross, 1906.
797. Off Petropavlovsk, Kamchatka....... 682 green mud, coarse black sand, 1 Do.

broken shells.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 299

HENRICIA LONGISPINA Fisher.
Pl. 76, figs. 1, 2; pl. 111, figs. 3, 3a.
Henricia longispina Fisner, Zool. Anz., vol. 35, March 29, 1910, p. 572.

Diagnosis.—Rays five, moderately robust, constricted at base, cylindrical,
tapering gradually to a very blunt, recurved extremity; disk small, slightly inflated;
skeleton forming fairly coarse mesh work with stout plates, and irregular, mostly
quadrate or roundish, though often irregular, papular areas, with rather numerous
papulz to an area; spinelets in well-defined, spaced fasciculate groups of two to nine
(usually five to seven) and 1 to 1.5 mm.long; extensive intermarginal series of plates;
adambulscral armature consisting of a zig-zag series of six or seven long bristling
spinelets decreasing in size from furrow edge, and in addition one spinelet deep in
furrow; actinal intermediate plates extending only about half length of ray.

Description —Type, R=47 mm., r=9 mm.; R=5.2 to 4.5 r; rays unequal.
Breadth of ray at base, 10 mm.; slightly beyond base,12 mm. Abactinal and lateral
skeleton forms a fairly coarse meshwork with stout plates which at frequent and
fairly uniform intervals bear on a slight elevation fasciculate groups of two to nine,
usually five to seven, unequal sharp radiating spinelets. The groups are largest
usually at the intersection of the skeletal ridges of plates. Between two such nodes
there is usually one smaller intermediate fascicule of two or three spinelets. These
spinelets, which are skin-covered and thicker in life than when dried, are finely
denticulate, sharp, and long for this genus (1 to 1.5 mm.), and much longer and
stouter than in any other species. Papular areas irregularly quadrate or roundish,
often broken up by a few isolated plates bearing one or two, or less often, more,
smaller spinelets than those of the regular plates. As viewed from the outside the
papular areas are irregular on account of the smaller secondary plates. Papule
usually five or six to an area on proximal half of ray, but groups of two or three are
sometimes found, and rarely as many as eight or nine. On disk the pseudopaxille
are closer together and papule range from about three to ten to an area. Anus
surrounded by several pseudopaxille, one probably representing the central plate,
having twelve spinelets.

A fairly regular series of pseudopaxille extending from interbrachial line,
about midway between abactinal and actinal surfaces, reaches tip of arm, and
represents the superomarginal plates. About midway between this and edge of
furrow is a similar series of inferomarginal plates which are a trifle larger than
superomarginals, The plates of both series bear six to nine or ten spinelets, and are
somewhat transversely elongated at base of ray. Between supero- and inferomar-
ginal series is a series of small pseudopaxille, with one or two to five smaller spinelets,
extending one-half to two-thirds length of ray, and at base of ray one or two addi-
tional incipient intermarginal series are added to the above. In the small papular
spaces between superomarginal and intermarginal series and between latter and
inferomarginal are about one to three good-sized papulw. Just external to adam-
bulacral plates is a series of very small actinal intermediate pseudopaxille extending
about one-half length of ray, each having two, three, or four spinelets. Between
this series and adambulacral plates is a series of single papulw, and another series
between inferomarginal and actinal intermediate plates. After the actinal inter-

300 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

mediate plates die out these two series of papulx merge into a single series (one or
two papulz to an area) between inferomarginals and adambulacrals.

Adambulacral armature consisting of (1) a saber-shaped spinelet situated deep
in furrow, unusually prominent, and capped by a short membraneous tip; when
bent outward this spinelet reaches about three-fourths across furrow. (2) On acti-
nal surface a rather long slender tapering bluntly pointed spine stands on margin
and behind it two similar spines in an oblique transverse series, the inner of these
two forming also an oblique series with the marginal spine. Behind these are three
or four slightly shorter and sharper spinelets, in a coordinate group, the outermost
spinelets the shortest of all. The spinelets stand in a zig-zag transverse series.
Occasionally the spinelets form a nearly straight series across the plate. Adam-
bulacral spines larger than any others.

Madreporic body large, situated near edge of disk and covered with numerous
spinelets similar to those of pseudopaxille.

Color in life, milky white; in alcohol, the same, or yellowish.

Variations—A smaller specimen from station 4228, vicinity of Naha Bay,
southeast Alaska, and consequently from near the type-locality, maintains the
general facies of the type, but has only one or two papule to each area, few inter-
marginal pseudopaxille, and spinelets which are relatively slightly longer than in
type (R=25 mm.).

Type.—Cat. No. 27783, U.S.N.M.

Type-locality.—Station 4199, Queen Charlotte Sound, off Fort Rupert, Van-
couver Island, British Columbia, 107 to 68 fathoms, soft green mud, volcanic sand.
Distribution.—F rom the vicinity of Vancouver Island to southern Alaska.

Specimens examined.—In addition to the type, one specimen from station 4228,
vicinity of Naha Bay, Behm Canal, southeast Alaska, 41 to 134 fathoms, gravel,
sponges.

Remarks.—This species is notable for the prominent fasciculated spinulation
and the prominence of the adambulacral spinelets. The spinelets are much longer
and sharper than in sanguinolenta, the groups much more spaced on account of the
open character of the skeleton, and the adambulacral spinelets much longer.

HENRICIA LONGISPINA ALEUTICA, new subspecies.
Pl. 77, fig. 2.

Diagnosis.—Similar in the character of its armature to Henricia longispina,
with which it agrees in having the spinelets in spaced fascicules, but from which it
differs in having a much more open abactinal skeleton with larger papular areas,
in having larger intermarginal papular areas, and extensive actinal intermediate
areas with numerous plates and papule.

Description —R =98 mm.; r=18 mm.; R=5.4 r. Breadth of ray at base, 20
mm. The whole skeleton is weaker and more open than in longispina and more as
in asthenactis, but the spinelets instead of being distributed singly along the skeletal
pieces (as in asthenactis) are in fascicular tufts and much longer than in that species.
Meshes of abactinal and lateral surface large, three-, four-, or five-sided, without
regularity, and ordinarily about four times the diameter of the skeletal pieces, while
in longispina they are not more than twice the diameter. The papular areas con-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 301

tain upward to ten minute isolated grains or plates, sometimes bearing a small
spinelet. Papulx conspicuous, three to ten to an area. Spinelets three to five to
a fascicule, delicate, pointed and thorny. They are relatively smaller than in long-
ispina, being 1 to 1.5mm. long. The tufts of spinelets are situated at the intersection
of the skeletal pieces, on slightly raised tabula or bosses, and also one or two may
occur on the internodes. The spinelets attached to the grains in the papular spaces
are much smaller.

Marginal plates small, the two series well separated by large irregularly quad-
rate papular areas containing four or five papule. Occasionally one of these areas
is bisected longitudinally. Intermarginal plates small, extending’ nearly to end of
ray. They are one of the elements of the transverse bars connecting infero- and
superomarginal plates. The latter are about the same size as the abactinal pseudo-
paxille, but the inferomarginals are larger, with about eight to ten spinelets.

Actinal intermediate plates small, extending about half length of ray (as in
longispina), but on the disk and base of ray there are numerous plates, as the
inferomarginal series bend outward and reach the ambitus at the interradial angle.
Along the interradial line from the mouth to the inferomarginal plate five or six
intermediate plates can be counted. These bear one to four short papilliform
spinelets. A row of papule along ray between inferomarginals and intermediate
plates and numerous papule among intermediate plates of disk.

Adambulacral plates with an unusually long spinelet in the furrow (2 mm.)
and an irregular transverse row of five or six spines, of which the inner is long and
slender (twice as long as width of plate), the next two or three decreasing gradually,
and the outer one, two, or three are commonly much shorter than the rest. Some
of the longer spines are flattened, grooved, or bifid at the tips. The marginal
mouth spines are often conspicuously grooved or flattened.

Type.—Cat. No. 27784, U.S.N.M.

Type-locality.— Albatross station 4784, near Attu Island, Aleutian Islands, 135
fathoms, coarse pebbles; one specimen.

Remarks.—Although the specimen upon which this race is based is quite dif-
ferent in general appearance from longispina, an analysis of its characters shows
many points of close similarity. The general type of armature, including that of
the adambulacral and mouth plates, is the same. Both species have the long
furrow spinelet and the bristling series of adambulacral spines, with certain trivial
differences.

The features in which the two forms depart the most widely from one another
are in respect to the wide-meshed skeleton of alewtica and the larger actinal inter-
radial areas. The abactinal papular areas contain numerous minute plates or
grains, some of them bearing small spinelets.

The present form is much more closely related to longispina than to any other
known species, and in view of the many points of resemblance it has been made
a race of that species. More specimens may show aleutica to be even nearer
longispina than the type indicates.

302 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
HENRICIA POLYACANTHA Fisher.

Pl. 78, figs. 1, 4; pl. 111, figs. 2, 2a-b.
Henricia polyacantha Fisuer,>Proc. Wash. Acad. Sci., vol. 8, Aug.:14, 1906, p. 129.

Diagnosis.—Rays moderately slender (R=about 6 r), very flexible, tapering
gradually to bluntly pointed, upturned tip; abactinal surface usually more or less
collapsed; abactinal and lateral surfaces with small evenly spaced paxille; single
large papule arranged in fairly regular oblique transverse series on either side of
median radial line (at least when viewed internally); very small marginal plates;
a very regular series of inframarginal papule; adambulacral spinelets very numer-
ous, and on the furrow face of plate two to six small saber-shaped spinelets.

Description.—Rays, five. R=66 mm.; r=11 mm.; R=6r. Breadth of ray
at base, 13mm. Abactinal and lateral surfaces closely covered with small, unequal
pseudopaxille, leaving, however, papular areas considerably larger than the plates;
papule, one to an area, large. Without aid of a glass the papular areas appear
roundish. Paxillee more or less elongated in one direction; convex, covered with
exceedingly small spinelets, which are numerous, but vary greatly in number,
according to the size of pseudopaxilla; ten to forty is the usual number. Paxille
form a more or less evident median radial line along ray. Frequently two rows of
pseudopaxille intervene between two papular areas or pores.

External to adambulacral plates is a regular series of actinal intermediate
plates, and separated from the latter by a conspicuous regular series of papule is
a row of smaller inferomarginals. Separating the regular papule of this series is a
second row of small transversely oriented compressed plates which do not extend
the whole length of ray. There is considerable variation in this character; the
marginal plates are often quite irregular. Sometimes a superomarginal series can
be distinguished just above the inferomarginals, especially on outer part of ray,
where the two series are fairly regular. At base of ray the serial arrangement is
broken up and two or three additional series of small intermarginal plates are
interpolated, a single series of very small ones extending far along ray.

Adambulacral plates separated by a distinct suture. Armature very dense,
consisting of many spinelets, as follows: (1) on furrow face of plates two to six
small saber-shaped spinelets in a vertical series or group. The. number varies in
different individuals. Usually there are four to six at base of ray and two or three
to each plate beyond middle. Rarely specimens have more than three on plates of
distal portion of arm. (2) On actinal surface of plate are thirty to forty slender
pointed spinelets arranged in three or four transverse series on inner half of plate,
but too crowded on outer half to form rows. Even the inner spinelets are often
without regularity. Spinelets decrease rapidly in length and caliber from the
furrow outward, the outer spinelets being sharper than the inner and about the
same size as those on other actinal plates.

Mouth plates small, triangular, with about four to seven small papilliform
spinelets on furrow face. The actinal surface is densely covered with spinelets, of
which eight or nine stand on the free margin.

: ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 303

Madreporic body prominent, tubercular, situated midway between center of
disk and interbrachial angle, there being small spinelets scattered on the surface.
Striations coarse, irregularly radiating.

Color in life unknown; in alcohol whitish.

Anatomical notes.—The plates are rather delicate and slender, as would be
expected in a species of this genus living at such a depth. The best idea will be
gained from the figure. The papule are single, and when viewed from the inner
surface of the integument are seen to be arranged in regular oblique transverse
series extending between the well-defined median radial series of plates and the
superomarginals. Externally the regularity is not very evident.

Type.—Cat. No. 21931, U.S.N.M.

Type-locality.— Albatross station 2936, off San Diego, California, in 359 fathoms,
on mud.

Distribution.—Known from the vicinity of type-locality only.

Specimens examined.—From type-locality, four; from station 2935, near type-
locality, 124 fathoms, fine gray sand, one specimen; locality unknown, one specimen.

Remarks.—This species is characterized by the presence of several spinelets
on the furrow face of the adambulacrals and by the small pseudopaxille covered
with numerous very minute spinelets; by the large single papulw regularly arranged
in oblique transverse rows, and by the small mouth plates and numerous adambu-
lacral spinelets. It stands near to ZZ. clarki. and to H/. pauperrima Fisher of the
Hawaiian Islands.

HENRICIA CLARKI Fisher.

Pl. 78, figs. 2, 3; pl. 111, fig. 5; pl. 112, fig. 3.
Henricia clarki Fisner, Zool. Anz., vol. 35, March 29, 1910, p. 573.

Diagnosis.—Rays long, very slender, and flexible, tapering from a narrow
base to an attenuate, blunt extremity; rays subcylindrical; abactinal surface
depressed; interbrachial angles abrupt, about 90°, not rounded; disk very small;
plates delicate, forming a sinuous meshwork inclosing smaller, lower, secondary
plates (also more or less joined in a close mesh) between which emerge single papule;
adambulacral plates with thirty-five to forty spinelets on proximal part of ray, and
in the furrow two spinelets, except on the proximal six or eight plates, where there
are three to five in a vertical series.

Description —R=54 mm.; r=7 mm.; R=7.7 r. Breadth of ray at base,
7 to 9mm. Abactinal surface with an open meshwork of narrow small primary
plates, bearing contiguous roundish or elongate groups of about a dozen very small,
slender, and sharp spinelets which are not so long as the height of the ridges bearing
them. These primary plates form sinuous ridges inclosing fairly large, very irregular
spaces or meshes which are filled by lower, smaller plates bearing single spinelets
or groups of two to six. These secondary plates in turn form a close meshwork
filling the larger primary meshes, but the summit or top of the secondary plates is
sunken below the level of the primary ridges and the secondary spinelets are also
smaller and fewer in a group. This mesh within a mesh is very characteristic of
this species. In the intervals between secondary plates single papule protrude.

3804 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Toward end of arms primary meshes become smaller and the distinction is not so
evident.

On basal third of ray, adjacent to adambulacral plates, is a series of fairly
regular plates slightly larger than other plates of general perisome, having twenty
to thirty spinelets, but this series soon becomes irregular. One or two tiers of
plates above this series there is an indication of a very irregular marginal series, but
it does not extend beyond basal fourth of ray. Meshwork on actinal and lateral
surfaces much closer than on dorsal, the secondary plates playing a very subsidiary
part in the make-up of skeleton. Frequently two or three or even more papule
occur in the papular spaces.

Adambulacral plates about twice as wide as long, separated by a distinct and
conspicuous interval or suture. Armature dense, consisting of (1) two small saber-
shaped spinelets placed one above the other high in furrow, except on first six to
eight plates, which have a vertical series of three to five. Occasionally on distal
portion of ray there is but one furrow spinelet, but two is the usual number. (2) On
actinal surface of plate are thirty-five to forty slender pointed spinelets arranged in
three or four transverse series on inner half of plates and too crowded to be regular
on outer half, decreasing rapidly in length and caliber from the furrow outward,
the outer spinelets being sharper than the inner.

Mouth plates forming a salient angle, closing actinostome, the sutures between
first adambulacrals and mouth plates marking off a regular pentagon inclosing the
ten mouth plates. Of this pentagon the prominent interradial dental sutures mark
the radii, while the closed ambulacral furrows form equally regular interradii.
Armature very dense, the six or seven marginal spinelets being somewhat enlarged,
much flattened, and with blunt or truncate tips; two or three small spinelets form
a longitudinal series on furrow face of the plates. The mouth plates are larger and
more conspicuous than in H. polyacantha. (See Pl. 78, fig. 3.)

Madreporic body prominent, situated midway between center of disk and
interbrachial angle, the surface being covered with tufts of small spinelets.

Color in alcohol, whitish.

Anatomical notes.—The appearance of a mesh within a coarser mesh presented
by an exterior view of the plates is due partly to the presence of the spinelets which
hide the slight gaps between the small secondary plates, although many of the latter
are firmly united into a ridge as the figure will show. Undoubtedly the plates are at
first independent grains and later coalesce. Comparison of the figures of H.
polyacantha and H. clarki will demonstrate the difference in the skeleton structure.
H. clarki has no sign of the regular arrangement of the papule, characteristic (so far
as known) of polyacantha. (Compare Pl. 112, fig. 3, with Pl. 111, fig. 2b.)

Variations.—The specimens from stations 2923 and 2935 are placed with this
species not without serious misgivings. The lack of several specimens from the type-
locality by which to determine the range of individual variation has led me to place
the specimens in question with clarki, although at first sight they look like a very
different form. The abactinal spinelets are fewer, and more spaced as regards the
small clusters. The meshes are very irregular, but in the example from 2935, which
is large (R=89 mm., r=12 mm.), the characteristic mesh within a mesh formation

ASTEROIDEA OF NORTH PACIFICO AND ADJACENT WATERS—FISHER. 305

is largely obliterated by the growth of the secondary intermediate plates to nearly
or quite the coarseness of the primary. This gives a finer-meshed appearance to
the abactinal surface. Two young specimens from station 2992 are remarkable in
having three or four spinelets in a vertical series on the furrow face of the adambu-
lacrals and fewer adambulacral spinelets, of which those near the furrow are abruptly
much larger than the outer ones; the last is true also of the specimens from 2935
and 2923. In these specimens the spinelets near the furrow edge (three to five) are
relatively larger than the outer ones, than is the case in the type-specimen. They
are thus more like pauperrima, but that species has square adambulacral plates and
more numerous abactinal spinelets; this species has the spine-bearing surface of the
adambulacrals wider than long. The specimens from 2992 (Revillagigedo Islands)
are too young to make their identification with this species certain; they are, how-
ever, nearer this form than to pauperrima, or any described species.

Type.—Cat. No. 27785, U.S.N.M.

Type-locality— Albatross station 4427, off Santa Cruz Island, California, 475 to
510 fathoms, black mud, broken stones; one specimen.

Distribution —Santa Cruz Island, California, to the Revillagigedo Islands,
Mexico, 124 to 822 fathoms.

Specimens examined.—Besides the type, four from the following stations:
2992, Revillagigedo Islands, off Mexico, 460 fathoms, black sand, rocks, two; 2923,
off San Diego, California, 822 fathoms, green mud, one; 2935, same locality, 124
fathoms, fine gray sand, one.

Remarks.—Henricia clarki differs from H. polyacantha (which it resembles
in having more than one furrow spinelet and numerous adambulacral spinelets)
in the form of the abactinal skeleton, in having the papule very irregularly arranged,
in having no distinguishable series of marginal plates, in having much larger and
more conspicuous mouth plates. The spinelets are longer and sharper. From J//.
pauperrima, of the Hawaiian Islands, to which clarki bears considerable outward
resemblance, the species differs in having more numerous adambulacral spinelets,
and oblong instead of square adambulacral plates. In pawperrima there is an indi-
cation of the characteristic skeletal structure of clarki, but the primary meshes
are smaller and do not inclose numerous secondary plates.

This species is named for Mr. Austin Hobart Clark.

Family SOLASTERIDZ Perrier, 1884, emended.

Spinulosa with a more or less open and irregular reticulate abactinal skeleton
(though exceptionally with small roundish or irregular independent plates), the
plates bearing paxilliform groups of spinelets; marginals paxilliform, in one or two
series; adambulacral spines in a furrow comb and a series at right angles to this;
actinal intermediate plates present; with interbrachial septa; no pedicellariw; no
supradorsal membrane; anus present; tube feet in two series with well developed
sucking disks.

57444°—Bull. 76—11——20

806 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

KEY TO THE KNOWN GENERA OF SOLASTERIDE.

a', Marginal paxille in either one or two series, but if in two then only the inferomarginals well devel-
oped, the superomarginals being much smaller, and similar to but larger than abactinal paxille.
Rays more than five.
b!. Superomarginal plates, if distinguishable, in a series distinct from inferomarginals; not one or
two superomarginals alternating with one inferomarginal in a single linear series.
c!. Abactinal plates not independent, but joined by their lobes, or by means of secondary inter-
mediate ossicles.
d'. Actinal intermediate plates extending only part way along ray, or confined to disk and not
bearing a comb of spines but a paxilliform tuft...........-..-.--- Solaster Forbes, p. 306.
@. A complete series of actinal intermediate plates, each bearing a curved comb of spines, extend-
ang to. end. Of Taya a o<niafi season ec ieee ieee em rae een eee Rhipidaster Sladen.
ce, Abactinal plates small, unequal, roundish, very numerous and independent. Actinal inter-
mediate plates extending nearly two-thirds length of ray ........-...-- Lextmaster® Fisher.
b?. Marginal plates of two kinds in a single linear series—conspicuous transversely oriented, promi-
nently spinous, paxilliform plates alternating with one or two low, longitudinally placed plates
with short spinelets; small actinal intermediate plates extending to end of ray; abactinal
skeleton wide meshed; paxillee small; papulee numerous and large. . Heterozonias Fisher, p. 330.
a?, Marginal paxille in two prominent, nearly equally developed series, rays 5. Lophasterb Verrill, p. 334.

Genus SOLASTER Forbes.

Solaster Forpes, Mem. Wernerian Soc., vol. 8, 1839, p. 120. Type, Asterias endeca Linnseus.
Crossaster MULLER and TroscHEL, Monatsber. preuss. Akad. Wiss. Berlin, April, 1840, p. 103;
Archiy Naturgesch., 6 Jahrg., vol. 1, 1840, p. 183. Type, Asterias papposa Linneeus. ~
Solaster (subgenus Endeca) Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p.183. Type, Solaster

endeca.
Solaster (subgenus Polyaster) Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 183. Type, Solaster

papposus.

Diagnosis.—Solasteride with one series of well developed marginal paxille,
the superomarginals being always markedly smaller than the inferomarginals;
abactinal skeleton of cruciform or stellate plates, either closely placed or forming,
by means of slender intermediate ossicles, an open meshwork. Abactinal pseu-
dopaxille spaced or crowded, small to large, fascicular, tabulate, or penicillate.
Actinal intermediate plates extending more or less along ray. Rays seven to
fourteen.

a Lxtmaster Fisher, Smiths. Miscell. Coll. (Quarterly Issue), vol. 52, 1908, p. 88. This is a substi-
tute name for Ctenaster Perrier; type, L. spectabilis (Perrier). L. Agassiz, in the Mém. soc. sci. nat. Neu-
chatel, vol. 1, 1835, p. 192, used Ctenaster as a substitute name for Asterina Nardo, 1834. Its status
corresponds to that of Cribrella. This prior use of Ctenaster (‘‘Once a synonym, always a synonym”)
necessitated the renaming of Perrier’s genus.

I examined the type in the Museum of Comparative Zodlogy (No. 910). Perrier, in my opinion,
has overlooked the chief character by which this form may be separated from Solaster, namely, the
structure of the abactinal skeleton. The presence of actinal intermediate plates ‘‘rather far along the
ray” upon which he places primary importance may be duplicated in Solaster (in its narrowest sense).
But the abactinal plates are very many, small, unequal, delicate, and roundish, with a low column
bearing three to six delicate sharp diverging spinelets. They appear to be quite independent, and
not forming a meshwork as in other genera of the Solasteride. Papule extending half way to end of ray.

6 Including Sarkaster Ludwig.

r

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 307

KEY TO THE SPECIES OF SOLASTER HEREIN DESCRIBED.

a'. Two rows of marginal plates clearly distinguishable, the superomarginals much the smaller and
situated just above the inferomarginals, though sometimes alternating with them. Abactinal
skeleton close meshed.

b'. Furrow spinelets much shorter than the actinal adambulacral spinelets, which are not conspicu-
ously long and bristling. Abactinal pseudopaxille small, spaced or crowded; marginal plates
of both series low and not forming a conspicuous border.

cl. Pseudopaxille very small and crowded; disk larger, rays shorter; inferomarginal pseudopaxille
larger with upward to thirty-two spinelets; superomarginal pseudopaxille very small.

y endeca, p. 307.

c?. Pseudopaxille larger (but with few spinelets) and more spaced; rays longer, disk smaller.
Less difference in size between supero- and inferomarginal paxille, the latter with upward
to eighteen spinelets (a combination of all these characters necessary rather than a single one
without the others, individual variation being wide in both endeca and stimpsoni).

stimpsoni, p. 311.

b?. Furrow spinelets long, reaching nearly to tip of actinal spines, which form a bristling fringe all
along either side of furrow; pseudopaxillze more or less tabulate and flat-topped except in
exiguus, and larger than in 6'. Marginal plates forming a conspicuous border to rays.

c!, Actinal adambulacral comb of spines straight, not curved aborally at inner end; actinal inter-
radial areas small; superomarginals typically opposite inferomarginals....... dawsoni, p. 313.

@. Actinal adambulacral comb of spines curved at inner end; superomarginals typically alternating
with inferomarginals.

d', Disk large, actinal interradial areas large; inferomarginal paxille prominent, much broader
than long, and somewhat fan-shaped, the spinelets numerous and not longer than the
pedicel; furrow spinelets proximally three or four, suboral spines numerous, about
iL Gen ees eee ns ne eeme enone oan eee Ueene peo e eames paxillatus, p. 315.

@, Disk small; actinal interradial areas small, inferomarginal paxille subcircular, not fan-
shaped, well spaced, with a tuft of delicate spinelets longer than the pedicel; furrow spines
proximally two or one, unusually long; suboral spines two to four........ exiguus, p. 319.

a, Marginal paxillz of one kind, very conspicuous, well spaced, in a single series (superomarginals
practically indistinguishable from abactinal paxille).

b!. Abactinal skeleton not wide meshed; paxille not large and penicillate, but small with two to
seven short spinelets; marginal plates with a few prominent sharp spines; a bare interradial
streak along each interradius.

c!. Suboral spines small, one or two, or none; actinal comb of adambulacral spines with two to
five, usually four, sharp spines; furrow spinelets small, typically much shorter than the first
actinal (less than half); abactinal pseudopaxille with three or four short spinelets.

borealis, p. 320.

©. Suboral spines five to seven very prominent; actinal comb of adambulacral spines with six
long spines; furrow spinelets typically more than half as long as first actinal; abactinal
pseudopaxillz higher, with seven to ten spinelets..............-.------ hypothrissus, p. 324.

b?. Abactinal skeleton wide meshed; paxille large, spaced, and penicillate; marginal paxille large

with numerous prominent spines.

c!, Furrow spines usually three or four not joined by a web to the actinal comb; marginal paxille
less numerous, and separated from each other by more than their own width.

papposus, p. 325.

c. Furrow spines seven to nine webbed, and joined by a web to the actinal comb; marginals

more numerous and separated by less than their own width............... japonicus, p. 330.

SOLASTER ENDECA (Linnmus).
Pl. 81; pl. 82, figs. 1, 2, 4.

Asterias endeca Linn£v8, Mantissa Plantarum, Appendix, 1771, p. 543.—Rerzivus, Nya Handl.
Kongl. Svenska Vet. Akad., 1783, p. 237.

Asterias aspera O. F. Mitrer, Zoologia Danice Prodromus, 1776, p. 234.

Asterias alboverrucosa Branpt, Prodromus, 1835, p. 71 (according to Ludwig).

3808 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Stellonia endeca Acassiz, Mém. soc. sci. nat. Neuchatel, vol. 1, 1835, p. 192.

Solaster endeca Forses, Mem. Wernerian Soc., vol. 8, pt. 1, 1839, p. 121. Authors since.a—
Verritt, Amer. Nat., vol. 43, 1909, p. 553 (Bering Sea).

Solaster (Endeca) endeca Gray, Ann. and Mag. Nat. Hist., vol. 6, 1840, p. 183.

Solaster galavides Verrw1, Amer. Journ. Sci., vol. 28, July, 1909, p. 59, fig. 2 (Victoria, British
Columbia).

Diagnosis —Rays seven to thirteen, usually nine to eleven; disk broad; rays of
moderate length. Distinguished by the numerous small crowded paxille with very
small spinelets; papulsz usually single; inferomarginal plates low, close together,
transversely elongated; superomarzinals small, but little lence than adjacent
paxille; furrow spinelets two or three, exceptionally four, short, not reaching end
of actinal spines; latter in a curved series of six to eight, the inner longest, but not
of conspicuous length, the longest not exceeding length of base line of comb.

Notes on North Pacific s pecimens.—Before the publication of Professor Verrill’s
Descriptions of New Genera and Species of Starfishes from the North Pacific Coast
of America,’ in which Solaster galazides is described, I had set apart the specimens
here listed, as S. endeca. Professor Verrill kindly identified two examples from
station 4246 as his S. galarides. I then went over the whole genus again in the light
of other identifications made by Professor Verrill, but have not been able to coincide
with his views in regard to the present form. From station 2851, off the Shumagin
Islands, Alaska, 35 fathoms, is a fairly large nearly typical Solaster endeca. A com-
parison of this with the so-called Solaster galaxides from southern Alaska and other
localities has not brought to light any constant differences. Rather has it enforced
the conclusion that galarides is not a distinct species, but at most a North Pacific
variety of the variable S. endeca.

For comparison Atlantic specimens of S. endeca from the following localities
have been used: off Cape Cod, 334 to 834 fathoms (two); Bay of Fundy (one);
west coast of Sweden (one). These present considerable variation among them-
selves, the two examples from off Cape Cod exhibiting almost the maximum of
difference. Grieg, in Echinodermen von dem norwegischen Fishereidampfer
Michael Sars in den Jahren 1900-1903 gesammelt indicates* the variability in pro-
portions and in the number of rays of Atlantic specimens, while a good description
may be found in Clark’s Echinoderms of the Woods Hole Region.?

Number of rays, proportions.—Alaskan examples have nine, ten, or eleven rays,
usually ten. Atlantic specimens vary in this respect from seven to thirteen, having
commonly nine to eleven, and occasionally seven, eight, twelve, and thirteen.

There are not constant differences in proportions between Atlantic and Pacific
specimens. For instance, a nine-rayed example from off Cape Cod has R=3.82 r,
while the Swedish specimen has R=2.35 r. In the large specimen from off the Shu-
magin Islands, R=2.46 r, and another from station 4246 has 3.3 r. Another from
the same station has R=2.2 r. The types of S. galaxides have the ratio R=2.7 r.

« Ludwig in Fauna Arctica, vol. 1, 1900, p. 464, gives a considerable number of references to this
species. S. syrtensis Verrill, however, is not a synonym of endeca.

b American Journal of Science, July, 1909.

¢ Bergens Museum Aarbog, 1906, No. 13, p. 66.

@ Bull. U. S. Fish Comm. for 1902, 1904, p. 556; pl. 3, figs. 18, 14; pl. 4, fig. 23.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 309

Abactinal surface —Abactinal pseudopaxille typically very small and crowded.
Specimens from 2851 and 4246 are exactly matched by an example from off Cape
Cod. There is less difference between these than between two Cape Cod examples,
or between the latter and a Swedish specimen. In large Pacific specimens the
spinelets of proximal midradial paxille vary from five to about nine or ten. In the
Swedish example there are commonly as many as nine, less often ten, and as few as
three or four spinelets. The paxillw on sides of rays are usually very slightly larger
than those of the midradial regions and center of disk.

Marginal plates—In the original description of S. galaxides: ‘The marginal
spines are about as in S. endeca, but the inferomarginals are more elongated trans-
versely, and bear a decidedly greater number of more minute spinules.”” The
available specimens have been examined with reference to this character. In two
equal-sized examples, one from off the Shumagin Islands and one from off Cape Cod,
the inferomarginals are the same size, or if anything a little larger in the latter (that
is, more elongated transversely). An average plate in the Alaskan example bears
thirty-three spinelets, while a corresponding plate in the Atlantic specimen has
thirty-one, and in the Swedish example thirty-two; neighboring plates in all three
specimens vary three to eight on either side of these figures. Some of the specimens
from 4246 have the inferomarginals decidedly narrower transversely than any of the
above examples; while among all the available material none can be said to have
the inferomarginals more elongated transversely. The characteristic is evidently
individual and not specific. Again comparing the spinelets for size, we find varia-
tion in both Atlantic and Pacific material; the Shumagin Islands example (several
times referred to) has coarser spinelets than Atlantic specimens. Specimens from
4246 (“S. galawides’’) have the spinelets variable, some slightly finer, some identical,
some slightly coarser than the Cape Cod material. The differences in the infero-
marginals of Atlantic and Pacific specimens are no greater than between different
individuals from the same locality, or from different localities of the same ocean.

The superomarginals are somewhat more conspicuous than the immediately
adjacent abactinal pseudopaxillx, carrying twelve to fifteen spinelets near the base
of ray. They vary in size with the inferomarginals, and as in Atlantic specimens,
are sometimes opposite the interspace between two inferomarginals, or else directly
above and close to the latter. In shape they are roundish, irregular, or less often
subquadrate.

Actinal interradial areas —In the description of 8. galazides: ‘‘The synactinal
series of pseudopaxille extends only to about the basal third of the free part of the
ray. They are relatively smaller than in endeca, being here only about half size of
the inferomarginals proximally. The actinal interradial areas are apparently rela-
tively larger than in endeca and bear a larger number of compressed pseudopaxille,
the larger ones similar to the inferomarginals and synactinals.”

The size of the actinal interradial areas depends upon the number of rays and
breadth of disk, and while it is possible, by comparing small disked Atlantic endeca
with larger disked Pacific examples, to find larger interradial areas in the latter, it is
also true that a reverse process will reverse the results. As a matter of fact, there is
certainly too little difference to be of importance. In the Shumagin Island example

310 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

the synactinals, instead of being smaller than in Atlantic endeca, are distinctly
larger, on comparison; but in specimens from 4246 the synactinals are just about
the same as in Cape Cod examples. The distance which the actinal intermediate
plates extend along the ray is variable in both sets of specimens. In specimens
from 4246 the plates extend slightly less to over one-half the length of free part of
ray. A comparison of specimens with reference to the above features does not
enforce the differences alluded to in the description of galaxides.

Adambulacral and dental armature—Furrow spinelets two or three, rarely four,
or near tip of ray only one, and conspicuously shorter than the actinal comb. There
is, however, some variation in the length of the furrow spinelets (in both Atlantic
and Pacifie examples). On the whole, the Pacific examples do not seem to have
longer spinelets than the Atlantic, and the limits of variation are about the same in
the two lots. When there are two furrow spinelets either one may be the longer,
or the two subequal; when three are present usually the median is longest and some-
times one is rudimentary. The actinal comb is curved (not straight as in dawsoni)
and composed of six to eight sharp skin-covered spinules, the two or three inner the
longest and stoutest, the others graded in length toward the outermost. These
spines stand little or not at all above the level of the marginal pseudopaxille, being
therefore different from dawson and pazillatus.

The armature of the mouth plates is typical of endeca—seven to nine marginal
spinules, of which the inner two or three are much larger than the rest, and six to
fifteen suborals, in two series (or a roughly triangular group, the inner spinules
being the larger).

Ty pe-locality —‘‘In Oceano Norvegico” (Linneeus).

Distribution.—Bering Sea south to Victoria, British Columbia (not yet known
from the Asiatic coast). In the west Atlantic from Cape Cod to Labrador, thence
to Greenland and Davis Strait, Iceland and the Faroe Islands; Ireland, the Irish
Sea, Orkneys, Shetland Islands, Scotland, and England (south to lat. 55° N.); north-
ern North Sea to the Kattegat, thence along the Norwegian Coast to Finmark, thence
to Spitzbergen and westward to lat. 80° 03’ N., the most northern station yet
known; eastward through Barents Sea, along the Murman coast, Kara Sea, thence
along the Siberian coast to the Gulf of Khatanga (long. 113° 30’ E.). From here
to Bering Sea the species is not yet known. (Emended, from Ludwig, Fauna
Arctica, Vol. 1, p. 465.)

Specimens examined.—Twelve, from the following localities:

Specimens of Solaster endeca examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
| Fathoms. | |

2851......| Off Shumagin Islands, Alaska........ | 35 gray sand, broken shells......... 1 U.S. Nat. Mus.
2862......, Queen Charlotte Sound, British Co- | 238 | gray sand and pebbles.......... 1 Do.

lumbis.
4246...... Kasaan Bay, Prince of Wales Island, | 123-101 | gray green mud, coarse sand ... -} 7 Albatross, 1908.

Alaska. |
4272......| Afognak Island, near Kadiak Island, 17-12 | sticky: mud {252222 2o-n ee tees 2 Do.

Alaska.

1 Stanford University.

Rearlulky Mada: on ccnsncaceacewes ace Low tidet)';.~ .~cesenacesn ecisser ee ceenease
| | |

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 311

Remarks.—If the specimens listed above are true endeca, as they seem to be,
the known range is thereby considerably increased. The form described by Brandt
as Asterias endeca, var decemradiata, and since called Solaster endeca decemradiata
probably refers to Solaster stimpsoni, but the name is a nomen nudum and does
not invalidate stimpsoni.

SOLASTER STIMPSONI Verrill.
Pl, 82, fig. 3; pl. 83, figs. 1-5.

Asterias endeca, var. decemradiata Branvrt (?), Prodromus 1835, p. 71 (nomen nudum), Sitka.

Solaster decemradiatus Stimpson (?), Boston Journ. Nat. Hist., vol. 6, 1857, p. 529.

Solaster Stimpsoni Verrm1, in Appendix ©. Report of Progress; Geol. Surv. of Canada for

1878-79, 1880, p. 192 b. (Ramsay Id., British Columbia).—J. F. Warreaves, Trans. Roy.
Soc. Canada, vol. 4, sec. 4, 1886 (1887), p. 116. (Port Neville, British Columbia, Johnstone
Strait.)—Verrm, Amer. Nat., vol. 43, 1909, p. 548.

Crossaster Vancouverensis DE Lorton, Mém. soc. phys. hist. nat., Genéve, vol. 32, pt. 2, No. 9,

1897, p. 12, pl. 1 (16), fig. 5 (Saanich Inlet, Vancouver Island).
Solaster endeca decemradiata Cuark, Proc. Bost. Soc. Nat. Hist., vol. 39, No. 15, 1901, p. 325
(Puget Sound).

Diagnosis.—Apparently most nearly related to S. endeca but differing in having
typically a smaller disk, longer rays, larger and less crowded abactinal paxille (which,
however, have comparatively few spinelets), and smaller inferomarginal pseudopaxil-
le, which have fewer spinelets. Rays ten, occasionally nine. R=82mm.; r=20mm.;
R=4.1 r. A shorter rayed variety: R=3.17 r.

Description.—In a typical form this species, or perhaps only subspecies, occurs
from Washington to Kadiak. At Kadiak there is some evidence that stimpsoni
and endeca run together, for it is difficult to assign certain specimens to their proper
species. As the two forms range together here they may possibly hybridize.

The photographic figures will reveal the tangible differences nearly as well
as a detailed description. The description will be in the nature of a comparison
of stimpsoni with endeca.

Typically the rays are longer and somewhat slenderer than in Pacifie endeca.
The pseudopaxille are perhaps two or three times as large, having coarser and
blunter spinelets, six to twelve to a group on the proximal radial regions. These
spinelets are sometimes sheathed to their tips in a continuous web and sometimes
radiate apart forming a floriform group of six to ten peripheral and one or two
central spinelets. These paxille are not always relatively more widely spaced
than in endeca, but on account of their smaller number and larger size appear to
be. In certain extreme variants—those from the southern part of the range—
the paxille are actually more widely spaced. The most obvious difference how-
ever is in the size of the paxille and component spinelets.

@ Verrill (1909a, p. 60, figs. 3 and 4) has described Solaster constellatus, ‘‘an eight-rayed species
with small disk and long, tapered arms.’’ R=3.7r. It resembles S. stimpsoni, but the pseudopaxille
are larger, though with only four or five to eight spinelets. The inferomarginals are small, with eight to
twelve spinelets, while the superomarginals are subequal to the abactinals. The adambulacral spines
consist of a furrow series, with two or sometimes three rather short, tapered spines and an outer comb
of six or seven nearly equal tapered spines, webbed nearly to the tips; the inner ones are usually rather
longer, so that the rows are a little graded. Professor Verrill kindly sent me abactinal and actinal
photographic views of this form. In view of the variation in the abactinal paxille of S. stimpsoni, and
the great similarity, almost identity in marginal and adambulacral armature of stimpsoni and constel-
latus, the main distinction remaining is the difference in the number of rays. I am strongly inclined
to regard S. constellatus an eight-rayed variety of stimpsont. S. pavillatus has an eight-rayed form.

312 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

The inferomarginal pseudopaxille are not so wide as in endeca and bear fewer
(twelve to eighteen) coarser spinelets. The plates are only slightly wider than
long. Superomarginals slightly larger than adjacent abactinal paxille and about
one-third to one-half as large as the adjacent inferomarginals. There is variation
in the size of the superomarginals. They are about as large as the midradial paxille.

Actinal interradial areas typically smaller than in endeca, and containing
fewer paxille similar to those of abactinal surface, but with fewer (four to seven)
spinelets. A single series of paxillae extends a very short distance along free part
of ray (perhaps one-fifth to one-fourth length).

Furrow spinelets two or three (only one near the end of ray), short, of the type
of S. endeca. Actinal comb curved aborad at inner end, and containing six to
eight blunt rough-tipped spinules decreasing in length from the furrow. These
combs extend little if any above the level of the marginal plates to which they are
adjacent throughout most of the ray. The adambulacral armature does not present
any very striking differences. One difference, however, is apparent though perhaps
not constant. In endeca the actinal comb is shghtly curved throughout its whole
length, but in stimpsoni it is straight on the outer part and abruptly curved at the
inner end. In this feature stimpsoni is intermediate between endeca and dawsoni.

Mouth plates similar to those of endeca; marginal spinelets six to eight, the
inner two or three enlarged as in endeca but in extreme specimens shorter and
stouter. Suborals four to eight, in a single series (usually double in endeca) shorter
and stouter as a rule than in endeca, but not invariably so.

Color in life: A specimen from Friday Harbor, Washington, has the center of
disk and a broad band proceeding thence along the middorsal region of each ray
leaden blue, shading to yellow ochre on interradii and rest of rays, except inner
half of each subambulacral comb, and mouth plates, which are like dorsum, bluish.

Cotype.—In the Yale University Museum.

Type-locality—Ramsay Island, British Columbia.

Dstribution.—Southern Bering Sea (Commander Islands), and Kadiak, south
to Oregon, in shallow water.

Specimens examined.—Twenty-one, from the following localities:

Specimens of Solaster stimpsoni examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
2874...... Off Cape Flattery, Washington .. 27 rocks and shells............. 1 | U. S. Nat. Mus.
3096...... Off Orford Reef, Oregon .......-- dts) & Aleta:
Kadlak;Alaskae 2. sceca--scs~ ace 9 | U.S. Nat. Mus., W.J. Fisher.
Bering Island, Commander Is- 3 | U.S. Nat. Mus., L. Stejneger.

lands.
Belkoteki, Alaskaess.<'sceus ose n 2 | soe eeeeee | eee ee eee ee 1 | U.S. Nat. Mus., W. H. Dall.
4205...... Admiralty Inlet, vicinity of Port | 26-15 | rocks, shells............-..-- 1 Albatross, 1903.
Townsend, Washington. |
| Do.

4200 eset} sk. GO se cotce as secncseoenena 25 | rocky, coarse sand, shells... . 1
Karluk, Kadiak, Alaska ......... Low tide.| scccucsncstuneasceccsussteoees 2 | Stanford collection.
Port Renfrew, Vancouver Island, | Lowtide.|..............--..--.--..----- 1 Stanford collection, J. C.
British Columbia. Brown.
Friday Harbor, San Juan Islands, | Lowtide.|.............0--seceeecceereee 1 Stanford collection, F. W.

Washington. | Weymouth.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 313

Remarks.—As the foregoing description is largely a comparison of stimpsoni
and endeca it will not be necessary to state again the differences which separate
the two forms. It is not possible to determine satisfactorily the relationship of
the two species. Although the extremes are certainly very distinct, at Kadiak
Island they either hybridize or approach each other more closely than elsewhere,
judging by nine dried specimens in the collection. The Bering Island specimens
are not typical stimpsoni as the rays are too short and the marginal plates larger
than in typical specimens. The abactinal paxille resemble those of stimpsoni.
These specimens may really be only a variety of endeca with unusually large paxille.

SOLASTER DAWSON I Verrill.

Pl. 84, figs. 1, 2; pl. 85, figs. 1, 2; pl. 86, figs. 1, 2; pl. 113, fig. 1.
Solaster dawsoni VERRILL, in Appendix ©, Report of Progress, Geol. Surv. of Canada for 1878-79,
1880, p. 193 6.—J. F. Wurreaves, Trans. Roy. Soc. Canada, vol. 4, sec. 4, 1886 (1887), p.
116 (Powell Island, Straits of Georgia, Melaspina Inlet, Johnstone Strait, Goletas Channel).
Solaster endeca Murvocn, Report of the International Polar Expedition to Point Barrow, Alaska,
1885, p. 159.

Diagnosis.—Rays eight to thirteen (commonly eleven or twelve) stout; disk
moderately large. A large specimen: R=163 mm.; r=48 mm.; R=3.4 r. R
varies to 2.5r. Differing from S. endeca in having larger, more spaced, and tabulate
pseudopaxille, larger marginal plates with more numerous spinelets, smaller actinal
interradial areas, and especially in having much longer furrow spinules and longer
bristling actinal adambulacral spines in a straight transverse series.

Description—Rays variable in number, but usually eleven or twelve, less
often thirteen, ten, nine, or eight. The width varies with the number, being greater
when the number is small. Typical specimens have the pseudopaxille much larger
and more distantly spaced than in endeca, tabulate in structure, and resembling
superficially in alcoholic specimens the parapaxille of Mediaster. They are ellip-
tical or roundish in contour, largest on disk and proximal portion of rays, decreasing
in size toward the end and sides of rays, being arranged in regular quincunx on the
lateral portions of the rays. In endeca the pseudopaxille decrease in size toward
the center of disk, but in dawsoni they are subequal, or slightly larger. The paxille
vary considerably in size in different localities and the number of spinelets and
their length is also variable. An example from Monterey Bay has on the larger
paxille of disk, surmounting the low tabulate eminence of the plate, sixteen to
thirty peripheral and five to fifteen central, short, rough-tipped stubby spinelets so
immersed in membrane that only the tips protrude and give a granuliform appear-
ance, the summit of the group being usually flat. The spinelets themselves are
longer than the height of the tabulum, and usually but not always shorter than
width of large paxille. A variety from deeper water (stations 3459, 3466) has the
paxille more convex and the spinelets more diverging, and less compact. The
spinelets are sometimes much less numerous, ten to twenty to a paxilla. Papule
numerous in the meshes of the skeleton, five to ten or even more on disk and usually
one to five on rays except distally where there are only one or two. Very large
specimens may have as many as twenty papule to an area near center of disk.

814 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Marginal plates numerous, separated by their length or less; larger than in
endeca, the inferomarginals being slightly higher and with fifty or more spinelets,
the inferior of which are much the longest, and exceeding the height of the pedicel,
the others graduated in length toward the upper end of the plate, being there about
the same height as on the paxille. In shallow water specimens the pedicel is no
higher than in endeca, although certain examples from deeper water (as for instance,
station 3466) have higher marginals and are not typical in respect to the abactinal
paxillee which are convex, not flat-topped. Typically the superomarginals are larger
than in endeca and are larger than the immediately adjacent abactinal paxille,
having about twenty peripheral and ten to twelve central spinelets, but the number
is variable, like every other character. Sometimes the superomarginals are directly
above the corresponding inferomarginals, but sometimes they alternate, both arrange-
ments being found on the same ray.

Adambulacral spines long and bristling, the furrow series especially being
much longer than those of endeca. Furrow spines stout, though slender, tapering,
bluntly pointed, skin-covered, united basally by a web, three or four in number
(on the first adambulacral, five or six). Actinal series straight (occasionally very
slightly curved) containing five or six subequal, tapering, stout, bluntly pointed,
or chisel-tipped spines, the innermost equal to or a little longer than the furrow
spines, the others successively slightly shorter. The innermost spine does not stand
out of line and aborad to the rest, nor is the series conspicuously curved as in endeca,
except sometimes at base of ray. The straight actinal series will serve to distin-
guish this from other closely related forms. These actinal spines, when erect,
extend conspicuously beyond the level of the marginal paxille.

Mouth plates broader than in endeca, even when the rays of the latter are less
numerous. Marginal spines nine to eleven, the outer ones conspicuously longer
than in endeca, being subequal to the adambulacral furrow spines (these being
short in endeca). The inner four or five are still larger, the innermost equaling
or exceeding the interradial dimension of the plates. The actinal surface bears a
single or double row or group of five to ten prominent spines, the inner longest; or
only one or two spines. There seems to be great variability in this character.

Actinal interradial areas very small in typical shallow water examples and
containing about twenty to twenty-five plates bearing a circle or group of about
six to twelve webbed spinelets much longer than those of dorsal paxille and about
equaling or slightly exceeding the longest inferomarginal spinelets. The variety
from stations 3466 and 3459, having ten and eleven rays, has decidedly larger than
typical interradial areas, as has also an eight-rayed specimen from station 4247.
In these the number of plates to an area is upward of sixty-five. As scarcely two
specimens in the collection are alike and the variations do not seem to be related
to locality, no varieties have been named.

Madreporic body conspicuous, midway to margin (or nearer center of disk)
and bearing several paxille near or on its margin; strix irregular.

Color in life, reddish brown above and yellow beneath (J. C. Brown, from a
specimen taken at Postelsia Point, Renfrew District, Vancouver Island, British
Columbia). A specimen from Friday Harbor, Washington, in formalin, is grayish

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 315

brown above, shading to lighter grayish buff on sides and dull yellowish brown on
actinal surface. Mr. Weymouth states that the life colors are similar, but brighter.
Cotype.—In the Yale University Museum.
Type-locality— Virago Sound, British Columbia, 8 to 15 fathoms.
Distribution Monterey Bay, California, to the Aleutian Islands, thence to
Kuril Islands; through Bering Strait to Point Franklin.
Specimens examined.— Twenty-eight, from the following localities:

Specimens of Solaster dawsoni examined.

— he

Station. Locality. Depth. Nature of bottom. Number. | Collection.
Fathoms.
34 TOURS Sn eetoinnn Jee uk ade eee 1 U.S. Nat. Mus.
a4 RTO Y AAI een nine naa dekowe 1 Do.
38 gray sand, gravel... 1 Do.
61 black sand, mud 1 Do.
1233 ROA BONO cco tenaceacewaseea 1 Do.
92 dark sand, rocky... 1 Do.
56 gray sand, shells... 4 Do.
Atralts of i uca, Washingtons 3 .}..5- .cencca|avuwcccscemcesnsneeae oe 1 U.S. Nat. Mus., D.S. Jordan.
Cape Mendocino, Californla......|........... [es Ao Se GA 1 U.S. Nat. Mus., Lieut. Com-
| | mander H. E. Nichols.
Sanborn Harbor, Shumagins, | Lowest |.........----------+-sceseeeee 1 | U.S. Nat. Mos.,W.G. Hall
Alaska. water. | | (W. HL. Dall).
Point Franklin (10 miles west), 18be (| sands)! .23.5-4. eemeede ete 2 U.S. Nat. Mus., Point Bar-
Alaska. row Expedition.
42474.....| Kasaan Bay, Prince of Wales | 89-114 green mud, fine sand........ | 1 Albatross, 1903.
Island, southeast Alaska. i
4777».....| Near Semf{sopochnoi Island, 5243 ONG Pavel ene en daacearaenes 1 Albatross, 1906.
Aleutians.
RT18 Oa ceae| eae OU candies ne cena anes ann <n 43-33 fine black gravel............ 3 Do.
48045..... Simushir, Kuril Islands . 4 229 coarse pebbles, black sand... 3 Do.
Monterey Bay, Callfornia. Offshores.| .. ae 2 Stanford University.
Renfhew District, Vancouver | Low tide. 1 Stanford University, J. C.
Island, British Columbia. Brown.
Friday Harbor, San Juan Islands, ...do .....|........--------++.-+- aoheeee | 2 Stanford University, F. W
Washington. | Weymouth.
!

a Eight-rayed, large actinal interradial areas.
> Variety, some with eight rays, resembling somewhat S. parillatus; possibly hybrids.
¢ Brought up on fishermen’s lines; less than 70 fathoms.

Remarks.—This species may be distinguished from both endeca and stimpsoni
by the long furrow spines, which are nearly as long as the actinal spines, by the
straight or nearly straight comb of the latter, and by the larger, usually flat-topped
pseudopaxille.

SOLASTER PAXILLATUS Sladen.
Pl. 87, figs. 1, 2; pl. 88; pl. 89, figs. 1, 2; pl. 113, fig. 3.
Solaster paxillatus SuavEeN, Challenger Asteroidea, 1889, p. 452, pl. 71, figs. 1-3; pl. 72, figs. 1, 2.

Diagnosis.—Rays eight to ten. R=177 mm.; r=58 mm.; R=3+r; varying
to 3.7 r. Disk moderate to large; rays, stout; no marked interradial streak free
from paxille as in borealis, but a slight indication in some specimens; paxille
fairly close-set, subtabulate with numerous small spinelets heavily invested with

316 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

membrane ; paxillz not so widely spaced as in borealis, nearly as in dawsoni; on the
rays and peripheral portion of disk about their own diameter apart, more crowded
on center of disk; marginal plates in two series, the superomarginals about the size of
larger paxille, and standing above and between the prominent fan-shaped infero-
marginal paxille; latter with constricted base and semicircular summit bearing
numerous spines decreasing in size toward upper end; adambulacral furrow spine-
lets long, four or three, then three, two, and finally one at end of ray; actinal series
curved aborad at inner end and with proximally five to eight, distally three to five
longer slender spinelets; mouth plates with eight to eleven marginal and about
fifteen suboral spines.

Description.—Abactinal paxille very numerous, low, tabulate (similar to those
of dawsoni), with a roundish crown of numerous very short denticulate spinelets
immersed in membrane. On the disk the paxille are not very uniform in size, and
larger toward center, where they are spaced less than their own width. On the
peripheral half of the disk and on rays the paxille are not so close together, but are
spaced about their own width apart or slightly more. Here they are arranged in
very oblique transverse rows, and also in fairly even cross rows, but a longi-
tudinal arrangement is not so evident (so-called quincunx order). The larger
paxille of disk have thirty to forty thick, fleshy spinelets forming a slightly convex
crown, the peripheral series of ten to fifteen being the largest, the rest decreasing in
size toward center. On the proximal part of the ray the paxille have about the
same number of spinelets, but the whole paxilla is smaller, and they decrease regu-
larly in size toward tip of ray and to a less extent toward the margin. The spine-
lets are less heavily enveloped on the ray. The spinelets themselves are slender,
even on the disk, with numerous denticles at the tip. An undried paxilla is about
as high as the width of the crown of spinelets; when dried, slightly higher.

Abactinal plates along lateral area of ray regularly four-lobed, each plate
imbricating with four others by the length of a lobe, the regular meshes thus formed
containing one, two, or near median area, three papule. Along midradial line and
center of disk (out about to madreporic plate) the plates are very irregular, having
three, four, or five lobes, and the primary plates on disk are irregularly connected
by oblong or irregular ossicles forming very irregular and unequal meshes which
contain two to five papule. Along midradial region there are two or three to a
mesh.

Marginal plates in two series, and of very unequal size, the superomarginals
alternating with inferomarginals. Superomarginals considerably larger than
adjacent abactinal paxille, and about as large as largest paxille of corresponding
midradial region. Each plate stands opposite the interval between two infero-
marginals, and on a level with the upper edge of the latter (not crowded between).
Inferomarginals (about sixty to a ray) large, paxilliform, with a much compressal
fan-shaped pedicel with a curved outer border or summit bearing thirty to forty
spinelets, which decrease very rapidly in length from the lower ones, which are in
two series and about as long as half the width of summit of pedicel, to the upper
which are in three or four series and the same size as spinelets of abactinal and
superomarginal paxille. Inferomarginals spaced about their own height apart

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 317

(but in a variety are closer) and usually are not visible from above, on inner half of
ray. They do not extend inward from the margin any great distance on the inter-
brachial ares, being there smaller and closer together. They differ from the mar-
ginals of borealis in having much more numerous and shorter spinelets, in haying
a relatively higher pedicel, with a more evenly semicircular edge and much more
contracted base, in not forming a conspicuous marginal fringe as in borealis, and in
being more numerous and closer together. They differ from those of dawsoni in
being higher, with a more convex or semicircular edge, and in being more widely
spaced.

Adambulacral plates of the same general type as in dawsoni. Yurrow spines long,
at base of ray three or four, then three (the adoral becoming shorter and shorter),
then two, and finally, near the end of ray, only one. The spines are slender and
webbed at the base, the outer part having a flange of tissue, and the middle spine is
longest, about as long as the distance between the inner end of two adjacent actinal
series. Subambulacral spines (proximally five to eight, distally five, four, or three)
longer, tapering, pointed, similarly webbed and membrane-sheathed, in a transverse
series the inner end of which bends aborally so that the inner spine is opposite the
aboral furrow spine while the outermost is at the adoral edge of plate. These
spines are slenderer than in borealis, stand on a ridge of the plate and are subequal in
length, or the outermost shortest, and in length about equal the base line of the
series, or the distance between the inner end of three adjacent actinal series or
slightly more than the combined length of two adambulacral plates. The maximum
number of spines in the largest specimen is six; in a smaller example, eight. In
dawsoni the actinal comb is straight, not curved, as in this species.

Mouth plates similar in form to those of borealis, with eight to eleven marginal,
and upward of fifteen suboral spines on each. The former increase very markedly
in length at inner angle, the inner three or four being long, the innermost as long as
the interradial diameter of combined plates. The suborals usually form a crowded
series with those of the companion plate, the innermost being nearly as long as the
teeth, the outer rapidly shortening or formimg two or three series parallel to suture.

Actinal interradial areas larger than in borealis and much larger than in dawsoni, .
with spaced low pseudopaxille in chevrons, bearing a variable number, but usually
numerous (as few as five, as many as fifteen) slender spinelets united at the base by
membrane. A single series of small intermediate plates bearing one to few spinelets
extends over half the length of ray.

Madreporic body halfway between margin and center of disk, surrounded and
partly obscured by five to eight large paxille.

Variations.—The specimens from stations 4792, 4784, 2853, and 3223 constitute
a well-marked variety with eight rays. The number is usually nine (station 3331),
less often ten; specimens from 3500 are either eight-, nine-, or ten-rayed; the eight-
raved being similar to those from 2853 and 4784, the others resembling examples
from 3331. The eight-rayed variety (pl. 89), as compared to the form from station
3331 (pl. 87), has broader rays, more compactly placed paxille, with more numer-
ous spinelets and larger superomarginal paxille, which usually stand above the
inferomarginals rather than opposite the interval between. The inferomarginals

‘

318 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

in extreme examples are closer together, and both series have more spinelets,
which may or may not be shorter than in specimens from 3331. In the variety,
also, the actinal spinelets are more heavily invested with membrane and the
actinal adambulacral spines are more numerous (eight or nine proximally). The
suboral spinelets form a dense cluster on each plate. In extreme cases the actinal
interradial area is smaller (pl. 89, fig. 2), with relatively larger and fewer paxille. The
general appearance of the actinal surface is that of more dense spinulation and ~
more compact marginals. The specimen from 2853 (pl. 89, fig. 1), however, although
typical of this variety abactinally, has an actinal surface similar to specimens from
3331, and thus combines characters of each form. The specimens from 3500 also
belong to both forms. The superomarginal paxille present gradations between the
two extremes.

The specimen from 4784 is an extreme example of the variety with unusually
large superomarginals. It resembles aberrant specimens of S. dawsoni, but is
probably referable to pawillatus as the actinal combs of adambulacral spines are
curved, not straight, as in dawsoni.

It is certainly very tempting to consider this variety a distinct species, but the
evidence of intergradation can not be ignored. There is a possibility that the
variety is really a northern form, which is hybridizing with true pazillatus, derived
from the south.

Type.—British Museum.

Ty pe-locality.—South of Yokohama, Japan, Challenger station 232, 345 fathoms,
green mud.

Distribution.—From south of Yokohama, Japan, to Bering Sea, thence to
vicinity of Kadiak Island, Alaska, 56 to 350 fathoms; the typical form in deeper
water.

Specimens examined.—Twenty-two, from the following localities:

Specimens of Solaster paxillatus examined.

Station. | Locality. Depth. | Nature of bottom. ; Number. Collection.
Fathoms.
South of Kadiak Island, Alaska. ..... 159 Gray Sande eee edaen te secrete 1 U.S. Nat. Mus.
Off Unimak Island, Alaska ... of 56 black pebbles 2 Do.
North of Unalaska.........-..-....-- 350 MUGS Cee eeeeey 6 Do.
South of St. George Island, Bering Sea. 121 fine gray sand 7 Do.
3608...... Bering Sea, between St. George and 276 gray SANG nec. ns te ee eee 3 Do.
Unalaska.
4784...... Near Attu Island, Alaska ...........- 135 coarse pebbles 2 Albatross, 1906.
4792...... | Off Commander Islands. ..........--- 72 pebbles. 22 2-=22-:scnpceteemente 1 Do.

Remarks.—This species differs from dawsoni in. having more prominent and
typically more widely separated inferomarginals, and in having the superomarginals
alternating with the inferomarginals. The actinal interradial areas are much
larger than in dawsoni, and the actinal comb of adambulacral spines is curved at
the inner end, while in dawsoni it is nearly or quite straight.

From S. borealis, pazillatus differs in having much larger abactinal paxille,
lacking the bare interradial streak, in having more numerous inferomarginals, with

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 319

much shorter spinelets, in having distinguishable, even prominent superomarginals,
and fairly long furrow spinelets.

A specimen from station 5050, off Honshu, Japan, while differing from Alaskan
specimens in having rather smaller paxille, seems to be referable to the form from
station 3331. At the same time it does not appear to be separable from Sladen’s
S. pawillatus. The Alaskan examples have therefore been ranked under pazillatus
which, if the identification is correct, has a fairly wide range and is extremely
variable.

The stomach of one specimen contained a holothurian, Molpadia intermedia.

SOLASTER EXIGUUS Fisher.
Pl. 90, figs. 1, 2; pl. 112, figs. 4, 4a—b.
Solaster exiguus Fisner, Zool. Anz., vol. 35, March 29, 1910, p. 573.

Diagnosis.—Similar in general appearance to S. pazillatus, but disk smaller,
paxille smaller, inferomarginal plates more widely spaced and much narrower with
fewer and longer spinelets, actinal interradial areas smaller (despite the fewer rays);
furrow spinelets two or one, long; actinal adambulacral spinelets three to five.
Rays seven; R=25 mm.; r=8 mm.; R=3 r.

Description —Rays moderately slender, the inferomarginals not visible from
above except on outer part of the ray. Paxille rather small (but not so small as
in endeca or borealis), and spaced farther than their width. They have the appear-
ance of being fairly well spaced, although the papular areas contain ordinarily but
one papula. Paxille arranged in quincunx on the rays, definite longitudinal and
oblique transverse rows being discernible. Each paxilla consists of a cruciform
base raised into a low tabulum surmounted by five to eight or even ten slender
spinelets considerably longer than height of tabulum and with denticulate blunt
tips. When the spinelets stand erect the paxillae appear small and well spaced, but
when the spinelets are spread out the paxille nearly touch one another and look
much larger. When spread out the spinelets are seen to be webbed for half their
length.

Inferomarginal paxille similar in general appearance to those of borealis, well
spaced, two to every three adambulacral plates, hence spaced farther than the
width of a paxilla. Each inferomarginal is subcircular, only very slightly com-
pressed and is crowned by five to twelve delicate spinelets, longer than the pedicel
and denticulate at tip. Above the interval between two inferomarginals is a small
superomarginal similar to but slightly larger than adjacent abactinal paxille.
This is less conspicuous than in pazillatus.

Adambulacral armature prominent; plates well spaced as in borealis. Furrow
spinelets unusually long, nearly or quite as long as the actinals; proximally two, and
distally 1. Actinal comb slightly curved aboral at inner end, and consisting of
three to five slender long spinelets webbed basally, the middle spinelets longest.

Actinal interradial areas smaller than in pavillatus and endeca, about as in
borealis, and containing fifteen to twenty paxilliform groups of five or six spinelets.

Mouth plates rather small, with six to eight marginal and two to five suboral
spinelets, the former webbed basally and the inner two or three enlarged into teeth.

320 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Madreporic body small, situated at about the middle of r.

Type.—Cat. No. 27786, U.S.N.M.

Type-locality.— Albatross station 2980, off Anacapa Island, California, 603
fathoms, green mud, two specimens. Known only from this locality.

Remarks.—Although this species was dredged along with S. borealis and Heterozo-
nias alternatus it is clearly not referable to either. It is probable that the number of
rays varies from the small number of seven in the types. It is at once distinguished
from H. alternatus in lacking the characteristic superomarginals of that species, and
it differs from borealis in having larger paxillze with more numerous longer spinelets
(the paxille being consequently higher), more numerous and delicate inferomarginal
spines, more prominent, alternating superomarginals, and two long furrow spine-
lets (not five or six short ones). It may be separated from paaillatus by the much
smaller actinal interradial areas, differently formed, and more widely separated
marginals which haye longer spinelets, by the few suboral spinules and typically
smaller paxille with fewer and longer spinelets. It is not possible to determine
whether intergrades exist between exiguus and pazillatus.

SOLASTER BOREALIS (Fisher).
Pl. 91, figs. 1-3; pl. 92, figs. 1, 2; pl. 113, figs. 2, 2a.
Crossaster borealis Fisher, Proc. Wash. Acad. Sci., vol. 8, Aug. 14, 1906, p. 134.

Diagnosis.—Rays nine to twelve, very rarely thirteen. R=140 mm,; r=47
mm.; R=3 r. Breadth of ray at base, 23 mm. Closely resembling Solaster
australis (Perrier). Marginal plates in a single series, very prominent, spaced,
paxilliform, with prominent bristling spines, and situated on the margin of ray,
becoming actinal in position only in interbrachial angle and then at the ambitus.
Abactinal pseudopaxille very small and spaced with very few spinelets. Papulz
large, when fully extended larger than paxille which they often obscure in big
specimens, one to three to a mesh on ray, three to ten on disk. Madreporie body
exposed. A bare interradial streak extending from the interbrachial angle half way
to center of disk. Adambulacral plates with five or six furrow spinelets (varying
to three or four distally and as many as eight proximally), and two, three, four, or
five very much longer and stouter actinal spines, in a transverse series. Mouth
plates with about eleven furrow spinelets and only one or two small suboral
spinelets, sometimes none; inner oral spines much lengthened. Actinal interradial
areas small, the spaced paxille with few spinelets; a single row of actinal imter-
mediate plates extending far along ray.

Description —Abactinal integument entirely obscuring underlying skeleton,
unless dried or treated with caustic potash. Paxille small, spaced, with a low
tabulum surmounted by one to six slender blunt or pointed, tapering spinelets. In
life these spinelets are thick, short, and stubby, owing to a membranous investment,
and are usually three or four to each paxilla. In center of disk and along distal
half of ray, paxille irregularly arranged, but between these two areas an arrange-
ment in longitudinal rows more or less evident. Base of paxille with three or four
slender unequal lobes impinging upon those of neighboring paxille or connected by
short irregular ossicles; latter not numerous; near center of disk there are one or

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 821

two isolated ossicles in many of the meshes. Anus surrounded by four or five large
paxille. As in preceding species a very narrow bare sulcus extends from inter-
radial angle about half way to center of disk. Papule prominent, but usually not
quite so large as in preceding species, about three to ten to a mesh on disk, one to
three in distal half of ray where skeleton is closer.

Marginal plates very prominent, spaced, in a single series, about twenty to
thirty on each side of a ray and confined to the margin. Superomarginals not
different from abactinal paxille and indistinguishable from them as a rule. Infero-
marginals with fairly high pedicels (relatively about as in papposus), bearing two
vertical or transverse palmate series of six to nine stout tapering pointed skin-
covered spines, the mesial of which are the longest. Sometimes there is one main
series, and two or three smaller spines stand adorally out of the series, or there may
be a second adoral series of less conspicuous spinules, but few in number. Spines
of proximal plates shorter than rest, except near tip of ray. The spines vary in
length but are about as long or a third longer than the extreme width of the top of
the pedicel and often twice as long as its height. The distance between the top of
two adjacent marginal plates is slightly less to one-third greater than the extreme
height of pedicel and longest spine. The prominent bristling spaced marginal
paxillz form one of the characteristic features of this species.

Actinal interradial areas rather small, about thirty-five to forty plates to
each area. Plates obscured by integument which has fine furrows or wrinkles
leading from interadambulacral sulcuses to marginal plates. Plates appear spaced,
each bearing one to four short stubby papilliform spinelets, very delicate when
dried. Plates arranged irregularly in rows, between the wrinkles. A series of
very small widely spaced actinal intermediate plates extends over three-fourths
length of ray. They bear usually one or two stumpy spinelets, or are spineless.

Adambulacral plates with (1) a palmate furrow series of five or six (distally
three or four) slender tapering skin-covered spinelets (united for about half their
length by a web) of which the second or third mesial are subequal, the laterals
much shorter. These spinelets are of about same length as in Heterozonias alter-
natus and about as long as the base line of comb. On the proximal part of ray
in some specimens there are seven or eight spinelets to a furrow comb. (2) On
actinal surface is a transverse series of four (three on smaller examples, varying
to two and five) much longer, slender, terete, blunt, skin-covered spines, the second
or third usually longest (exceeding in length the width of plate), the outer about
one-half length of inner (where there are three spines); when two spines only are
present they are subequal and long. The longest spines are about twice as long
as the longest furrow spinelet.

Mouth plates just a trifle narrower than in, H. alternatus. Free margin with
a webbed series of about eleven spinelets increasing in length toward inner end of
each plate to two or three much enlarged spines, the innermost stoutest. On
actinal surface of plate near inner end of each is a stout, though slender, spine.
Sometimes instead of this a small one stands on outer end of plates, or there may
be two or three small spines, or the surface may be entirely unarmed. The furrow
spinelets are usually eleven, but vary to nine or twelve even in otherwise typical

57444°—Bull. 76—11——21

822 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

“ae

examples. The mouth plates are decidedly narrower in specimens with twelve
rays than in those with nine (station 3607).

Madreporic body variable in size, exposed and situated at inner end of an
interradial sulcus. Two or three large paxille stand near it.

Color in life, of a Californian specimen: Abactinal surface, dull orange or saturn
red; dull light yellow below.

Variations.—All the specimens from station 3607 have nine rays, which are
broader than in the type. There seem to be no important constant differences
in the ornamentation however. Specimens with ten and eleven rays are present
in the collection but nine and twelve are the prevalent numbers. Size has appar-
ently nothing to do with the number of rays. Variations in the number of furrow
spines and subambulacral spines have been alluded to in the description. The
increased number of furrow spines can not be correlated with other constant differ-
ences. Certain specimens, especially those from station 3607, have curious slits in the
web at the base of the furrow spines of mouth and adambulacral plates. Examples
with fewer rays have the paxille more widely spaced than those with eleven or
twelve rays. Owing to the broader mouth angles in examples with nine and
ten rays, the mouth plates are broader than in specimens with twelve rays.

Specimens from California to Queen Charlotte Islands have a different appear-
ance from the more northern examples and perhaps deserve to be separated as a
race of typical borealis. The abactinal paxille are lower, both the spinelets and
pedicels being shorter, and the spinelets blunter, the rays are often slender (ten
to twelve in number), the marginal paxille have rather fewer spines and the
subambulacral spines are only two. or three and less heavily covered with
membrane. Oral spines eleven to thirteen, the suborals are one or two, or
absent. The differences are most pronounced in the specimen from station 4415
between Santa Catalina and Santa Barbara Islands, 638 fathoms.

I have examined a twelve-rayed specimen from station 5050 off the east coast
of Honshu, Japan (lat. 38° 11’ 30’’), 266 fathoms. The rays are shorter than
in typical examples (R=52 mm., r=22 mm., R=2.36 r). A nearly equal-sized
specimen from Monterey Bay has R=2.7 r. There is a thirteen-rayed specimen
from station 3489, Bering Sea, which has R=2.1 r. The length of ray is therefore
variable even in typical examples. The most important difference is the small
number of three furrow spinelets throughout most of the furrow. There are nine
oral and one to three suboral spines. The paxille are fairly typical with two
to four or five tapering spinelets. The bare interradial streaks or lines are well
marked.

Type——Cat. No. 21933, U.S.N.M.

Type-locality.— Albatross station 2858, east of Kadiak Island, Alaska, in 230
fathoms, on blue mud and gravel.

Distribution.—F rom off San Diego, California, to Bering Sea, thence to Honshu,
Japan, 225 to 1,044 fathoms.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 323

Specimens examined.—One hundred and seven, from the following stations:

Specimens of Solaster borealis examined.

ot — := a. as
Station. Locality. Depth. | Nature of bottom. Number.

Collection.

|
| U.S. Nat. Mus.

ane Off Sea Lion Rock, Washington.

2858...... | East of Kadiak Island, Alaska....... |

AGO. oe = South of Queen Charlotte Islands... .. | | Do.

SAT ar ot Off Gray’s Harbor, Washington. .... | Do.

DOTS ss Off San Diego, California........ Do.

2980...... Off Anacapa Island, California... Do.
’
’

aoe Monterey Bay, California ............
Gall nena Bering Sea north of Unalaska.........
3331...... North of Unalaska.............

eRe e nti e Rew eee Re oh ot ee

Nicholas Islands, California.

Do.

Do.

Do.

Do.

Do.

Do.

Do.

3387.2... South of Unimak, Alaska Do.

3338...... South of Shumagin Islands, Alaska... | Do.

3343...... Off Destruction Island, Washington. 516 | Do.

$8465 Off Tillamook Bay, Oregon.......... | 786 Do.

S887 cg-c0 Jewece 0, ask Sse oon e 345 Do.

S450 ou Bering Sea, west of Pribilof Islands... 184 Sel Do.

3601...... Bering Sea, south of Pribilof Islands.) 1,044 Do.

3007...... North of Unalaska..............--.--- 987 Do.

3608...... Bering Sea, between Pribilof Islands 276 | Do.

and Unalaska. |

3670...... Monterey Bay, California............. 581 | 6 Do.

4415...... Between Santa Catalina and Santa pay Albatross, 1904.
Barbara Islands, California.
4420...... Between Santa Barbara and San 238 | 1 Do.
Do.

ae
e
3
4
5
=
nm

Remarks.—The differences between borealis and hypothrissus will be found
below under that species.

The species to which borealis (and especially the Californian form) shows the
greatest resemblance is Perrier’s Crossaster australis from the Straits of Magellan.
The differences are evidently not great and it is not at all improbable that the
ranges of the two will some day be found to run together. The west coast of
South America is not as yet well worked in deep water. Crossaster neptuni Bell
may be a link in this chain. Solaster australis® differs in having a single papula®,
not several, in each mesh of the skeleton, especially on disk; in having a series of
suboral spines (as in hypothrissus); and in having less prominent marginal plates
with shorter spines. There seem to be differences in the paxille, judging by
Perrier’s figure, but it is perhaps not safe to make comparisons without specimens.
Australis has nine or ten rays which are about as slender as in Californian specimens
of borealis though longer (R=3.25 r).

Solaster regularis Sladen, collected by the Challenger in 175 fathoms, south of
Wellington Island, off Chile (lat. 50° 08’ 30’’ S.), is another closely related species,

“See Ludwig, Zeitschr. fir wiss. Zool., vol. 82, 1905, p. 65.
> Perrier, Miss. eci. Cap Horn, 1891, p. 115, pl. 10, fig. le.

324 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

and its relationship with australis is greatly in need of elucidation. If regularis is
only. a variant of australis, the remarks made above concerning the differences
between australis and borealis will have to be emended. Sladen refrains from
stating how many specimens he had (a defect of practically all his descriptions), so
that it is not possible to judge of the range of variation. Regularis has only eight
rays, which are unusually long (R=4.5 to 5 r), broad stumpy paxille with six to
ten spinelets, less markedly compressed marginal plates, with more numerous and
apparently shorter spines, less prominent actinal adambulacral spines (four or
five in number), and a row of five suboral spines.

SOLASTER HYPOTHRISSUS Fisher.
Pl. 93, figs. 1, 2; pl. 118, figs. 5, 5a.
Solaster hypothrissus Fisuer, Zool. Anz., vol. 35, March 29, 1910, p. 574.

Diagnosis.—Similar in general appearance and structure to S. borealis, but
differing in having a row of five to seven prominent suboral spines parallel to the
median suture, in having a maximum of six long subambulacral spines, instead
of four or less, and in having the abactinal pseudopaxille slightly higher with seven
to nine spinelets which are blunter than in typical borealis and end in several points,
not in an attenuate denticulate tip. Rays ten. R=118 mm.; r=38 mm.;
R=3 r+.

Description.—It is not possible to determine whether this is a distinct species
or a ‘‘form”’ of borealis. A close examination reveals many points of intimate
resemblance to borealis, but no intermediate specimens have been found. The
chief features have been alluded to in the diagnosis. The papule are prominent,
as in borealis, but the paxille are stouter with more spinelets (seven to nine instead
of three or four). The spinelets are less attenuate distally and form a rosette
with one in the center. There are about thirty marginal plates not differing in
any important respects from those of borealis. The actinal intermediate plates
extend in a single series nearly to end of ray; distally they are very inconspicuous.
The adambulacral armature is very bristling, the actinal series containing proxi-
mally six spines, the outer the shortest, and distally five, then four. The furrow
series has only three to five spinelets, five proximally, which are webbed for about
half their length. They do not differ in important respects from those of borealis,
but are slightly longer. The mouth plates and their armature depart more widely
from those of borealis, having a bristling array of suboral spines five to seven in a
row parallel with and close to the suture. The innermost spine is the stoutest
and about as long as the longest actinal spine of the first few adambulacral plates.
The marginal spines are rather longer than in borealis, especially the outer ones.

Type.—Cat. No. 27787, U.S.N.M.

Type-locality.—Station 3338, south of the Shumagin Islands, Alaska, 625
fathoms, green mud, two specimens.

Distribution.—Vicinity of the Aleutian Islands.

Specimens examined.—The types and one specimen from station 4784, near
Attu Island, 135 fathoms, coarse pebbles.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 325

Remarks.—It is a matter upon which opinion may well differ whether such a
form as hypothrissus should be dignified with a name. The structure of the mouth
plates and paxille is different from that of typical borealis. The former are among
the least variable parts of this variable genus, and may be relied upon perhaps more
safely than the adambulacral armature. The armature of the mouth plates is
variable, to be sure, but the limits are less wide than in the case of the adambulacral
or marginal plates, and they change much less with age. The general facies of the
two species is quite unlike.

SOLASTER PAPPOSUS (Linnaeus).
Pl. 94, figs. 1-6.

Asterias papposa Linn £vs, Syst. Nat., 12th ed., 1767, p. 1098.—O. F. Miter, Zoologiw Danicw
Prodromus, 1776, p. 234.—Fasricius, Fauna Greenlandica, 1780, p. 369.

Asterias helianthemoides PENNANT, British Zoology, vol. 4, 1777, p. 56.

Asterias affinis Branpt, Prodromus, 1835, p. 71. (Bering Strait.)

Stellonia papposa AGassiz, Mém. soc. sci. nat. Neuchitel, vol. 1, 1835, p. 192. -

Solaster papposa Forses, Mem. Wernerian Soc., vol. 8, pt. 1, 1839, p. 121; British Starfishes,

1841, p. 112, fig.

Solaster (Polyaster) papposa Gray, Ann, and Mag. Nat. Hist., vol. 6, 1840, p. 183.

Crossaster papposus MitteR and Troscuet, Monatsber. preuss. Akad. Wiss. Berlin, Apr., 1840,
p. 103; Archiv Naturgesch., 6 Jahrg., vol. 1, 1840, p. 183.—Murpocu, Marine Invertebrates,
Report of the International Polar Expedition to Point Barrow, Alaska, 1885, p. 159.—Lup-
wiG, Fauna Arctica, vol. 1, 1900, p. 460 (Review of literature and synonymy to 1900).

Solaster papposus Miitter and Troscuet, System der Asteriden, 1842, pp. 26, 127; pl. 3, fig. 1 a-b;
pl. 12, figs. 3, 4.—Dusarprn and Hur®, Hist. nat. zoophytes, Echinodermes, 1862, p. 353.—
Perrier, Révision des Stellérides., 1875, p. 94.—Dépertern, Wiss. Meeresuntersuchung,
neue Folge, vol. 4, Abth. Helgoland, Heft 2, 1900, p. 205, pl. 6, figs. 14.—Osrereren, Zool.
Anz., vol. 27, 1904, p. 615.—GrieG, Bergens Museums Aarbog, 1906, No. 13, pp. 46-66, text
figs. 7 (4) fig. 8; pl. 1, figs. 6-8. .

Diagnosis..—Rays eight to fourteen, usually ten or eleven, very rarely less
than nine or more than thirteen; in the North Pacific, commonly ten or eleven.
R=1.8 to 2.3 r. Rays stout, tapering; abactinal skeleton forming an open reticu-
late structure with widely spaced penicillate paxille and numerous intervening
papule. Paxille with stout pedicels and upward to fifty slender spinelets gradu-
ated in length from periphery to center, the latter spinelets being typically as long
as or longer than pedicel (in some varieties shorter). Marginal paxille in a single
series, well spaced, prominent, similar to but usually larger than the abactinal.
Adambulacral plates with usually three or four, three to five, four or five, or rarely
four to seven furrow spinelets in a comb, and a transverse actinal comb of about
five to nine slightly longer and stouter spinelets. Interradial paxille none to about
twenty-five, varying in different interradii; mouth spinelets, about eight to ten
marginal and two to four suborals.

Notes on North Pacific specimens.—The variability of Solaster papposus has
been described and figured by several authors, especially Déderlein® and Grieg,*

a‘‘Similis A. rubenti, sed radiis 13; undique muricata aculeis apice pectinatis s. fasciculatis.”’
Linnzeus.

b Wiss. Meeresuntersuchung, neue Folge, vol. 4, Abth. Helgoland, Heft 2, 1900, p. 205, pl. 4, figs. 1-4.

¢Bergens Museums Aarbog, 1906, No. 13, pp. 46-66. For a discussion of variations in Solaster
papposus and S. sqguamatus, as well as for a review of the literature, this article is quite complete.

326 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

who have confined their attention to examples from the North Atlantic and adja-
cent portions of the Arctic Ocean. The North Pacific specimens are also variable,
so much so, in fact, that the extremes are very unlike. The species tends to form
small varieties differing with locality and also with depth, but there are instances
where the same variation (or nearly the same) is repeated in widely separated
places. The forms listed herewith as Solaster papposus may in reality be small
species, but it has not been possible to separate them on account of intergrades.
With the exception of the Japanese specimens, which are believed to constitute a
distinct species, all the North Pacific varieties have been called Solaster papposus.

Variety A.—One of the most striking varieties is that from Point Franklin,
Bering Island, Kadiak, and Sakhalin. (A complete list of stations given in record
of specimens.) In this the rays are ten, less often eleven, short, and robust. The
paxille are lower than in typical papposus, globose, the spinelets being much shorter.
In this form the central spinelets of a paxilla are relatively short, the whole paxilla
being about as high as broad, while in typical papposus the central spinelets alone
are longer than the width of a paxilla. The midradial paxille are larger than the
rest and form a fairly straight series. The abactinal skeleton is decidedly of the
reticulate form characteristic of papposus and not closely imbricated % as in S.
squamatus (Déderlein). The marginal paxille lkewise have much shorter spine-
lets and pedicels, showing about the same relative difference from the typical form
as do the abactinal paxille. Adambulacral plates with three or four moderately
long skin-covered and partly webbed spinelets, the adoral when there are four
being shorter than the rest. In the transverse subambulacral series are six or
seven tapering spinelets, equaling or slightly exceeding the furrow spinelet. The
inner end of the series curves aborad. Mouth plates with six or seven marginal
and three or four suboral spinelets. Actinal interradial areas with two to twelve
small tufts of short spinelets. In general appearance this variety departs farthest
from the type, especially on account of the low globose paxille. I have examined
a specimen from station 5024, Okhotsk Sea, off the east coast of Sakhalin, vicinity
of Cape Patience, which is referable to this variety. The largest specimen (station
4795) measures: R=66 mm.; r=35 mm.; R=1.86 r. In this the midradial row
of paxillee is not quite so prominent as in smaller examples, but the rays are thick
and the interradial areas unusually large. (See pl. 94, figs. 4, 5.)

Two specimens from station 3231, having eleven and twelve rays, are inter-
mediate between variety A and typical papposus. The abactinal surface is nearly
as in A, but the numerous rays, reduced actinal interradial areas, and longer adam-
bulacral spinelets are more like the typical form.

Variety B.—From stations 4777 and 4778 are four specimens which show a
combination of the characters of variety A and variety C, and are only slightly
different from typical papposus. The midradial row of paxille is well marked,
but all the paxille are higher than in variety A, though not so high or so slender as
inC. Rays ten or eleven.

Variety C.—This form differs from typical papposus in having the midradial
paxille longer and slenderer and the rays slenderer as a rule. Rays ten or eleven.

@ Described and figured by Grieg, Bergens Museums Aarbog, 1906, No. 13, pp. 46-66, text figs. 7, 8;
pl. 1, figs. 6-8.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 327

The midradial row of paxille is clearly distinguishable from the rest, that is those
between the midradial and marginal paxillw, being relatively smaller than in pap-
posus. The paxillar spinelets are long, longer relatively than in most specimens
of papposus, and the actinal spines are all ‘long. The marginal plates are very
prominent. Furrow spines four; superadambulacral spines five or six, as a rule.
In general appearance this variety has a more open abactinal skeleton than most
specimens of typical papposus, and the slenderness of the paxille gives a different
facies, which is perhaps better appreciated from a comparison of figures. Stations
4787, 4788, 4791, 4792, 2847, and others. (PI. 94, figs. 2, 3.)

Practically the same variations are present in specimens all the way south to
Oregon and Washington, as, for instance, from stations 2848, 2865, 2866, 2874,
2877, 3051, 3445, 3459, 3461, 4286. The southernmost examples (2874, 3051, 3459,
3461) have the variation very well marked. The abactinal skeleton is wide meshed
and the paxille tall and slender, there being fewer than in typical papposus. The
midradial paxille are sometimes, though not always, as large as the marginals, and
are distinctly larger than the dorsolaterals. The actinal spines are long and slender,
the adambulacral plates having three or four furrow spines, and four to six or
seven in the actinal series. Mouth plates with nine to twelve marginal and two
to four suboral spines. If this variety is compared with A, the contrast is of
course very striking. In many particulars typical papposus is about intermediate
between the extremes.

Type-locality.—“ Habitat in O. Europeo et Asiatico.” (Linneus.)

Distribution —Cireumpolar; through Bering Straits, into Bering Sea, thence
south along the west coast of North America to Washington; on the Asiatic side
to the Sea of Okhotsk. Along the east coast of North America, from Newfoundland
and Labrador south to about 40° north latitude; northward to Davis Strait and
Smith Sound, the northernmost point being reached in Discovery Bay; westward
to Assistance Bay in Barrow Strait, thence to Point Barrow; eastward to Green-
land, Jan Mayen, Spitzbergen, northward to about 794°. South along the Scan-
dinavian coast to the coast of Denmark, Holland, and Belgium, and to the Faroe
Channel, Scotland, England, and Ireland, reaching the southern limit in the western
part of the channel (lat. 48° 30’ N.). From Finmark eastward through Barents Sea
to the Murman coast, thence to the Kara Sea and eastward to the Gulf of Khatanga
(long. 113° 30’ E.) on the Siberian coast. (Atlantic data from Ludwig, Fauna
Arctica.)

Specimens eramined.—One hundred and eighty-six. The letters refer to the
varieties mentioned in the preceding discussion.

328

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Specimens of Solaster papposus examined.

Station.

2843.....

Locality.

Unalaska, Alaska
.| Near Unimak Island, Alaska. - ---

...| Vicinity Shumagin Islands, Alaska

.| Near Kadiak Island, Alaska. ---- -
.| Off Afognak Island, Alaska. -....--

Admiralty Inlet, Washington.....
.| Near Flattery Rocks, Washington
Off Cape Flattery, Washington...

Heceta Bank, Oregon
.| South of Unimak Island, Alaska. .

82M ees ane North of Alaska Peninsula (near
Point Moller).

3445C....| Straits of Juan de Fuca, Washing-
ton.

3459C 7... 2]. <2. COs coe Seo meee hee

3461C....|.---. COs epee ete ean 1 eee ee

3480.....- Amukta Pass, Aleutian Islands...

3518A....| St. Matthew Island, Bering Sea...

3595. ...-- Admiralty Inlet, Washington

Hagenmeister Island, Alaska... ..
Alaska C

St. Paul, Kadiak, Alaska ........

Kadiaks AlASKAs ca. nnaeeenanee

Sanborn Harbor, Nagai, Shuma-
gin Island.

Round Island, Coal Harbor, Shu-
magin Islands, Alaska.

St. Paul Island, Alaska.

Popoff Strait, Alaska

Unalaska

Madiak Alaske ‘Ars ..cs2-6:-5- 5.
Bering Islands A

Depth.

Fathoms.
72
44
48
110
69
69
68
40
171
27
40
59
59
41
38
42
41
34
50
56
12

Nature of bottom. Number. Collection.
DEDDIEs seme sneniecc sone soe 1 U.S. Nat. Mus.
gravel. ifs. cst uecs 223 2c assk 1 Do.
fine'gray\sand=. .5-2----~--1e< 1 Do.
green muds c.----o--- eae es 1 De.
2 Do.
1 Do.
16 Do.
Dpepblesee ees sassece = teweee 2 Do.
ceo COscces sence eee 1 Do.
rocks and shells. ............ 1 Do.
poe GO. sco esate noes. ces 1 Do.
black sand and mud........ 1 Do.
coral, broken shells... ....-.- 1 Do.
black:sand oo. 22. 2~- see ose 1 Do.
gray Sand 352-2 h asec ee soc oe 2 Do.
black gravel..........-.-..-. 2 Do.
plack sand << -22--~-2--<--== 1 Do.
gravel, broken shells.......- 1 Do.
black sand, pebbles, shells. .. 1 Do.
Dlackipebblesis-< sence sees 3 Do.
SONG: coc ne seca oeeteoseces 2 Do.
pebbles, stones.........-...- 5 Do.
sand, pebbles... --......--.. 3 Do.
NoqpooaoRocoGedeSbeestedaGooHe 9 Do.
gravel, sand, shells.......... 3 Do.
gravel, sand, rocks.......... 4 Do.
TOCKYisoscsecenc est Sa eee eee 2 Do.
gravel, sand, pebbles. ......-. 2 Do.
gravel, sand, rocks.......... 1 Do.
black sand, rocky........... 1 Do.
PTeeN MUG. cence sees ae eee 2 Do.
gravel, sand, rocky.....-.... 3 Do.
1 Do.
1 U.S. Nat. Mus., Dr. W. H
Jones.
peor otoecocsc se hectscoscccaai 10 U. 8. Nat. Mus, W. J.
Fisher.
Sin athe nineteen 1 Do.
| weaagelt th natty pteade ceceed 1 U.S. Nat. Mus. Ss naa
snaitehlo ase eee aan sens 4 Do.
1 U.S. Nat. Mus.
3 U.S. Nat. Mus., W. H. Dall.
1 Do.
2 U.S. Nat. Mus., Point Bar-
row expedition.
| 1 U.S. Nat. Mus., W. H. Dall.
| 9 U.S. Nat. Mus., L. Stejneger,
N. Grebnitsky.
bated Suse suas ansee ae saaceeccen 1 Steamer Corwin,1881, ace. No.
16889.
Soote eee esece ne sss eee anaes 1 Steamer Corwin, 1884, acc.

No. 15278, tag 5.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FIS#H1ER. 329

Specimens of Solaster papposus examined—Continued.

] iat

Station. Locality. Depth. Nature of bottom. Number. Collection.
| Fathoms.
Aen nstaey as funatiesice hein omate mod ARES one shen ti Dei sbieeea lara steabtban ac iaeetieae 2 Steamer Corwin, 1884, 15278,
tag 6.
HRA Cesena tecen rest ere nears) cctew nd enero, ose tcc neem 2 | Steamer Corwin, 1884, 15278,
' tag 10.

Port Clarence, Alaska A..........|..-....00- 3 Stanford University.
4246 C ...| Kasaan Bay, Prince of Wales Is- 42447 1 Albatross, 1903.
land, Alaska.
4281 C ...| Chignik Bay, Alaska Peninsula... 42 1 | Do.
30-41 1 Do.
26-30 1 Do.
57-63 1 | Do.
42 1 Albatross, 1905.
4777 B ..-| Near Semisopochro! Island, 52 7 Albatross, 1906.
Aleutians. |
TAB Clemdo re eee 43 ae yo:
4784 B ...| Near Attu Island, Aleutians ..... 135 1 Do.
4787 C ...| Near Medni Island, Commander | 54-57 3 Do.
Islands.
4788 C'.. -|_.--- Oe ene eee 87 ter Dol
4791 C ...| Near Bering Island, Commander 76 1 Do
Islands.
MIST eed ee ion ac ahaa | 72 41 Do.
4795 A...| South of Petropavlovsk, Kam- | 69-48 1 Do.
chatka.
5024 A ...| Off eastern coast of Sakhalin, vi- 67 1 Do.
cinity of Cape Patience.

Remarks.—This species is the type of the genus Crossaster Miller and Troschel,
1840, recognized by Sladen, Perrier, Ludwig, Kehler, and numerous other writers.
Crossaster differs from typical Solaster (as for example, endeca) in having the abac-
tinal skeleton a reticulate structure with large spaces, the penicillate paxille being
correspondingly rather widely spaced. The papule are numerous in each area. In
Solaster the skeleton consists of stellate plates forming a close reticulation, the papule
are few or isolated, and the paxille small and more crowded. These differences are
not constant. Crossaster sguamatus,* which by some writers is regarded as a variety
of papposus, does not conform to the generic diagnosis, for the abactinal plates are
closely placed, with isolated papule, and are less reticulate in arrangement than sey-
eral species of true Solaster. Solaster borealis was first described by me as Crossaster;
the young are obviously Solaster, while the large specimens would be classed as
Crossaster. The difference in the paxille is all that is left to distinguish Crossaster
from Solaster. Even here the divergence is not nearly so great as indicated by
such extremes as papposus and endeca. Intermediate species with skeleton of
Crossaster and the paxille of Solaster, as well as the two types of skeleton in such
obviously close relatives as papposus and sqguamatus, baffle any attempt to make two
generic groups. Whether it is convenient to retain Crossaster as a subgenus depends
upon the.importance attached to such a group. It would seem that the same objec-
tions apply here also.

Japanese specimens.—PI. 95, figs. 1 and 2; pl. 113, fig. 4. In Japanese waters
there is a representative species in general appearance resembling papposus, but

9 See Grieg, 1906, pl. 1, fige. 4 and 5, and text, fig. 7 (1-3).

330 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

differing in several important details. So far as the limited number of specimens
allows one to judge, these differences are constant. Solaster japonicus, new species,
differs from Solaster papposus in having more numerous and less prominent mar-
ginal plates, and seven to nine furrow spinelets (four in North Pacific papposus).
Rays ten, rarely nine. A large specimen (R=80 mm., r=40 mm., R=2 r) has
twenty-six to twenty-eight marginals, while papposus of the same size have eighteen
to twenty. Typically these marginals are separated by less than their own width,
form a very regular series, their spinelets being very numerous, the peripheral quite
short, and the rest graduated toward the center ones, which are decidedly shorter
than the subambulacral spines. In some specimens the marginals are decidedly
compressed and the longest spines near the ventral side of the paxilla. The adam-
bulacral furrow spinelets are shorter than the actinals, less heavy, and in a regular
comb, the middle members the longest. The spinelets are joined by a web nearly
to the tip, the edges of the web forming around the end of each spinelet a flattened
pad or flap. This web is joined by the transverse web of the actinal comb, a
peculiarity not observed in North Pacific papposus. As a result the furrow and
actinal series of adambulacral spines are more or less closely joined. In the actinal
series are seven to nine tapering spinules (six in small specimens), longer and stouter
than the furrow spinelets, webbed for about half their length, the web continued
as a flange of tissue over the pointed tip of each spine. Mouth plates with twelve
to fourteen webbed marginal spinelets, the inner two or three abruptly the largest:
suboral spinules four to nine webbed, but not connected with the marginal series.
They are variable in length, though longer than the shorter marginals. Actinal
interradial areas paved with numerous small, four to eight spied paxille, which
do not extend far along ray. Abactinal skeleton open meshed as in papposus, the
paxille being of the same general type. Paxille numerous with long central and
short peripheral spinelets (apparently deep-water examples have longer spinelets).
No well-defined midradial series larger than the rest. (See pl. 95, figs. 1 and 2.)
Eighteen specimens of this species have been examined from the following localities:
Station 4818 (type-locality), off Sado Island, Sea of Japan, 225-245 fathoms, fine
brown mud, eleven specimens; 4859, off Matsushima, Sea of Japan, 93 fathoms,
green mud, one specimen: 4860; same locality, 122 fathoms, green mud, one speci-
men; 4867, Sea of Japan (36° 31’ N.; 129° 46’ E.), 150 fathoms, green mud, four
specimens; 4871, Sea of Japan (36° 29’ 30’’ N.; 129° 43’ 30’’ E.), 60 fathoms, one
specimen.

Type.

Cat. No. 27788, U.S.N.M.

Genus HETEROZONIAS Fisher.

Heterozonias Fisner, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 172. Type, Crossaster

alternatus Fisher.

Dragnosis.—Solasteridee with a wide-meshed abactinal skeleton, and small
pseudopaxille like Solaster, but differing from that genus in having a complete series
of actinal intermediate plates to tip of ray and in the arrangement of marginal
paxille, which consist of transversely oriented prominent inferomarginals alter-
nating with longitudinally oriented less prominent superomarginals, all in a single
linear series. Adambulacral armature as in Solaster. Papule numerous and
prominent.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 331
HETEROZONIAS ALTERNATUS (Fisher).
Pl. 96, figs. 1-3; pl. 112, figs. 5, 5a—b.

Crossaster alternatus Fisner, Proc. Wash. Acad. Sci., vol. 8, Aug. 14, 1906, p. 181.

Diagnosis.—Rays ten. R=2.6 to 2.9r. Breadth of ray at base, 4r. General
form flattened; abactinal surface of disk slightly convex, capable of inflation, but
flattened on central part; abactinal surface of rays slightly rounded; margins well
rounded; actinal surface nearly flat; interbrachial angles rather acute; abactinal
skeleton open reticulate, the ossicles slenderer than in S. papposus; papule large;
paxille small, well spaced; marginal plates of two kinds in a single series; more
prominently spinous transversely placed paxilliform plates alternating with usually
two longitudinally oriented plates with very short spinelets; marginal plates actinal
in position on basal half of ray; actinal interradial areas small, with few plates set
fairly close together; a single series of very small intermediate plates extending to
end of ray; adambulacral plates with four to eight furrow spinelets and a transverse
series of four actinal spinules.

Description.—Abactinal integument rather thin but tough and pliable, pareh-
ment-like, quite opaque, and obscuring the ossicles unless dried. Skeleton open
and forming a network with fairly wide meshes, which are irregular and largest on
disk; connecting ossicles slender, often irregular; inclosed within meshes, small
free irregular ossicles; these often absent, but usually present on disk and most
numerous near its center. Paxillee with a two to four slender-lobed base and a low
stout pedicel surmounted by usually four to six, or far along ray two or three, and
on disk’ sometimes as many as nine, rough, delicate tapering, pointed, spinelets
inclosed in a delicate membranous sac, which fits tightly about each spinule for
about half its length, leaving only its basal part obscured. In consequence of the
open character of skeleton, the paxille are well spaced, but are very much smaller
and more numerous than in S. papposus. They are largest and most widely spaced
on radial areas midway between center of disk and margin, thence rapidly diminish
in size toward extremity of ray and less toward center of disk. The paxille are
not arranged in very evident series, although along the lateral regions of the rays
an arrangement in quincunx is observable, though not so well marked as in S. borealis
or paxillatus. A bare suleus leads from each interradial angle halfway to center of
disk. These bare areas are about 1 mm. wide and are paved solidly with ossicles,
which are the upper edge of the incomplete calcareous interbrachial partition. At
the inner end of this bare area, in one interradius, is the madreporie body sur-
rounded by several paxille; in the other radii several slightly larger paxille hold
a similar position. Papule large, partially obscuring the small paxille; absent
from bare interradial areas; two to seven or eight to each mesh of skeleton on rays
and as many as fifteen on disk, or even more where meshes are incomplete. Papulx
commonly 3 mm. long, pointed. In the interradii a number of the abactinal plates
are actinal in position, because the marginal plates are drawn inward toward the
mouth. Thus in the type the distance between marginal plates and interradial
angle is 6 to 8 mm. Consequently the dorsal integument with plates and papule
is drawn upon actinal surface.

332 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

“

Marginal plates conspicuous; about fourteen to twenty prominent, compressed,
rather widely spaced, transversely oriented, paxilliform plates represent the infero-
marginal series, and between any two of these in the same linear series are one to
three, usually two, longitudinally oriented, much lower and slightly smaller plates
representing the superomarginal series, although now forming a single series with
inferomarginals. The prominent marginals become more conspicuous toward tip
of ray, acquiring a heavy, compressed pedicel often higher than its width at top
and very paxilliform in appearance, bearing two transverse rows of about eight to
sixteen tapering needle-like spines one and a half to twice as long as the pedicel,
which increase in length but decrease in number toward extremity of ray. Beyond
proximal fourth of ray there are two well-defined series of these spines, of which
the adoral are the shorter, and in the other series about three skin-covered spines
become much larger than the rest and have very sharp points. Distally thespines
form two palmate series, but there is more or less variation in their numbers. The
nonprominent longitudinally oriented plates vary considerably in size and decrease
markedly in size distad, whereas the others become more prominent. Except at
base of ray they are not nearly so high as transverse plates and are rounded to
elliptical-oblong, bearing upward to twenty-five very short spinelets in about three
or four longitudinal series. At tip of ray these plates are very small, bearing a
group of five or more delicate spinelets.

Actinal interradial areas small, with small, closely placed, paxilliform plates
bearing four to ten spinelets, which are more delicate than those of abactinal paxille,
although the latter are of about the same size. Interradial paxille about ten to
twenty in number. Proceeding along ray almost to its tip is an inconspicuous
series of very small actinal intermediate plates, often rather widely separated, a
plate usually opposite each adambulacral plate, and distally bearing only a single
small spinelet, or none at all, proximally with two to five spinelets.

Adambulacral plates with a palmate furrow series proximally of six to eight,
distally of four or five, very delicate, tapering sharp skin-covered spinules united
for about a third their length by a web. Mesial spinules longest (about one-third
width of plate in length), thence decreasing in length toward either end of series.
On actinal surface of plate is a transverse comb of four or five slender, needle-like
sharp spines, the two or three mesial the longest, the inner usually slightly longer
than furrow spinules though often nearly twice as long; outermost spine usually
nearly equal to the longest, which exceeds in length width of plate. These spines,
like those of furrow series, are invested in membrane, which forms vane-like lateral
expansions (causing the spinule to appear broad and flat near base) and unites them
in a common web by their bases. On distal part of ray the large adambulacral
spines are similar in size and appearance to the larger inferomarginal spines already
described.

Mouth plates of the usual shape, rather prominent actinally. Each plate has
three long slightly-tapering pointed spines at inner end, these decreasing in size out-
ward, so that third spine is about one-half length of innermost; thence the series
is continued to end of plate in seven or eight much shorter spines resembling those
of first adambulacral plate. All spines are skin-covered and united basally by a

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 333

web. On actinal surface, parallel with median suture and slightly nearer it than
free margin, is a comb of two to eight skin-covered sometimes basally webbed
spines similar to but smaller than corresponding series of first inferomarginal.

Madreporic body situated about midway between center of disk and margin,
irregularly circular or oval, convex, irregularly and centrifugally striated; about 3
mm. in diameter.

Color in life, “salmon pink.”

Young.—Young specimens agree very well with the large examples, except
that the papule are less numerous, and there is a slight reduction in number of
spines of interradial, marginal, and adambulacral plates, as well as fewer interradial
and marginal plates. In small specimens there is more often only one superomar-
ginal plate interpolated between the transversely oriented inferomarginals, and the
former are slightly more superior in position, at base of ray, than in adults. Adam-
bulacrals commonly with three to five furrow spinules proximally and three or
four actinal. Usually only one or two large papule to a mesh. Abactinal spine-
lets are not fewer in number than in adults, except in very young specimens, but
the suboral spines are often lacking. The marginal plates are very nearly on the
margin in the interbrachial ares.

Type.—Cat. No. 21932, U.S.N.M.

Type-locality —Albatross station 2839, Santa Barbara Islands, California, in
414 fathoms on gray sand.

Distribution From Washington to San Diego, California, in 316 to 603
fathoms.

Specimens examined.—Kighty-three, from the following stations:

Specimens of Heterozonias alternatus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
TOS soa Santa Barbara Islands, California.... 4i4 32 U.S. Nat. Mus.
2036...... | Off San Diego, California 359 2 Do.
ROT xaciva eee G0scesctesate enna rose sown Sct 404 $ Do.
W000 so6 == | on Anacapa Island, California........ 603 2 Do.
3073......| Off Sea Lion Rock, Washington ..... 477 20 Do.
SR 3. | South of Monterey Bay, California..... 316 2 Do.
Bons: cents | Off Destruction Island, Washington. 516 6 Do.
4307.....- | Vicinity of San Diego, California...... 490-496 7 Albatross, 1904.
APF 325: hae Goh Seas cscs szneiede- Suds daeeeh 471-510 5 Do.
4400...... Between San Diego and San Clemente, | 500-507 3 Do.
Islands. |
4427...... Off Santa Cruz Islands, California...) 447-510 1 Do.

Remarks.—This is one of the best-marked Solasteride of the North Pacific,
the remarkable arrangement of the marginal plates making it easily distinguishable
from any described form.

334 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Genus LOPHASTER Verrill.

Lophaster Verri1, Amer. Journ. Sci. and Arts, ser. 3, vol. 16, 1878, p. 214. Type, Solaster furcifer
Diiben and Koren.

Sarkaster Lupwic, Mem. Mus. Comp. Zodl., vol. 32, July, 1905, p. 185. Type Sarkaster validus
Ludwig. ’

Diagnosis.—Five-rayed Solasteride with cruciform or lobed abactinal paxilli-
form plates (without intermediate connecting ossicles) forming a quadrate or polygo-
nal meshed skeleton; with two well-developed series of marginal paxille, and with
usually a single series of spaced actinal intermediate plates extending far along ray,
with or without a paxilliform tuft of spinelets; adambulacral plates with a furrow
comb usually webbed, and with an actinal comb of spines, often oblique, behind the
furrow series.

Remarks.—Much may be said for merging this genus with Solaster. The type
is, I believe, more distinct than Crossaster, and is kept separate for convenience,
and in the belief that, although close to Solaster, intergrading species have not been
demonstrated.

A curious misapprehension has for some reason grown up around this genus.
Sladen states in his key that there are no actinal intermediate plates along the ray,
and Hamann in Bronn’s Thierreich follows Sladen. A specimen of L. furcifer from
the Kara Sea, which I have, has a single series of spaced actinal intermediate paxille
placed alternately with the inferomarginals, extending very nearly to the tip of the
ray (to within 6 mm., R being 45mm.). Specimens of L. furcilliger vexator described
below are similarly provided, although in some specimens they are suppressed
except for widely-spaced plates on the outer part of the ray. L. furcilliger vexator is
an undoubted Lophaster, and is believed to intergrade with furcilliger in deeper
water; with the latter Sarkaster validus is either identical or closely related. It fol-
lows, therefore, that Sarkaster can not be maintained, for its alleged distinguishing
character is shared by Lophaster. Even if in Lophaster furcifer intermediate plates
are not always present on the ray, the case against Sarkaster is not altered.

LOPHASTER FURCILLIGER Fisher.
Pl. 79, figs. 1, 2; pl. 114, figs. 1, la—g; pl. 116, fig. 5.
Lophaster furcilliger Fisuer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 312.

Diagnosis—Rays five. R=70 mm., r=16 mm., R=4.38 r. Breadth of ray at
base 18 mm. or a little less than one-fourth R. Disk rather small, rays long, tapering
from narrow base very gradually to pointed extremity, ambulacral furrow wide
and shallow; tube feet with small sucking disk. As compared with L. furcifer, rays
narrower and longer, abactinal skeleton more open, paxille much slenderer, higher,
with longer and more delicate spinelets having only two or three, less often four
sharp points; marginal paxille higher, more spaced, actinal interradial areas much
smaller with fewer plates, which with the intermediate plates along ray have more
delicate spinelets; adambulacral spinelets longer and more delicate, and the series
as wide apart as length of a series, not close together.

Description.—Abactinal surface beset with well-spaced paxille, which have
rather long (2 to 3 mm.) slender pedicels, with a capitate summit beset with two

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 835

glomerular tufts of numerous (twenty or even more) long, delicate, glassy spiculi-
form spinelets, flattened, flaring, and with two or three, rarely four, delicate rather
long teeth at the tip, these spinelets about 1 to 1.2 mm. in length and radiating in all
directions; paxille largest on disk, decreasing in size toward end of ray. Papule
two or three to a mesh of the skeleton in small specimens, gradually increasing with
the size of the individual to about six or eight in type. There are more on disk
than onthe rays. Abactinal plates with four lobes on the lateral portions of the area
and with two, three, four, or even five on the midradial region, where the meshes
are irregular, five- or four-sided, and larger than the lateral quadrate meshes. All
the meshes are much larger than in fureifer.

Marginal paxille forming two conspicuous series alongside of ray; they are well
spaced and the superomarginals stand above the inferomarginals—that is, do not
alternate. Superomarginal paxille much larger than the adjacent abactinal, and
with stouter and longer pedicels (3.5 mm., type), and the spinelets have a number
of sharp teeth scattered along the outer part (Pl. 114, fig. 1f); sometimes they are
trifurcate. Inferomarginal paxille similar to the superomarginal, but sometines a
trifle more robust; about twenty-three to twenty-five to a ray in type.

Adambulacral plates short and broad, with wide intervals between successive
plates, the suture being wider than the length of exposed surface of plate; furrow
series consisting of four spinelets at base of ray, which are soon reduced to three,
then to two beyond middle of ray, and finally to one; these spinelets rather long,
delicate, skin-covered and united for a third or a half their length by a web; when
there are four, they are graduated in length toward the adoral, which is shortest;
when there are three, the middle is usually longest, and commonly the adoral when
there are two; the separate series well spaced from one another and the spinelets as
long as or even longer than width ‘of the plate. On the actinal surface is a series
of four, three, or two longer and more robust, tapering, pointed spinelets, disposed
in an oblique transverse series, and decreasing in length as they recede from the
furrow, skin-covered, often having lateral expansions, and the bases united by
membrane; when there are three or four spines the outer is usually much shorter
than the rest. Northern specimens, as, for example, from station 3075, have the
spinelets much more heavily enveloped with membrane.

Mouth plates fairly large, spade-shaped, and prominent actinally; median
suture wide, roofed with membrane; furrow spinelets seven or eight, long, skin-
covered, united for about a third to a half their length by a web, increasing very
slightly in length toward the innermost, which is rather more robust than the
others. Near the middle of the suture margin are one to four much slenderer
spinelets, which are shorter than those on furrow. They are webbed basally.
The type has only one or two spines; apparently the specimens from very great
depths have the fewest and slenderest suborals.

Actinal interradial areas much reduced and with few spine-bearing plates in each
interradius. The type has about eight or nine low paxille within the area deter-
mined by the first marginal paxille. An extreme specimen from station 4237 not
typical has about twenty-four plates, including the area to the limit of two rows of
plates on each ray. These paxille have low pedicels and spinelets somewhat
longer ending in several points. A single series extends from the third inferomar-

336 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

ginal nearly to tip of ray, decreasing in size distally. These paxille usually stand
in the interval between two inferomarginals and are therefore widely separated.
Near the base of ray, however, the plates are closer and there may also be one
opposite the inferomarginal, although sometimes it lacks a paxillar develepment.
In the interradial region the plates touch one another.

Madreporic body convex, situated about midway between center and margin.

Color in life, salmon pink, tip of arms yellowish white.

Variations.—The following table shows variations in proportions of Californian
examples:

Breadth
Station. R. r. R:r. of ray at
base.
Mype tees t sess aes ee eee see teeeaciom 70 16 4.38: 1 218.0
AA05 AN ee |. kee e ee eer gre ce tease 32 8 ai 9.5
AAO eaieare soto eect mie eae = etataiare 30 7 4.2:1 9.0
ASOT ce mre ame ee Se ate nla mn alate 35 7 a 7.0
29 7 4.1: 1 9.0
The following are similar measurements of L. furcifer:
Breadth
Locality. R. r. Rit. of ray at
base.
36 13 2.8: 1 a16
53 20 2.65: 1 21-25
| 45 7 2.7: 1 19

a Compare with length of ray, which will give an idea of robustness of ray.
» Off Nova Scotia, 218 fathoms.
¢ Measurements from figure, Danielssen and Koren, Asteroidea: Norwegian North Atlantic Expedition, 1884, pl. 8, fig. 12.

The Alaskan examples are not quite typical, tending slightly toward vezator
in having less widely spaced adambulacrals, slightly compressed inferomarginal
paxillz: and not quite so high or so delicate paxille. Some examples recorded under
vexator from station 3331 stand nearly midway between the extremes, though nearer
vezator in the structure of abactinal and marginal paxille, and spinelets. Possibly
the intermediate depths at which these were taken may in a measure account for
the less extreme development of the characters which make furcilliger.

The type from deep water has the distinctive characters of the species most
pronounced; the long rays, long paxille pedicels and their crown of delicate glassy
fureate spiculiform spinelets usually radiating to form beautiful tufts resembling
miniature composite flowers; very prominent marginal paxille, small actinal
interradial areas with few plates, and long delicate furrow spinelets.

Type.—Cat. No. 22341, U.S.N.M.

Ty pe-locality.—Albatross station 4425, between Santa Barbara and San Nicholas
islands, 1,084 to 1,100 fathoms, green mud, sand, globigerina.

Mstribution.—From south of the Alaska Peninsula to southern California,
thence south to the Galapagos Islands, 192 to 1,100 fathoms.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 337

Specimens examined.—Twenty-four, from the following localities:

Specimens of Lophaster furcilliger examined.

D
Nicholas Islands, California.
Do.

Station. Locality. Depth. Nature of bottom. Number. | Collection.
Fathoms. | j
SEBO eas Off Queen Charlotte Island, B.C....| | 876 | TOBTENNTIAS 5 5 Si koe ca eke eee 16 | U.S. Nat. Mus.
8075...... Off Sea Lion Rock, Washington .... Oo) cca OG cearceseateat Rivecnemesansp 1 Do.
3338...... South of Unga Island, Alaskan 625 green mud, sand.................. 1 Do.
Peninsula. |
AZBT 5+ Vicinity of Yes Bay, Behm Canal, 192 reer NOL eae ook ooa- eee | 1 Albatross, 1903.
Alaska.
4407...... Vicinity of Santa Catalina Island, | 334-600 | rocks, shells, fine gray sand. ...... | 1 Albatross, 1904.
California.
4425...... Between Santa Barbara and San 1,084-1,100| green mud, sand, globigerina...... 3 | 0.
}
|
|

Remarks.—This species, although very well marked in great depths, seems
gradually to change in shallower water to L. furcilliger verator. However, an exam-
ple referable to furcilliger rather than to verator was taken in 192 fathoms off south-
east Alaska. The specimens of verator from 350 fathoms north of Unalaska are not
typical, but are nearer vezator than to the present form. The specimens of furcilliger
from stations 2860 and 3075 are not typical, but vary slightly in the direction of
verator. It is probable that there is a slight change northward in addition to the
change caused, it would seem, by shallow water. Specimens of verator from shallow
water show the characteristics of the race more prominently than do those from
deeper than 200 fathoms.

Venturing somewhat upon the speculative, one may suggest that verator varies
with latitude as well as with depth, and that ranging into deep water it gradually
becomes the form known as furcilliger, which assumes its extreme characters only
at great depths (as the type-locality). If northern furcilliger is derived from
northern verator, and southern furcilliger from the southern verator, specimens of
fureilliger from the two extremes would naturally show some differences even if
the remoteness geographically did not also exert an influence, which it probably
does. It is likely that furcilliger pushes farther south at great depths than its
possible originator does at shallower. Also, once established furcilliger may return
to shallower water where bottom and temperature are favorable; at least specimens
nearer the type than to any other form are found in less than 200 fathoms. It
hardly seems possible that these are individual variants of vezator.

L. vexator is much nearer to furcifer than is furcilliger. This seems very sug-
gestive. If intergradation occurs between furcifer and vexator and verator inter-
grades with furcilliger, as seems demonstrable, then logically, Sarkaster validus
(searcely if at all specifically separable from furcilliger) would be connected by a
complete series with L. furcifer, and would naturally become, along with furcilliger,
a subspecies of furcifer. The intergradation of furcifer and verator is, however, yet
to be demonstrated, but there is no doubt that Sarkaster is only an extreme of
Lophaster.

57444°—Bull. 76—11——22

338 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Sarkaster validus Ludwig? ranges from the Galapagos to the Tres Marias Islands
and from 680 to 286 fathoms. There are no important differences between this
species and the variable Lophaster furcilliger. The type of furcilliger happens to be
from deeper water than validus, and has more delicate paxille and spines generally.
The Alaskan examples of furcilliger from lesser depths are nearer validus, having
the inferomarginals larger than the superomarginals, lower paxille, and stouter
rays. Although Ludwig’s excellent description and figures seem to indicate slight
differences, these are more than compensated by the variations of the specimens
at my disposal. The Alaskan specimens of furcilliger are quite as distinct as is
validus but if one started to name the intergrading forms with verator there would

be no end.
LOPHASTER FURCILLIGER VEXATOR Fisher.

Pl. 80, figs. 1, 2; pl. 114, figs. 2, 2a.
Lophaster furcilliger vexator FisHER, Zool. Anz., vol. 35, March 29, 1910, p. 574.

Diagnosis —Similar to L. furcilliger, but with larger disk, thicker rays, much
stouter paxille both abactinal and marginal, and with stouter abactinal spinelets
having six to ten points instead of two to four, heavier adambulacral spines, and
more closely placed adambulacral plates. Perhaps intermediate between L. furcifer
and L. furcilliger. Type, R=89 mm.; r=27 mm.; R=3.3 r. Breadth of ray at
base, 30 mm., or about one-third R.

Description. —This form, variety, race, or small species stands near to furcilliger
although the extremes are decidedly different in appearance. Vezator is a much
stouter animal with thicker rays, more robust paxille, and stouter skeletal parts,
due possibly to the shallower habitat. In furcilliger the pedicels of paxille are
decidedly slender and the spinelets are about a half to two-thirds the extreme
length of pedicel. The spinelets are very delicate and glassy with two or three
extremely sharp points at the flaring tip (sometimes four), and commonly one or
two series of perforations along the middle of the spinelet. But in typical vexator
the pedicels are shorter and stouter, and the spinelets are about the same length as
in furcilliger or shorter, ending in six to ten points distributed along the outer
third or half of the spinelet. There are also several rows of perforations along the
center of spinelet where the wing-like expansions or buttresses which terminate in
points join to form the core of the spinelet. The abactinal plates or bases of paxille
are of the same general structure as in furcilliger, but usually stouter. Papule in
type and large examples ten to twenty-five to an area on disk and midradial region,
laterally five or six. Proximally they sometimes encroach upon the marginals and
lie intra-marginally, but this is not usual and occurs only in very large specimens
such as the type.

Marginal plates very robust, the inferomarginal paxille stouter than supero-
marginal and compressed toward the end in large examples. The difference in
size between the marginal paxille in typical furcilliger is slight or not marked and
the inferomarginals are not compressed at the tip. The marginal spinelets are
variable in length, but usually shorter, stouter, and more dentate than in furcilliger.

Adambulacral plates with the furrow combs typically nearer together than in
furcilliger, but some specimens having them slightly spaced and hence interme-

4 1905, p. 185, pl. 15, figs. 75, 76; pl. 29, figs. 171-173; pl. 30, figs. 174-177.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 339

diate (station 3331). Armature of four or three furrow spines, but specimens from
station 3331 with three, two, and finally one as in furcilliger, and the series are
slightly spaced also, as noted above. The subambulacral spines agree very nearly
in number with the furrow, to which they are subequal in length, but are consider-
ably longer and thicker in a large (and not in a small) specimen from station 3331.
The armature differs from that of furcilliger mainly in having stouter spines,
narrower spaces between consecutive series, and heavier membrane on the spines.

Mouth plates differ from those of furcilliger in having slightly stouter furrow
spines and more numerous actinal or suboral spines, there being in the type three
to six about as long as the furrow spines. In fureilliger (typical) only one or two
spines are present, although a northern example, which is not typical in some other
respects, has three suboral spines.

Actinal intermediate plates about as numerous as in furcilliger. In a specimen
from station 3330 the paxille along the ray are very inconspicuous, but in another
example also large, they are fairly large, fully as large as in typical furcilliger.
A specimen station 3331 lacks actinal intermediate paxille on the rays, although
here and there far along ray a small one may be seen.

Type—Cat. No. 27789, U.S.N.M.

Type-locality.— Albatross station 3350, off Point Arena, northern California,
75 fathoms, fine sand, mud; two specimens.

Distribution—Southern Bering Sea to northern California in 75 to 350 fathoms,
typically in less than 200, intergrading evidently in deeper water with the following
form. Probably westward to Spitzbergen.

Specimens examined.—Fourteen, all from the National Museum collection, as
follows:

Station 3330, north of Unalaska, 351 fathoms, black sand, mud, four; station
3331,same locality, 350 fathoms, mud, four; station 3339, east of Shumagin Islands,
138 fathoms, mud, gravel, two; station 3350, type-locality, two; mio locality, two.

Remarks.—This form may ultimately be found to intergrade with L. furcifer
of the Arctic Ocean and of the North Atlantic. Vexator differs from typical furcifer
in having a more open abactinal skeleton with consequently more widely spaced
paxille, higher paxille with longer spinelets, much smaller actinal paxille (about
as in furcilliger) and longer adambulacral spines. If equal sized specimens of the
two forms are compared, L. furcifer is seen to have wide rounded interbrachial
arcs which merge gradually into the ray. Vezator has a smaller disk, never rounded
interbrachial ares, but acute angles, the rays being sometimes swollen at base so °
that the marginal and adjacent abactinal paxille of two rays interlock. L. fur-
cifer reminds one of a five-rayed Solaster, whereas L. verator suggests a five-rayed
“ Orossaster.”’

Grieg * has furnished notes on the variation of L. furcifer from which it would
appear that a form resembling vexator occurs off Spitzbergen .? This can not

a See Report of the Second Norwegian Arctic Expedition in the Fram, 1898-1902, Echinodermata,
1907, 8 (pl. 1, figs. 3 to5; pl. 3, fig.1). Echinodermen von dem norwegischen Fishereidampfer Michael
Sars in den Jahren 1900 to 1903 gesammelt, III, Asteroidea, Bergens Museums Aarbog, 1906, No. 13,
p. 69, text fig. 9, and pl. 2, figs. 6-9.

> Text fig. 9, Asteroidea, Michael Sars Expedition,

340 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

be confused with true furcifer as the rays have an entirely different shape, and
the paxille are more spaced. It is not possible to determine whether this form
figured by Grieg is a separate species or only a variation of furcifer. At any rate it
must be considered separately from furcifer, and if the same as the race inhabiting
Bering Sea and the depths off northwestern North America, should be designated
as distinct from fureifer, the typical form of which appears not to be found in this

area.
Vexator may be a species which invades the North Atlantic and which is quite

distinct from the resident furcifer. We seem to have to do with at least two dis-
tinct forms which range together to the north of Scandinavia. Even if verator is
only a variety or “forma”-of furcifer, conditions are such that it is the type which
has survived in the North Pacific at lesser depths, and as it evidently intergrades
with furcilliger its name has been linked with that species rather than with furcifer,
with which intergradation is not demonstrated.

Family KORETHRASTERID Danielssen and Koren, 1884.

Korethrasteridz DANIELSSEN and Koren, 1884; Bex, 1892; GreGory, 1900; FisHer, 1905.

Korethrasterine SLADEN, 1889; PERRIER, 1904.

Spinulosa resembling superficially the Solasteride, but lacking actinal inter-
mediate plates, and with the armature of the adambulacral plates forming with
that of the inferomarginals a single transverse series, never webbed; inferomar-
ginals never paxilliform; interbrachial septa partly calcified; abactinal skeleton
composed of roundish plates, or lobed plates forming a wide mesh, bearing a tuft
of spinelets either independent or united by web; mouth plates plowshare-shaped,
resembling those of the Pterasteride.

Remarks.—This small family is usually placed under the Solasteride which it
resembles in respect to the abactinal skeleton and spinulation, but differs in lack-
ing actinal intermediate plates and in having the mouth plates of the Pteraster
type, not the Solaster form. The armature of the adambulacral plates is also differ-
ent, being nearer that of the Pterasteride, though entirely lacking a web. The
inferomarginals are never paxilliform as in the Solasteridz and Pterasteride.

The Korethrasteride seem to stand naturally between the Solasteride and
Pterasteride, combining a few of the characteristics of each family. The Solas-
teride are more homogeneous if Peribolaster and Korethraster are removed, for the
main likeness is to be found only in the abactinal skeleton. The reticulate structure
such as is found in Peribolaster and that of independent plates as in Korethraster
is paralleled in the Solasteridw, where the principal genera have a reticulate skeleton,
but Letmaster (olim Ctenaster) is more comparable to Korethraster, having the
abactinal platelets not united by intermediate ossicles or by their own processes.

The structure of the actinal skeleton is more homogeneous, however, and is
distinctly nearer that of the Pterasteride than of the Solasteride, for the mouth
plates are of the characteristic Pteraster form and the actinal intermediate plates of
the Solasteridx are lacking. The inferomarginals are never paxilliform but bear a
prominent spine (or two spines) and are therefore different from those of the
Pterasteride. The adambulacral armature is more like that of Pteraster than of
Solaster.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 341

KEY TO THE KNOWN GENERA OF KORETHRASTERID®.

Abactinal plates polygonal or roundish in outline, the spines composing the paxilliform tufts naked

ANI OM IZOG Dee Pere tae ees tar rassee = Gh a4 SANS Des oho hd se enone | che Korethraster Thompson.
Abactinal plates cruciform forming wide meshes; the spines composing the paxilliform tufts in
membranous sacs which are united together in the interior of the group....-.... Peribolaster Sladen.

Genus PERIBOLASTER Sladen.

Peribolaster StaDEN, Narr. Challenger Exp., vol. 1, 1885, p. 616 (no type-species); Challenger
Asteroidea, 1889, p. 464. Type, P. folliculatus.

Diagnosis.—Korethrasteride having the abactinal surface paved with cruci-
form plates whose delicate lobes are connected by supplementary ossicles forming
a fairly regular network of large quadrangular meshes over the entire surface.
On the center of each primary abactinal plate is borne on a small boss a fascicule
of delicate, subequal spinelets, enveloped in membranous sheaths, united in the
interior of the fascicule. One to several papule present in each mesh; no actinal
papule. No actinal intermediate plates. Adambulacral armature forming with
that of the inferomarginal plates a transverse series of four to six large single
isolated spinelets invested with membranous sacs. Mouth plates fairly large, of
the type of the Pterasteride, and with a prominent median keel, a few marginal
and a single or no suboral spine. Tube feet in two or four rows, with fleshy button-
like terminal disk. No pedicellariz.

PERIBOLASTER BISERIALIS Fisher.
Pl. 97, figs. 1, 2; pl. 114, figs. 3, 3a-c; pl. 115, fig. 5.
Peribolaster biserialis FisHer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 313.

Diagnosis.—Differing from P. folliculatus Sladen in having the tube-feet in
two series, in having a much larger disk and relatively shorter rays, a more open
abactinal skeleton, with fewer and more widely spaced primary plates, and longer
spinules; adambulacral plates with three and inferomarginals with a single spine,
the four forming a single transverse series (six in folliculatus) no suboral spine
as in folliculatus. R=25 mm.; r=10 mm.; R=2.5 r. (Type, R=20 mm.; r=
7.5 mm.; R=2.6 r; breadth of ray at base, 8.5 mm.) Abactinal surface convex,
rather flexible; form stellate; rays rather short and robust, wider than high,
tapering, the outer part somewhat slender and often recurved; interbrachial ares
acute, the rays being constricted slightly at the base; edges of ray rounded, due
-to inflation of abactinal integument; tube feet strictly biserial.

Description.—Abactinal plates ordinarily obscured by the skin unless dried.
These plates are usually four- (sometimes three-) lobed, the lobes being con-
nected by rod-like intermediate ossicles typically longer than the lobes them-
selves, forming a very open reticulate skeleton with wide interspaces for the papule.
No serial arrangement of plates is very apparent but they are nevertheless in fairly
regular order, there being in the largest example about five longitudinal series
of plates. The plates of adjacent rows alternate and the papular areas are quad-
rate. Each plate has a slight eminence or tabulum surmounted by a brush-like
fascicule of three to five delicate, glassy, sharp spinules about 3 mm. long, which
are united into a compact group by a membrane, the spinules sometimes radiating

342 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

a trifle; spinules decreasing in length toward tip of ray, but increasing slightly
toward margin. Papule lomg and conspicuous, three to seven to an area. The
spinules give a bristling appearance to the abactinal surface, and the membrane
connecting the spinules is frequently emarginated at their tips. The groups appear
distinctly spaced owing to the large meshes of the dorsal skeleton. Anal aper-
ture surrounded by a row of short obtuse spinelets, and to one side of it the flat
subcircular central plate may be seen in a dried specimen.

Actinal surface paved with two longitudinal rows of band-like plates, which
are comparatively wide and short: (1) the adambulacrals bordering the furrow,
and (2) external to them a series of equally regular plates, probably the inferomar-
ginals. Adambulacrals much wider than long and separated by prominent sutures;
each plate with a transverse series of three prominent spinules, sheathed individ-
ually in membrane, which extends beyond the tip in a short sacculus; each spinule
surmounting a slight boss on the plate, one situated on the furrow margin, another
on the extreme other margin, and one exactly midway between the two; inner
spinule 2 mm. long, and the other two successively slightly longer and stouter.

Inferomarginal plates a trifle longer than adambulacrals and not quite so
wide, so that they do not exactly correspond to the latter, even though the sutures
are not so prominent. Surface of each plate actinal in position and slightly con-
vex, and the outer, lateral end bearing a prominent membrane-invested spinule,
3 to 3.5 mm. long, which superficially appears to stand in a linear series with the
adambulacral armature; inferomarginal spinules, of which about thirty-five can
be counted to the side of each ray, decreasing in size toward extremity, where
both marginal and adambulacral plates with their armature are very small; about
ten lateral cruciform plates immediately adjacent to (above) inferomarginals,
form perhaps the superomarginal series.

Mouth plates very prominent actinally, the exposed surface rising toward
the median suture; the combined pair thus with a very prominent but rounded
beak about midway between inner and outer angles, or a little nearer the outer;
margin of each plate rounded, but somewhat angular where the furrow edge merges
with that turned toward the mouth; general form of mouth plates strongly re-
calling that of Pteraster and allied genera, the median beak on the actinostomial
margin, in line of the median suture, being present; armature consisting of four
marginal spinules, that nearest the mesial beak being largest, the third sit-
uated on angle between furrow and actinostomial margin at a lower level in the
furrow than the other spinules, and often directed across or down into the furrow,
while the others are reflexed; fourth spine commonly standing somewhat on actinal
surface; all enveloped in membrane, which is prolonged beyond the tip in a sacculus.

Actinal interradial areas very small; apparently no actinal intermediate
plates.

Madreporie body of medium size, situated about midway between center of
disk and margin, irregularly circular in outline, with coarse, often branching, cen-
trifugal ridges and narrower striations; anal opening somewhat eccentric, sur-
rounded by low, cylindrical, granuliform spinelets; ambulacral furrows wide;
actinostome not very wide, often completely closed over by mouth plates; tube feet
in two series and with a button-like terminal disk.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 343

Anatomical notes.—Stomach large and eversible, the hepatic corca being large
and extending to end of ray. Intestinal coecum of medium size, saccular, depressed,
slightly subdivided on the periphery into a few irregular broad lobes. Gonads on either
side of the interradial septum and opening on the side of the ray close to interradial
line. Septa partly calcified. Ampullex double. Tube feet without calcareous deposits.

Type.—Cat. No. 22342, U.S.N.M.

Type-locality.— Albatross station 4410, between Santa Catalina and Santa
Barbara Islands, California, 178 to 195 fathoms, fine gray sand and rocks.

Dstribution.—Southern Bering Sea and off southern California, in 57 to 313
fathoms.

Specimens examined.—Forty-six, from the following stations:

Specimens of Peribolaster biserialis examined.

=

Station. Locality. Depth. Nature of bottom. umber, Collection.
wa ashe Sa an fae
Fathoms. |

2927....... Off San Diego, California............. 313 Ree UOC co csetees.ccnsdeccsanse 1 U.S, Nat. Mus.

BS17. co. North of Unalaska................... 165 coarse sand, gravel, rocks.......... 1 Do.

aaa Between San Diego and San Clem- | 264-285 | fine gray sand..................... 2 | Albatross, 1904.
ente Island. }

4410.......| Between Santa Catalina and Santa | 178-195 | gray sand, gravel.................. 2 Do.
Barbara Islands.

4421.......| Between Santa Barbara and San | 291-298 | gray mud, rocks................... | 7 | Do.
Nicholas Islands. | |

4788.......| Off Medni, Commander Islands... ihn NBER SATE cs san ee oes oe 3 | Albatross, 1906.

4790:......|) Off. Bering Taland . . 5... ..0.025..0-5 4 ADDIE as oe cans an <cnacnacksn eames 3 Do.

WIG e akan ata one TOs sone iak ace vanpasacasesasecue sn 76 POOKY 0 o sires donee denen ciciple staan naw 17 Do.

BiWasedac cal toess WDD Mia Saints ues eaaew aay acon ax 4 2 DRUDIES ccramedwacacuncisssdscenauaa 10 Do.

Remarks.—One other species of this genus is known, P. folliculatus Sladen, from
45 fathoms off Chile (46° 53’ S.; 75°12’ W.). The more striking differences between
the two forms are the arrangement of the tube feet in two series in biserialis (four
in folliculatus), its more open abactinal skeleton, fewer plates, longer paxillar
spinules, and less numerous actinal spines (four instead of six in each transverse
row). A prominent furrow spinelet is present in biserialis, the mouth plates have
four marginal and no suboral spines, the inferomarginal plates may be easily
distinguished from the adambulacral if dried, and the madreporic body, though
variable in size is relatively smaller than in folliculatus. Judging by Sladen’s figure
the rays of folliculatus are longer and the fascicules of spinelets much more numerous
and compactly placed than in biserialis.

In the arrangement of tube feet, biserialis is not precisely typical of the genus,
but in other characters the two known species are essentially similar.

Family PTERASTERID.® Perrier, 1875.

Spinulosa with cruciform or lobed plates bearing paxilliform groups of spinelets
which support a supradorsal canopy or membrane distinct from the abactinal wall;
between the two a nidamental cavity is formed, opening centrally by a valved
aperture, the osculum; actinolateral spines (arising from adambulacral plates) pre-
sent supporting an actinal web, or merged in actinal floor; external to adambulacral
plates, a series of segmental apertures all along ray, opening into nidamental cavity
and guarded by a specialized opercular spinelet or papilla; no actinal intermediate
plates; mouth plates broad and plowshare-shaped; membranous interradial septa.

344 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

KEY TO THE KNOWN GENERA OF PTERASTERID#.

a’. Armature of adambulacral plates forming transverse combs, the spinelets united by a web.
b!. Supradorsal membrane with muscular fibrous bands; actinolateral spines not completely merged
in actinal floor, but the webbed tips forming a free independent ventro-lateral fringe.

cl. Muscular bands of supradorsal membrane not conspicuous; usually not regularly reticulated
and circumscribing the spiracula, but meandering or crisscrossing, usually very inconspicuous;
sometimes spicules are present in supradorsal membrane . Pteraster ¢ Millerand Troschel, p. 344.

c. Muscle bands conspicuous, regularly reticulated; no supradorsal spicules.
d', Adambulacral plates of two kinds, less prominent alternating with more prominent, and
correlated with them two sorts of furrow combs, one longer and set farther into furrow, the
other set back slightly; actinolateral membrane broad and fleshy; tube feet in four

Wh ea nbe oc sooceoasedscosoonsmenscccosmUnessoncosecscaqes Diplopteraster Verrill, p. 370.
d@, Adambulacral plates not of two kinds; tube feet in two rows; actinolateral membrane
THATTO Were onc oteve vote raisto lave te tore ic eietete ate avers ate ts Teleta slatora te Sra ane te Tatas eer eee Retaster Perrier.

b?. No muscular fibrous bands in supradorsal membrane; actinolateral spines merged in actinal floor;
no lateral fringe.

c!. Paxille spinelets (15 to 30), long and hair-like, protruding freely through the membrane; adam-
bulacral armature partly horizontal in disposition; one pair of suboral spines united by web
tormarpinal mouthispinessjec eee see jee eetel mele et eetaleeieisiaie ele ate eer eer Marsipaster Sladen.

c. Paxille spinelets (five or six) short, robust, not protruding; adambulacral armature perpen-
dicular in disposition; three pairs of suboral spines, free and independent. . Calyptraster Sladen.

a?, Armature of adambulacral plates not forming transverse webbed combs; spinelets free.

b!. Nidamental cavity spacious; supradorsal membrane well developed; spinelets of paxille support-
ing the membrane, not protruding through; muscle fibres of supradorsal membrane well devel-
oped; spiracula present.

c!. Mouth plates large with suboral spines; actinolateral spines long, projecting beyond border
OL WAY ope ce acein e ciets sions sinais een tei la alelateio stots tne eee nie ett nator Hymenaster Thomson, p. 373.

c?, Mouth plates small, without suboral spines; actinolateral spines short not reaching to interra-
dial line; supradorsal membrane regularly reticulated, each mesh containing a spiracle.

Cryptaster Perrier.
b?. Nidamental cavity small, the spines projecting through the thin supradorsal membrane which
lacks or has only very rudimentary muscle fibres...............------+---- Benthaster Sladen.

Genus PTERASTER Miller and Troschel.

Pteraster MULLER and TroscHEL, System der Asteriden, 1842, p. 128. Type, Asterias militaris
O. F. Miiller.

Hexaster PERRIER, Comptes rendus, vol. 112, No. 21, May 1, 1891, p. 1227; Mém. soc. zool.
France, vol. 4, 1891, p. 267. Type, Hexaster obscurus Perrier.

Temnaster VERRILL (subgenus), Proc. U. 8. Nat. Mus., vol. 17, 1894, p. 275; Amer. Jour. Sci.
vol. 49, 1905, p. 202 (as genus). Type, Pteraster (Temnaster) hexactis Verrill=Pteraster
obscurus (Perrier).

Pterasterides VerritL, Amer. Naturalist, vol. 43, Sept., 1909, p. 547, footnote. Type, Pteraster
aporus Ludwig.

Diagnosis.—P terasteridee having muscle-bands in the supradorsal membrane,
but these as a rule not regularly reticulated; adambulacral armature in the form of
transverse webbed combs; actinolateral spines forming a free independent lateral
fringe, not merged in actinal floor; supradorsal membrane frequently with small
spicules (but not invariably).

@ Lophopteraster Verrill (Amer. Journ. Sci., vol. 49, 1895, p. 202) is stated to differ from Pteraster ‘‘in
having a very prominent, solid crest or keel-like prominence on the center of each jaw; it forms the
inner angle of the jaw, separating the two groups of oral spines in the middle.”’ Otherwise it is like
Pteraster. Type, Pteraster abyssorum Verrill. This may be a good genus, but it seems rather slightly
differentiated from Pteraster, the shape of the mouth plates constituting the only difference.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER, 345

KEY TO THE SPECIES OF PTERASTER HEREIN DESCRIBED.@

a‘. Rays constantly five.
b'. Formstellate; R exceeding 1.8 r, usually 2 r or more.
c!, Supradorsal membrane reticulated, very thin and delicate; spiracvla six to ten to an area;
GY ON HINER RWO sats cas cen oa meni aa ds c)8 dam aw wma wisn bJalerd nang Melanin ecincine naw gracilis, p. 349.
c?, Supradorsal membrane not reticulated.

d'. Tube feet in two series; supradorsal membrane with spiny spicules; oral spines webbed;
suboral stout.

e’, R=2 to 2.5r; adambulacral comb with more than five spines; oral spines six to eight; sub-
oral spine cylindrical, not three edged; paxilla stalk low with three to five spines.

militaris, p. 346.

e?. R=1.87 r; adambulacral comb with five spines; oral spines five; suboral very large, three-

edged; paxilla stalk high with six spines.................----------- trigonodon, p. 348.

d*. Tube feet crowded, in four series, at least proximally; supradorsal membrane either without
deposits or with a very few simple rods; oral spines partly or wholly webbed; suboral
slender.

e!, R=2r; adambulacral comb of three or four spines; oral spines four; no small lateral pockets
back of aperture papillz; no deposits in supradorsal membrane; paxilla spines usually
Pwo; LODPEr ApITACLIa PresenkUakio see oc sey see cue ss. -eaidam ons aa einen jordani, p. 350.
e?. R=1.8 to 2.5 r; adambulacral comb of five spines proximally; oral spines five; small shallow
pockets just back of aperture papilla; supradorsal membrane with few scattered simple
rods; paxilla spines three, shorter; spiracula absent .........-.-..-..-- marsippus, p. 352.
b?. Form more or less pentagonal, R less than 1.8 r.

cl. Oral spines six, free, suboral spine very slender; spiracula very numerous; lateral fringe defin-
ing ambitus; no spicules in supradorsal membrane; R=1.4r........-.. coscinopeplus, p. 354.

c. Oral spines five to seven, the entire ten to fourteen united by a continuous membrane.

d'. Size large; membranes thick; supradorsal membrane more or less reticulated by muscle bands,
or else creased by fine furrows into which open the spiracula; actinolateral membrane very
narrow. R=1.44 tol.8r.

e'. With well developed rays; reticulations usually evident at least on sides of body. R=1.6

HOMES ee ated a Cet eee pele cunt esleee coe aWensenl raw ecdsw eaves. 3 tesselatus, p. 359.
é. More arcuately pentagonal in form. R=1.44 r; no evident reticulations; supradorsal mem-
brane with many fine irregular disconnected creases. ....-.....---.----- arcuatus, p. 363.

@. Size medium or small; membrane rather thin, often translucent. R=1.3 r; deposits present
or absent; abactinal surface swollen and more or less warty in adult.

e!. Paxille with high pedicel and five to fifteen (five to seven in Alaskan specimens) spines;
suboral spine slender cylindrical; muscle bands connecting tips of paxillar spinelets;
aperture slits small; oral spines six or seven; no deposits in supradorsal membrane;
lateral fringe not defining ambitus............-..-------+-++-+-+-+--- pulvillus, p. 358.

e. Paxille low, with low pedicel and five to seven short spines; suboral spine with three sharp
edges; no muscle bands connecting tips of paxillar spinelets; aperture slits unusually
large; oral spines five; in supradorsal membrane deposits in the form of branched rods;

lateral fringe defining ambitus.............-.-.----+-+-+--++--++-- temnochiton, p. 355.
3. Oral spines three, each series independently webbed; abactinal surface not much elevated and
MO’ BEA ALL WELOV sek « cas na oan Snwwe Pom Aem ene side Ov UN eswennaneneseneess multispinus, p. 359.

a?, Rays six, seven, or eight, very rarely nine; lateral fringe when evident narrow, defining ambitus;
supradorsal membrane, actinolateral membrane, and webs heavy, spiracula numerous; abactinal
surface rough and spiny; adambulacral spines four to seven; oral spines five or six, completely
webbed. R=1.4 to 1.65 7.2... 2.22.2 c eee eee eee cence ee een eee cent ee enceeeneces obscurus, p. 363.

a | have not included in the key Verrill’s Pteraster hebes (1909, p. 61), a8 Tam not sure that it is a
Pteraster. Professor Verrill has sent me two views showing the lateral and abactinal surfaces. The
creature greatly resembles a Diplopteraster. The type lacked mouth plates, and the presence of muscle
bands and spiracula in the supradorsal membrane is not indicated. If it is a Diplopteraster it would
seem to be distinct from multipes. (Departure Bay, British Columbia, 23 fathoms, mud and sand.)

346 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

PTERASTER MILITARIS (O. F. Miiller).
Pl. 98, figs. 1, 2.

Asterias militaris O. F. MULLER, Zoologize Danicee Prodromus, 1776, p. 234.

Asteriscus militaris MULLER and TroscHEL, System der Asteriden, 1842, p. 44.

Pteraster militaris MiLLER and TroscHEL, System der Asteriden, 1842, p. 128, pl. 6, fig. la and
b.—Fisner, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 167.

Pteraster aporus Lupwie, Echinodermen des Beringsmeeres, Zool. Jahrb., Abth. f. Syst., vol. 1,
1886, p. 293; Fauna Arctica, vol. 1, 1900, p. 470, footnote —CiarxK, Bull. Mus. Comp. Zodl.,
vol. 51, No. 11, 1908, p. 287 (key).

For an extensive list of references (none referring to the region covered by this report), see
Ludwig, Arktische Seesterne. Fauna Arctica, vol. 1, p. 469.

Diagnosis.—Rays five. R=2 to 2.5 r. Form stellate, rays evenly tapered,
often slender at tip; supradorsal membrane fairly thick, without muscle bands,
crowded with calcareous deposits in the form of minute very spiny fenestrated or
reticulate bodies; spiracula few, fairly large, scattered whole length of ray abac-
tinally and on sides; pedicels of pseudopaxille low, slightly higher than broad, with
three to five slender spines. Adambulacral plates, proximally, with a transverse
webbed comb of seven to nine spines (according to size of specimen), the free edge
of web being slightly thickened, undulating, and continued outward as a ridge on
the actinolateral membrane. Mouth plates with six to eight webbed marginal, and
a single stout suboral spine, sheathed in membrane, and with a clear glassy tip.
Actinolateral membrane narrow with a free margin.

Notes on North Pacific specimens.—An example, station 2846, has the following
measurements: R=75 mm.; r=33 mm.; R=2.2+r; breadth of ray at base, 38
mm. A specimen from station 3445 is about the same (large) size.

The calcareous deposits in the supradorsal membrane can be readily seen with
a strong glass, without clearmg. The membrane itself is fairly thick, especially in
the large example from station 2846, and is wrinkled more or less. The spines form
only slight protuberances. The adambulacral combs are very regular, and the man-
ner in which their web is continued laterally across the actinolateral membrane is
characteristic of all the Pacific specimens. In very small examples this continua-
tion is for only a short distance toward the margin, but in the large ones it forms a
ridge nearly to the border. In three specimens accidentally dried these ridges are
largely obliterated. The inner adambulacral spine is very short, the next one or
two successively longer, so that the third or fourth is nearly as long as the longest,
which are the following four or five. The large specimens have eight or nine adambu-
lacral spines, while the small ones have about seven on the proximal half of ray.
None of the spines have a terminal membranous sacculus as in Pt. jordani, but the
edge of the web is, in life, slightly thickened and emarginated between the spines
as in Pt. stellifer Sladen. The only Atlantic specimen of militaris available (from
Bay of Fundy) has stouter adambulacral spines, six or seven in number, but the
specimen is small (R=30mm.). The mouth plates have six to eight slender gradu-
ated webbed marginal spines, but the companion combs are not joined. The long
inner spine is longer than the median suture of jaw plates. The suboral spine
is clavate, pointed, very much heavier than any other spine, has a glassy distal
part, and is normally sheathed in membrane. In the largest specimen this spine

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 347

has a length of 7 mm., and equals the inner marginal spine. It will be seen that
there may be two more mouth spines than is usual in smaller Atlantic specimens.
However, a specimen (R=35 mm.), from station 4772 has six mouth spines as also
one from Kadiak (R=18 mm.) while a young example from station 2857 (R = 10
mm.) has five adambulacral and five mouth spines. The ridges of tissue continu-
ing the furrow webs onto the actinolateral membrane are less developed in the
Bay of Fundy example, but the North Pacific specimens vary somewhat in this
character, which Sladen emphasizes for his variety prolata.

Type-locality— Norway (not formally stated in the Zoologie Danicw Prodro-
mus, but Denmark and Norway are to be understood; the species does not occur
off Denmark).

Distribution —This circumpolar species, heretofore not certainly known from
this region,? evidently has, like so many other forms, invaded the North Pacific
via Bering Strait. It ranges south to Washington, and from shallow water to
344 fathoms. Ludwig (Fauna Arctica, vol. 1, p. 470) summarizes the distribution
as follows: East coast of North America from Cape Cod to the Gulf of St. Law-
rence, north through Davis Strait and Smith Sound to Grinnell Land, reaching the
most northerly station in lat. 79° 40’ N.; Greenland, Jan Mayen, Spitzbergen,
Edgeland, and west of West Spitzbergen; south to Faroe Channel, and the
Norwegian coast (Finmark to Bémmelfjord); eastward from Finmark to Barents
Sea, Murman ccast, and Kara Sea to Cape Tscheljuskin (longitude 116° E.).
From here it evidently ranges to Bering Strait, although there are no records.

Specimens examined.—Nine, from the following stations:

Specimens of Pteraster militaris examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
2846...... South of Unimak Island, Alaska...... 44 BIRVOM ha. Aenea ad. scceecteuenedes 1 U. S. Nat. Mus.
SeUt ake Near Afognak Island, Alaska......... 51 gray sand, broken shells........... 1 Do.
Pt Near Unimak Island, Alaska......... | 41 | black sand.............. ines 1 | Do.
3445...... Straits of Fuca, Washington .........| 100 XOUK Vidiewwusecutsasweccenscuunus dws 3 Do.
Madiak Island-sses..-e-.--cecss-e 8 favre eid een ee een tutes vod react enaecee oe Do.
778s cass Bowers Bank, Bering Sea, north of 344 green mud, brown sand........... 1 | Albatross, 1906.
Andreanof Island. } |
4780...... Near Agattu Island, Aleutians. ....... 70 | green sand, pebbles............... 1 Do.
|

Remarks.—The presence of this species in the North Pacific is not surprising
when considered in connection with Diplopteraster multipes, Pseudarchaster parelii,
Henricia sanguinolenta, Leptychaster arcticus, and others.

I think there is little doubt that Pteraster aporus Ludwig is an abnormal example
of militaris. In all respects, except in the absence of the oscular orifice, the deserip-
tion given by Ludwig admirably fits militaris. Ludwig's specimen came from
Bering Sea.

@ Ludwig, Fauna Arctica, vol. 1, p. 470, states that Pt. aporus is perhaps a synonym of militaris.

348 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
PTERASTER TRIGONODON Fisher.
Pl. 99, figs. 1, 2; pl. 115, fig. 2; pl. 116, figs. 2, 2a.
Pteraster trigonodon Fisuer, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 168.

Diagnosis.—Rays five. R=30 mm.; r=16 mm.; R=1.87 r. Breadth of ray
at base 18 mm. General form stellate with angular interbrachial ares. Disk mod-
erately high; vertical diameter about two-thirds breadth of ray. Supradorsal mem-
brane thin, without definite reticulations; calcareous deposits abundant, in the form
of irregular spiny rods; spiracula numerous, large; pseudopaxille with high
pedicels and six divergent spines; five adambulacral and five webbed marginal
mouth spines; aperture papilla very broad; suboral spine very large, sharp, and
three-edged; actinolateral membrane narrow, not defining ambitus.

Description.—Supradorsal membrane thin and translucent, the minute calca-
reous bodies showing under a moderate magnification; no regular reticulations,
but fine muscle fibers proceed radially from the tips of the abactinal spines which
push up the supradorsal membrane, giving the surface a papillose appearance.
Spiracula conspicuous, numerous, but not uniformly distributed, although occur-
ring everywhere on the abactinal and lateral surfaces. Some of them seem to
have been enlarged by the stretching of the membrane. Here and there six or
eight form a circle around the central spine of a paxilla. Membrane contains a
multitude of minute calcareous bodies in the form of very irregular branched spiny
rods, fenestrated spiny bodies, or irregular plates with one or two perforations
and long spines attached. Each pseudopaxilla consists of an elevated slender
pedicel about 2 mm. high, surmounted by a central spine and a circle of five around
it, all slightly shorter than the pedicel, and forming the sight protuberances on the
supradorsal membrane. Osculum large.

Ambulacra narrow, feet in two rows. Adambulacral plates with a curved
transverse fan of five spines, the innermost shortest (one-half or two-thirds the
longest), thence increasing in length to the outermost, which is usually longest
and situated close to the actinolateral membrane. As in many species of the genus,
the inner end of the series curves aborally. There may be four spines, especially
far along ray. The web is deeply emarginate between the spines, and does not
extend laterally on the actinolateral membrane. Aperture papilla wider than
usual in proportion to length and with a very convex free aboral edge. Aperture
at base of outer comb spine and about one-half its length.

Mouth plates rather narrow, each with a prominent central elevation or boss
for the large actinal or suboral spine. This is one and one-half to nearly twice
as long as the longest marginal spine, sharp, very heavy, and with three wide lateral
flanges or edges, so that the spine has three concave faces, appearing like a minia-
ture spear-head. Marginal spines five, graduated toward the innermost, which is
about as long as the interradial dimension of plate; all are united by membrane,
but companion combs are independent.

Actinolateral membrane narrow, with a narrow free border which does not
define margin of ray when viewed from either surface. Actinolateral spines gradu-
ally decreasing in length; proximally about twice as long as outer adambulacral
spine.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 349

Type.—Cat. No. 27790 U.S.N.M.

Type-locality — Albatross station 4427, off Santa Cruz Island, California, 447
to 510 fathoms, black mud, rocks.

Distribution —Known only from type-locality.

Specimen examined.—The type.

Remarks.—This species, which stands about halfway between those which are
stellate and those which are more or less pentagonal, bears superficially more resem-
blance to militaris than to any other form, especially as there are spiny often
fenestrated calcareous bodies in the supradorsal membrane. The differences are
important, however. In trigonodon the stalk of each paxilla is high (not low) and
crowned by six (not three to five) spines shorter than pedicel; supradorsal mem-
brane is thin in trigonodon, and the deposits are not so crowded. The adambulacral
and marginal mouth spines are fewer (five, not six to eight), and the suboral spine
is very large and of a special three-edged form, not cylindrical, as in militaris.

PTERASTER GRACILIS (Clark).

Retaster gracilis CLARK, Proc. Boston Soc. Nat. Hist., vol. 29, No. 15, May, 1901, p. 326, pl. 3, figs.
3 and 4.

The status of this species is somewhat doubtful. The two specimens are
probably young, and, coming as they do from Puget Sound, might be considered
as immature tesselatus. This form, however, differs from young tesselatus in having
a very thin and delicate supradorsal membrane, longer rays (R=2 r), and only
two oral spines. The spiracula are fewer.

Diagnosis—Rays five. R=18 mm.; r=9 mm.; R=2r. Breadth of ray at
base equals r. Disk depressed, 6 to 8 mm. high. Supradorsal membrane very
thin and delicate; tips of spines from the edges of the paxille connected by delicate
lines giving abactinal surface a reticulated appearance; no spicules; spiracula
quite numerous, six to ten in each mesh of reticulum; paxille high, with six or more
(often many more) spines surrounding edge, and several smaller ones erect in
middle; oscular orifice small. Adambulacral plates with five webbed spines
arranged in a row at right angles to furrow. Of these the first is very small, the
second larger, the third still larger, the fourth still larger, the fifth equals the third.
Actinolateral spine slightly longer than longest adambulacral; actinolateral mem-
brane with a free margin, narrow. Aperature papille very small, free only on one
side. Mouth plates prominent, each with two marginal spines united by a mem-
brane, and a very large free spine on oral surface near base, pointing downward and
outward. Ambulacra broad; feet comparatively few, in two rows. Color, light
brownish gray, darkest on disk, lightest near tip of rays. :

Type.—Said to have been sent to Columbia University, New York; not to be
found, 1910.

Type-locality— Probably Puget Sound, Washington.

Remarks.—This species is known only from the two specimens described by
Clark. Even the locality is doubtful, inasmuch as the specimens were not labeled,
and the collection contained several species evidently from some locality south of
the United States.

850 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
PTERASTER JORDANI Fisher.
Pl. 100, fig 2; pl. 101, fig. 1; pl. 117, figs. 1, la-b.

Pteraster jordani Fisuer, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 314; Ann. and
Mag. Nat. Hist., ser. 8. vol. 5, Feb., 1910, p. 67.—Ctarx, Bull. Mus. Comp. Zoél., vol. 51,
No. 11, April, 1908, p. 286;

Diagnosis.—Similar in general form to Pteraster militaris, but supradorsal
membrane lacking spiny calcareous bodies, tube feet in four instead of two rows,
only three or four adambulacral spines, four mouth spines, and two (one to three)
paxillar spines. R=2r. ;

Description.—Rays five. R=75 mm.; r=37 mm.; R=2 r. Breadth of ray
at base, 37 to 42 mm. Thickness of ray at base, about 28 mm. (type).

Form stellate, depressed; abactinal surface of rays and disk convex, actinal
surface subplane; edges of ray rounded; rays tapering, with fairly straight sides,
bluntly pointed, though sometimes swollen near tip and the extreme tip recurved,
giving an entirely different appearance; interbrachial angles obtuse; a well devel-
oped lateral fringe present, though actinolateral membrane is narrow.

Supradorsal membrane rather thin, with no deposits of calcareous matter;
paxillar spines very prominent; two spines to each pseudopaxilla (sometimes one
or three), one considerably longer than the other, protruding 3 or 4 mm. above
general level of the membrane, but carrying the latter with it; pedicel very low,
and the spine correspondingly rather longer than usual in this genus; fine criss- -
crossing muscle fibers exceedingly abundant, and the summits of the paxillar
spines connected by faint muscular bands, which are nearly invisible in the type-
specimen but show with some distinctness in a smaller example; in the irregular
meshes thus formed one to three large spiracula, these often absent; oscular orifice
rather small, though variable in size; valves with a much higher and stouter pedicel
than other pseudopaxille, and with about fourteen spines in two series on top of
pedicel, all united by single membrane. The pedicel is about 2 mm. high in the
midradial region of the type, the entire pseudopaxilla being about 11 to 12 mm.
high. The papulze extend about 5 mm. up the pseudopaxilla and have numerous
short lobes at the top.

Adambulacral plates with a transverse series of three spines and one spinelet,
united by a web; inner member of the series quite small and situated slightly
aborad from the others; the next three are two to five times longer, and increase
slightly in size outward, being united by membrane nearly to their tips, which are
capped by a membranous sacculus; outer spine close to the corresponding actino-
lateral spine, to which it is united by membrane. Aperture papille prominent,
jawbone-shaped, free on the aboral margin. On a smaller specimen the inner small
spinelet is absent, and in the type it is sometimes absent, especially on plates beyond
the middle of the ray. A specimen from off the coast of Washington also has three
spines on most of the plates. Thus the combs may contain three long and one
shorter spinelet or three long subequal spinelets only. The aperture papilla is
not tented over by an auxiliary membrane, forming a special pocket between each
pair of adambulacral combs. (See next species.)

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 351

Armature of mouth plates consisting of a marginal series of four spines and
spinelets, the inner long and slender, the next nearly as long, and the outer two
much shorter; inner three united by membrane, and usually also the whole four,
but the outermost, which is shortest, sometimes stands alone; on the actinal surface
commonly one long, slender spine about the size of the inner marginal, with which
it stands in a linear series, directed over the outer end of the plate; this spine entirely
free from the others. Sometimes there are only three marginal spines. The outer
end of the combined mouth plates forms a peak in the center, the median suture
sloping upward at a rather steep incline.

Actinolateral membrane narrow, only a trifle wider than furrow at base; spines
short, the ninth or tenth the longest; for a considerable distance along ray they
shorten only slightly or not at all. Membrane rather thin, forming a lateral fringe
with an undulating edge. In some states of contraction this edge is free from the
supradorsal membrane for one-third or one-fourth of its width.

Color in life, ‘‘dragon’s blood”’ red.

Anatomical notes.—Hepatic ceca relatively enormous, extending to tip of ray,
the interior a very spacious cavity from which food is probably absorbed. Between
the large dorsal stomach and the anus is a spacious saclike intestine with wrinkled
walls which probably function as the intestinal caeca of other forms. There is a
powerful sphincter between the dorsal stomach and intestine. Dorsal stomach
. Spacious with large dorso-ventral slits opening into each of the ten hepatic ceca,
which are connected with the first ambulacral ossicles by strong bands of muscle,
probably to draw the ceca toward the mouth when the stomach is thrown out.
The first ambulacral ossicle is very large and arched. To it are attached along the
inner face numerous muscles apparently protractors and retractors of the ventral
or eversible stomach. The first tube foot has a single large ampulla, the succeeding
tube-feet have two. The upper ends of the ambulacral ossicles are strongly imbri-
cated and are very flaring. Tube feet with large sucking disks, in two zigzag or
virtually four series; alternate pores are placed nearer tie adambulacral plates so
that the feet are really four ranked. No Polian vesicles; madreporic canal rather
long and sinuous. Interradial septa of tough membrane, attached to either side of
which is the gonad, forming a large much branched tuft. Gonads packed with
sperm.

Abactinal plates with four long slender arms widening at the end. These
overlap the lobes of adjacent plates and form square papular areas.

Type.—Cat. No. 22343, U.S. N. M.

Type-locality.— Albatross station 4354, off San Diego, California, 646 to 650
fathoms, green mud.

Distribution.—From northern Lower California to Washington, in 266 to 984
fathoms, on mud or sand.

352 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Specimens examined.—Eleven, from the following stations:

Specimens of Pteraster jordani examined.

Station. Locality. | Depth. Nature of bottom. Number. Collection.
Fathoms.
2019...... Near Los Coronados Islands, Lower | os4 2 U. 8. Nat. Mus.
California. |
Off Santa Cruz Island, California... .. 266 1 Do.
Off Sea Lion Rock, Washington... .. | 859 1 Do.
Off Half Moon Bay, California........ | 301 1 Do.
.| Off Tillamook Bay, Oregon ...-...- Al EG 1 Do.
-| Near San Diego, California............ | 646-650 |..... LO rete te ea tetataas ee tere ater etetereraiatate 1 Albatross, 1904.
Between San Clemente and Santa | rG5aenll|bazee CO Selec s nee ees eee eee 1 Do.
Catalina Islands, California. |
44125 .cee Between Santa Catalina and Santa Bite ee sae COCs cee eee eee ere 1 Do.
Barbara Islands, California.
BANDS oaecn|aimacte CO Neeseocnae aaneeriee ee eceemaeee ne G3Scai ees LO Se eyalewretatetsicie cee eee eine reattoars 1 Do.
dao} <a59: Between Santa Barbara and San| 291 |..... OS paceeteacten< sees aarti 1 Do.
Nicholas Islands, California.

Remarks.—Pteraster jordani is a well-characterized species, having a rather
thin supradorsal membrane lacking calcareous deposits, but possessing faint muscle
bands connecting the two paxillar spines. The pedicels are very low and the spines
correspondingly long, being six or seven times the length of the former. Tube feet
are crowded in four rows, and the adambulacral combs have only three or four
slender spines, of which the inner is commonly shorter than the rest. From Pteraster
militaris the species is readily distinguishable by the absence of copious supradorsal
calcareous deposits, fewer adambulacral and mouth spines, fewer paxillar spines,
and in having four-ranked instead of two-ranked tube feet.

Attention should be called to the fact that the mdividual dissected was a
male and that it was not different from the rest. The suggestion has been made?
that in the Pterasteride the male may be some entirely different looking animal,
perhaps now described as a different species. Such is certainly not the case.

PTERASTER MARSIPPUS Fisher.
Pl. 100, fig. 1; pl. 101, fig. 2; pl. 115, fig. 4; pl. 116, fig. 1.
Pteraster marsippus FisHer, Ann. and Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 168.

Diagnosis.—In general appearance and proportions similar to jordani; differs
in having three shorter paxillar spines; sparse calcareous deposits in the form of
simple straight rods of small size in supradorsal membrane; proximally five adam-
bulacral spines (far along ray only four) and especially in having the aperture
papilla partially tented over by a membrane passing from one comb to the next,
and merging outwardly with the actinolateral membrane. A shallow side pocket
is thus formed at the mouth of which is the aperture papilla. No spiracles.
Type, R=100 mm.; R=40 mm.; R=2.5 r.. Breadth of ray at base, 45 mm. R
varies to 1.8 r.

@Lang, Text-book of Comp. Anat. (English translation), Pt. 2, 1896, p. 503.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 3853

Description.—This form is evidently a northern representative of Pteraster
jordani, and without careful examination the two might be easily confused. The
paxillar spines are not, however, so long as in jordani, although the pseudopaxille
have the same character. The spines are 7 or 8 mm. long, and one is commonly
slenderer than the other two; pedicel about 1 to 1.25 mm. high. In the alcoholic
specimens the supradorsal membrane is smooth, fairly thick, and few spines pro-
trude through. At any rate, it has not a rough, swollen, or warty appearance. A
piece of the membrane cleared shows in scattered places a few minute straight or
slightly curved, sometimes incipiently branched, rods, wholly unlike the crowded
spiny deposits of militaris. Spiracles so few or inconspicuous that none can be
detected. The supradorsal membrane is decidedly thicker than in jordani. Oscu-
lum as in jordani, each valve with about fourteen spines in two rows.

Proximally the adambulacral plates have five long, slender spines heavily
sheathed and capped by a sacculus, and united by a membrane which joins the
actinolateral membrane usually beyond the end of the outermost spine. The outer
part of successive webs is joined by a slight horizontal membrane (not present in
jordant), which forms a sort of lateral cul-de-sac at the mouth of which is the
aperture and its papilla (similar to those of jordani). Far along the ray each
comb has four spines; proximally the first spine is sometimes much shorter than
the second, sometimes nearly as long. On account of the four-ranked tube feet,
alternate adambulacral plates and armature are crowded back just a trifle. In
Diplopteraster this crowding is carried farther, and the alternate adambulacrals
have a less number of spinelets.

Mouth plates similar in form to those of jordani; marginal spines five, united
by membrane, rapidly graduated in length from the long (10 mm.) inner to the
short (2.5 to 3 mm.) outer; suboral spine slender, as long or longer than inner
marginal, just behind which it stands.

Actinolateral membrane narrow and similar to that of jordani.

Type.—Cat. No. 27791, U.S.N.M.

Type-locality.— Albatross station 4784, near Attu Island, Aleutians, 135 fath-
oms, coarse pebbles.

Distribution.— Bering Sea, 52 to 351 fathoms, on gravel.

Specimens examined.—Twenty, from the following stations:
Specimens of Pteraster marsippus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
| Fathoms.

9 sass: Near Unalaska.....-.....2s-.sesee0e. BR Sell SODDIML EE cy egsccetecntn adic ‘ 1 U.S. Nat. Mus.
SER ed nce einstein eta tea 56 | black pebbles.............. Te Do.
$486.5 5..5 West of Pribilof Islands.............. 150 green mud, fine sand.... 1 Do.
$330.....4 INGEY UDMIORRR.« chen << scntccruduses 351 | black sand, mud...... sabe 3 Do.
3500...... South of Pribilof Islands............. 121 fine gray sand, gravel..... ; le | Do,
4779. c= 00 Near Semisopochnoi Island, Aleutians. 52 fine QrAVEl a ccee see seses svesenaccs 2 | Albatross, 1906.
4784...... Near Attu Island, Aleutians.......... | 135 | coarse pebbles............. aaocae 3 Do.
4702...... Near Bering Island, Commander Is- | 72 | pebbles...............200.0.000045 | 5 Do.

lands.

a Two others probably from this station; label lost during earthquake 1906.
57444°—Bull. 76—11 23

354 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Remarks.—Pteraster marsippus differs from jordani in a number of details
mentioned in the diagnosis. The calcareous microscopic rods in the supradorsal
membrane are not found in jordani. In young jordani the spiracula are readily
seen, while in marsippus they can not be detected. Neither have I been able
surely to make them out in the adult. The additional mouth and adambulacral
spine seems to be a constant difference, and especially the little side pocket just
back of the aperture papilla, between any two fans of spines. This disappears in
dried specimens. Unique examples of Pteraster should never be dried, as many
valuable characters of the membranes are obliterated, never to be restored. A
dried Pteraster is comparable to a dried fish; neither is a satisfactory verisimilitude
of the live object.

From Pteraster militaris, which dwells in the same region and is similar in
proportions, the present species differs in having four-ranked tube feet (except
young), in lacking the crowded deposits in the supradorsal membrane, and in
having fewer adambulacral and mouth spines. The actinal mouth spine is slen-
derer than in militaris, which also does not possess the little pockets between the
outer ends of the adambulacral combs.

PTERASTER COSCINOPEPLUS Fisher.
Pl. 102, figs. 1, 3; pl. 116, fig. 4.
Pteraster coscinopeplus Fisher, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 169.

Diagnosis.—Stellato-pentagonal. R = 30 mm.; r= 21 mm.; R=1.4+4r.
Breadth of ray, omitting lateral fringe, 22 mm.; with lateral fringe, 25mm. Supra-
dorsal membrane without special muscle bands connecting summits of spines, but
with a fine reticulation; spiracula very numerous, small; no spicules; paxille with
high pedicels and six to eight spines; actinolateral membrane very wide, the free
border defining entire ambitus; six webbed adambulacral spinelets; one very slen-
der suboral spine; six slender free marginal mouth spines.

Description.—Abactinal surface covered with protuberances, the ends of the
paxillar spines. Membrane moderately thick, slightly translucent, without coarse
muscular reticulations, and without calcareous deposits, but with a fine reticulum
of muscle fibers, the tiny not well-defined meshes of which inclose the small spi-
racula, one to a mesh. The fine fibers radiate from the summits of the spines,
crisscrossing in every direction. Spiracula very numerous and small, occurring in
short rows between the spines, or radiating from the central spine of a group or in
a circle about a spine; much more numerous than in trigonodon. Pedicels of
pseudopaxille relatively high (3 mm.), of about the same proportions as in trigono-
don, and much higher than in militaris, jordani, and marsippus. Spines six to
eight to each pseudopaxilla, subequal, and about as long as the pedicel. Osculum
large, surrounded by many spines.

Ambulacra of medium width; tube feet biserial, not crowded. Adambulacral
spines six, the inner the shortest and set aborally to the rest; the second about
twice as long, the third slightly longer; the next three subequal; the outermost, or
next one, longest. Web slightly emarginated, the spines having a blunt membra-
nous tip; web extends a slight distance beyond end of outermost spine on actino-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 355

lateral membrane. Aperture papilla longer than wide, with slightly convex aboral
border; the aperture is at the base of the outermost adambulacral spine, and is
less than one-third its length, shorter even than the small papilla.

Mouth plates with six slender independent marginal spines (no web), the inner
the longest, and longer than the interradial diameter of plates, the outer very short,
the others evenly and rapidly graduated between. The single suboral spine is
very slender and a little longer than inner marginal spine.

Actinolateral membrane very broad, the conspicuous lateral fringe extending
1 to 3 mm. beyond the border of ray. The spines are longest at about the middle
of ray (seventh to ninth spine); thence they decrease rather evenly to the tip, the
fringe always defining ambitus.

Type.—Cat. No. 27792, U.S.N.M.

Type-locality.— Albatross station 2925 (probably), off San Diego, California,
339 fathoms, mud. The bottle in which this specimen was received was broken
by the earthquake of 1906; the specimen was found beside the above label and
the remnants of the bottle.

Distribution.—Off southern and central California, 287 to 1,062 fathoms.

Specimens examined.—Three; besides the type, one from station 4339, off San
Diego, California, 287 to 369 fathoms, green mud; one from station 4537, off Mon-
terey Bay, 1,062 fathoms, hard sand.

Remarks.—This species is related to Pt. multispinus Clark, Pt. reticulatus
Fisher, and Pt. multiporus Clark, the first from Washington, the second from the
Hawaiian Islands, and the third from Sagami Bay, Japan. From multispinus the
present form differs in having five free slender mouth spines (not three united by
membrane), in having a slender suboral spine, much more numerous spiracula,
fewer paxillar spines (not eight to ten or more), and apparently differently formed
aperture papille (in multispinus ‘‘nearly free on both sides’’). From reticulatus,
coscinopeplus differs in being more pentagonal in form, with a very much wider
actinolateral membrane, and in having slenderer suboral spines (these heavy and
club-shaped in reticulatus), no supradorsal reticulations, fewer paxillar spines,
which consequently form a less closely papillated dorsal surface, and in having
many more spiracula of a smaller size. The adambulacral armature differs chiefly
in size, the fans being larger (and spines longer) in coscinopeplus. From multiporus
this species differs in having a much wider actinolateral membrane (it is narrow in
multiporus), higher paxillze with fewer spines, and in having the mouth spines free,
not the whole twelve (six of each plate) united by a continuous membrane.

PTERASTER TEMNOCHITON Fisher.

Pl. 103, figs. 2, 3; pl. 115, fig. 3; pl. 116, figs. 3, 3a.

Pteraster temnochiton Fisner, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 169.

Diagnosis.—Rays five. R=21 mm.; r=16 mm.; R=1.3+r. Breadth of ray
at base, without lateral fringe, 9 to 11 mm.; with lateral fringe, 12 to 12.5 mm,
General form stellato-pentagonal; interbrachial arcs shallow; lateral fringe defining
ambitus; pseudopaxille low with five to seven spines; no reticulation; groups of
spines independent and spaced; scattered, branched, irregular rods in membrane;

356 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

adambulacral spines five or four, webbed, the outer standing perpendicular, not
subparallel to actinolateral membrane, which is wide; apertures long; aperture
papilla slender, sharp, with slightly convex aboral margin; oral spines five, slender,
all ten united by continuous membrane; suboral spine sharp, three-edged, much
heavier than oral spines.

Description.—Abactinal surface very arched and swollen by reason of the rays
being more or less drawn together ventrally, and the actinal surface correspondingly
concave. Supradorsal membrane rough by the spaced groups of paxillar spines,
these groups of five to seven forming a slight or decided protuberance, the mem-
brane between the groups being smooth and everywhere punctured by very numer-
ous fair-sized spiracles, the whitish sphincters of which clearly define the edges.
Pale muscle fibers meander here and there through the translucent membrane,
but there are no discoverable muscle bands running between the tips of the spines.
It should be added that the spiracles occur also among the spines, but are not more
abundant than in the considerable space between the different groups. In other
words, the groups are not crowded together, but are well enough spaced to be
counted. Unlike pulvillus, the pedicel of each pseudopaxilla is low, scarcely if
any higher than wide, with a crown of five to seven diverging slender spinelets,
of which one commonly stands in the center of a circle of four to six; spinelets
about 2.5 mm. long, and the pedicel about 0.75 mm. high. The whole paxilla is
thus low, so that the supradorsal membrane is not much raised above the abactinal
surface. Paxille with strong lobes (four to six). Membrane with scattered spicules
in the form of crooked rods with irregular branches along the sides, these sometimes
anastomosing and forming loops. There is no regularity in the form of these rods,
which are not crowded as in militaris. They are never in the form of fenestrated
bodies and are more like those of trigonodon than militaris, though simpler, but
yet not so simple as those of marsippus. The arcade-like space beneath the supra-
dorsal membrane is filled with embryos oriented with either the dorsal or ventral
surface uppermost. The specimen is a female. Osculum rather small.

Ambulacra narrow; feet in two rows. Adambulacral armature consisting of
a series of five (distally four) webbed spines curved aborally on the inner end;
the inner very short, the next nearly twice, the third about two and a half, and
the fourth about three and a half times length of inner, the fifth subequal to fourth.
Membrane deeply emarginate between spines which are not prolonged in a mem-
branous sacculus. A peculiarity of this species is that the outer spine is not close
to the actinolateral membrane, but stands nearly perpendicular, with as much
web between it and the membrane as there is between the fourth and fifth spines.
Aperture papille long, slender, and sharp (about one-half length of outer spine),
situated about half its length laterally from base of outermost spine. The aper-
ture slit is unusually long, extending laterally beyond the edge of the papilla,
especially on outer part of ray. The apertures form a conspicuous row of slits
along the outer edge of the series of adambulacral fans.

Mouth plates with five marginal spines graduated fairly evenly from the long
inner to the very short outer, the whole ten united by a continuous membrane
extending to the tips of the spines. The spines are slender and slightly tapering.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 857

The suboral spine is sharp and glassy, the outer half flaring, with three sharp
edges and three concave faces which can be easily discerned even when the spine
is sheathed in membrane. The shape of this spine, which is situated on a boss,
at about the middle of the plate, is similar to that of trigonodon.

Actinolateral membrane wide, with a wide fringe defining the entire ambitus.
About the seventh actinolateral spine is longest in type; thence they regularly and
rapidly decrease in length toward the tip of ray. When viewed from above, the
membrane projects a conspicuous distance (about one-sixth width of ray) beyond
the edge of the supradorsal membrane. Edge of membrane scalloped; spines
showing through membrane.

Color dirty whitish, or brownish gray in alcohol.

Type.—Cat. No. 27793, U.S.N.M.

Type-locality.— Albatross station 4784, near Attu Island, Aleutians, 135 fathoms,
coarse pebbles.

Distribution Known only from the vicinity of the Aleutian Islands.

Specimens examined.—F rom type-locality, two; from station 3223, near
Unalaska, 56 fathoms, black pebbles, three specimens.

Remarks.—Superficially, at least, this species resembles pulvillus, especially
the abactinal view, the groups of spines forming slightly raised independent spiny
protuberances. The differences are as follows: Low paxillar pedicels with spines
much longer (reverse in pulvillus), no muscle bands between spines (figured in
pulvillus), adambulacral spines fewer (five); aperture slits of a peculiar and charac-
teristic form; mouth spines five; suboral spine not slender but stout and three-
edged. The deposits in the supradorsal membrane may constitute a further differ-
ence (I have not seen Atlantic specimens of pulvillus). From Pteraster obesus Clark
(Sagami Bay, Japan) temnochiton differs in having calcareous spicules in mem-
brane, low paxillw with fewer spines, fewer adambulacral spines (six or seven in
obesus). In contrast to temnochiton, obesus has independent oral spines (five to
seven), short actinolateral spines, so that the fringe is narrow and concealed from
above. Even the suboral spine is different, that of obesus ending in a triangular
point, while in temnochiton the spine is really three-edged for a considerable part
of the length. The aperture papilla is small in obesus, and the aperture is not
unusually large, as in temnochiton. From trigonodon, temnochiton differs in having
a wide, not narrow, actinolateral membrane, differently formed deposits, low,
not high paxille, slender and long, not short and thick aperture papille, in having
the outer adambulacral spine distant from, not close to the actinolateral membrane,
in having much more conspicuous apertures, in having companion combs of oral
spines bound together mesially by a continuous web. The abactinal facies of the
two is quite different; the lateral fringe, for one thing, does not define the ambitus
in trigonodon.

358 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.
PTERASTER PULVILLUS Sars.
Pl. 102, figs. 2, 4.

Pteraster pulvillus M. Sars, Oversigt af Norges Echinodermer, 1861, p. 62, pl. 6, fig. 14-18; pl. 7;
pl. 8; pl. 9, fig. 1-6.—Lupwre, Fauna Arctica, vol. 1, 1900, p. 470 (here is given an exten-
sive list of references, referring to Atlantic and Arctic localities, but not to the region of this
report, where the species has not heretofore been taken).—FisHEeR, Ann. and Mag. Nat. Hist.,
ser. 8, vol. 5, Feb., 1910, p. 168.

Two specimens from the vicinity of Bering Island are referred with some
hesitation to this species. They have the following characters:

Diagnosis.—Rays five. R=21 mm.; r=16.5 mm.; R=1.27 r. Form nearly
pentagonal with breadth of ray at base about equalling minor radius; disk convex,
uniformly roughened by prominent tips of paxillar spines; pseudopaxille with
high pedicel and six, seven, or eight slightly shorter diverging spinelets, the tips of
which are connected in the supradorsal membrane by rather faint muscle bands,
about six radiating from central spine of each paxilla, the meshes thus formed con-
taining one to three papular pores. No spicules in membrane. Osculum small,
ambulacra narrow, feet in two rows. Adambulacral spines six proximally, then
five, then four distally, the innermost much the shortest, the next to outermost
longest; membrane emarginate between tips of spines, aperture papilla rather
narrow and free along aboral convex border. Marginal mouth spines seven (six
in a smaller specimen), slender, tapering, the inner very short; all fourteen united
by a common web to tips; suboral spine a trifle longer than inner oral spine (which
is longer than interradial diameter of plates) slender, glassy, sharp, encased in a
pulpy sheath which makes it appear fairly robust. Actinolateral membrane
narrow, about the fourth spine longest whence the rest gradually decrease in length
to tip of ray, free edge of membrane narrow and slightly emarginate between tips
of spines.

Color in aleohol dull yellowish gray or light yellowish brown.

Type-locality.—Three miles north of Bergen, Norway, 50 to 60 fathoms.

Distribution.—Ludwig (Arktische Seesterne, p. 471) summarizes the distribution
as follows: Maine to Newfoundland (lat. 42° to 48° N.); on the Norwegian coast
from lat. 60° N. to Finmark; Barents Sea; Murman coast; Matotschkin Strait and
Kara Sea; Spitzbergen (to lat. 79° N.). To this must be added Bering Sea.

Specimens examined.—Two, from station 4792, near Bering Island, 72 fathoms,
pebbles.

Remarks.—These two specimens do not seem to be typical, although I have
had only the Latin diagnosis of Sars and a copy of figs. 2 and 9, pl. 7, and fig. 2, pl.
9, as a basis for comparison. According to Doctor Clark’s key these examples
would be pulvillus.

The differences seem to be as follows: fewer paxillar spinelets (eight to fifteen
in the type), a weaker reticulation (according to Sars’ pl. 7, fig. 2 the bands are
quite heavy and the meshes contain one spiracle); more spiracula; the supradorsal
membrane has not a number of humps due to more prominent paxille as shown by
Sars’ figure, but instead the surface is roughened all over by the prominent spines;
the adambulacral spines are six only at the base of furrow near mouth plates; the

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 359

mouth spines are seven. However, it may be stated that the character of the oral
armature, paxille, and supradorsal membrane of the Bering Sea specimens agrees
fairly well with Sars’ diagnosis and figures.

PTERASTER MULTISPINUS Clark.

Pteraster multispinus CLarK, Proc. Bost. Soc. Nat. Hist., vol. 29, No. 15, May, 1901, p. 326, pl.
3, figs. 1 and 2.—Fisner, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 168.

Description.—Rays five. R=27 mm.; r=18 mm.; R=1.5r. Breadth of ray
at base, 17 mm. Interbrachial arcs rather shallow. Disk almost flat, 10 mm.
high, the rays not clearly marked off. Abactinal surface of rays almost flat; actinal
surface flat, edges rounded. Supradorsal membrane thick, stout and apparently
muscular with no signs of reticulations. Spiracula seem to be absent on disk, but
abundant near tips of rays. Paxille high with numerous spines (eight to ten or
more), those in the middle little more slender than those on edge of paxilla. On
account of the abundance of paxille spines, it is almost impossible to determine
whether the membrane contains any other calcareous bodies or not, but there
seems to be none. Oscular orifice large, surrounded by long spines imbedded in
the membrane. Ambulacral furrows broad; feet in two rows. Adambulacral
plates with six spines arranged in a row at right angles to furrow and united by a
delicate membrane. The outermost spine is clearly united to the fifth by this
membrane, and each series of spines is united by this membrane to the edge of the
supradorsal membrane (italics mine). Between each pair of outermost spines is a
small sharp spine, the aperture papilla, which is nearly free on both sides. Mouth
plates prominent, each with three marginal spines united by a delicate membrane,
and with a very large free, superficial spine, placed near the base, and pointing
downward and backward. Color in alcohol, grayish pink-purple. (Clark.)

Remarks.—The type, originally at Columbia University, New York, seems to
have been misplaced. At least its whereabouts was not known in January, 1910.
The specimen was collected by a party from Columbia University, presumably at
Puget Sound, 1897. Nothing further is known of the species.

No mention is made in the description of an actinolateral membrane, which is
therefore probably very narrow. The italicized portion of the above description
would seem to indicate as much. As no lateral fringe is mentioned I am not cer-
tain whether this is present.

PTERASTER TESSELATUS Ives.
Pl. 104, figs. 1-5.
Pteraster tesselatus Ives, Proc. Acad. Nat. Sci. Phila., 1888, p. 421 (Kadiak Island).

Diagnosis.—General form very broadly stellate, as, for instance, R=88 mm.,
r=54 mm., R=1.63 r (station 4222); disk high, rays short and thick, with straight
or swollen sides and recurved blunt tip; interbrachial ares angular broad; supra-
dorsal membrane thick, spongy, marked by reticulations inclosing numerous small
spiracula; spinelets not protruding as a rule; paxilla spinelets eighteen to twenty-
six, the six or seven peripheral stouter, and slightly longer than pedicel; actino-

360 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

lateral membrane narrow, a slight fringe in young examples; adambulacral comb
with five to seven webbed spines, the three or four outer subequal and longest;
five to seven oral spines (usually six), both series united by continuous membrane;
suboral spine heavier and longer than inner marginal oral, but slender; madre-
poric body with a tuft of spines.

Description.—The general form is somewhat variable, but the animal is very
stout, and the abactinal membrane is more or less inflated, the creature resembling
a stellate cushion. Supradorsal membrane thick, tough, and spongy, regularly
reticulated, the quadrate or hexagonal areas in most individuals being better differ-
entiated on sides of ray, and there containing twenty to thirty-five spiracula. No
spicules in supradorsal membrane. Surface of membrane either smooth, or rough-
ened and papillated by the paxillar spines. Very small specimens are relatively
much rougher (though not invariably so) than the large ones, depending, probably,
on the contraction of the membrane at death. (See ‘‘ Variations,’ on p. 361.) Pseu-
dopaxille very numerous, with a rather low pedicel which is shorter than the spines,
Spines numerous, about twenty-six in full-sized specimens; of these, seven or eight
thicker ones form a diverging group on the periphery, the rest being much slenderer,
and delicate, and occupying the center. The membrane in which the spines are
immersed forms seven or eight lamin radiating from the center of the group and
ending with each of the seven or eight peripheral spines. All spines are rather
stouter at the distal end. The muscle bands, which show as reticulations exteriorly,
hang down on the inner side of the supradorsal membrane like tiny partitions or cur-
tains, so that the ‘“‘ceiling”’ of the nidamental cavity is fenestrated. The seven or
eight (fewer in small specimens) radiating spines with their laminz of membrane
rise in a quite regular manner and join these so that the appearance is that of a
groined vault in miniature. Length of pedicel, about 2 mm.; length of spines,
about 3mm. Osculum conspicuous with a raised border.

Ambulacra narrow; feet large in two rows. Adambulacral combs, close
together, in large specimens with seven spines (proximally at least), or six in medium-
sized and small examples (five in some very small ones). Membrane slightly emar-
ginated between spines, the four or five outer of which have swollen sacculated
tips; the three or four outer subequal; the third from inner end slightly shorter,
the second about half as long as third, and the first very short and placed aborally to
the series. Occasionally five spines are long and subequal (in a series of seven), but
four is the usual number, in large and medium-sized examples, and three in small
ones. Outermost spine close to actinolateral membrane with which it is joined by
membrane; aperture a little over half length of outer spine, the papilla narrow and
with free aboral edge.

Mouth spines five to seven, usually six, the outer very short and the rest grad-
ually lengthening to the inner, which is slightly longer than interradial diameter of
plate; the whole ten to fourteen of each angle are united by a common membrane
to tips. Suboral spine heavier and longer than any marginal spine, slender,
subcylindrical, clear-tipped, pointed, and incased in heavy membrane.

/
ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 361

Actinolateral spines very short, the proximal ones only one and a half to one
and three-quarters times as long as outer adambulacral spine; consequently, the
membrane is narrow and does not increase very much in width proximally. In
large specimens this membrane is flush with the general actinal floor, there being
no free edge or fringe except far along ray. In young examples and in dried spec-
imens there is always a more or less marked free border to the actinolateral mem-
brane.

Color in life, variable, generally bluish gray or dull purplish mottled with fawn
color; a specimen from station 4235 was very light yellowish (yellowish white) in
life, the tube feet canary yellow. The mottling usually persists in alcohol.

Anatomical notes.—Abactinal plates cruciform with four (or five) long slender
truncate lobes. Madreporic body very convex, as high as wide, with a tuft of
spines at the top (‘‘pseudoplacote’’). Anal opening very large, surrounded by a
membranous periproct 11 mm. broad, and bounded by the five large radial (osculum)
paxille. Anus opens from a spacious simple chamber with four short saccular
diverticula. This chamber in turn connects with the dorsal stomach by a large
aperture. Hepatic cceca large. Very strong muscles from the enlarged first
ambulacral ossicle to stomach wall. Gonads interradial, in large tuft. Apparently
no Polian vesicles; Tiedemann bodies large. First tube feet, on either side, with
only one ampulla (as in Pt. jordani); the rest with two. Interradial septa mem-
branous, tough.

Variations —The Alaskan and British Columbian specimens vary slightly in
proportions, some having more inflated and shorter rays than others. The tip of
the ray is more or less upturned or reflected, so that the ends of the ambulacra
assume a dorsal position. The degree of this upturning causes a variation in the
length of rays; the extremes are about R=1.5 to 1.85 r; those less than 1.6 r are
young; the usual is R=about 1.65 r. Besides a variation in color there is also a
great difference in the texture of the supradorsal membrane. Usually it is only
slightly roughened by the tips of the spinelets. The reticulation shows most plainly
on the lateral and ventrolateral surfaces, and often not at all plainly on the dorsal
surface. If the surface layer of the supradorsal membrane is much swollen, the
reticulations are wholly or partly obliterated and the spiracula appear to open into
fine furrows or creases. Variations in the adambulacral spines and actinolateral
membrane have already been alluded to.

Type.—No. 292, Academy of Natural Sciences, Philadelphia.

Type-locality—Kadiak Island, Alaska. The locality was omitted in the
original description, but was recorded later.*

Distribution —Bering Sea (Bering Island, Pribilofs, Hagemeister Island) south
along the North American coast to Washington.

Proc. Acad. Nat. Sci. Phila., 1889, p. 175.

362 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Specimens examined.—Thirty-nine, from the following stations:

Specimens of Pteraster tesselatus examined.

Station. Locality. Depth. Nature of bottom. Number. Collection.
Fathoms.
Hagemeister Island, Alaska. ...--. 8-15 1 U.S. Nat. Mus., W. H. Dall.
Shumagin Islands, Alaska. ......- 48 oe 1 Do.
Unalaska, Alaska......--....---- Extra low 2 Do.
water.
Bering Island..........- 2 U.S. Nat. Mus.
Popof Straits, Shumagins....... se 1 U.S. Nat. Mus., W. H. Dall.
Victoria, British Columbia.....-.. 1 U.S. Nat. Mus., C. F. New-
combe.
Barclay Sound, British Columbia.|..............|..------20-ceeseeeceserees 1 U. S. Nat. Mus., Albatross,
1889.
Departure Bay, Nanaimo, British |.............-|..-.--2--eeeeeeeeee eee eee 1 U. S. Nat. Mus., Geo. W.
Columbia. Taylor.
QR62E Seek Off Vancouver Island.......-....| 238 gray sand, pebbles...... 1 U.S. Nat. Mus.
$213...... South of Alaskan Peninsula (San- 41 black sand 1 Do.
nak Islands).
South of Unimak Island, Alaska. OOMe melee OSs cssieesee eee ee 1 Do.
.| Near Unimak Island, Alaska..... 43 black stones, rocks... ... 1 Do.
| South of Alaska Peninsula (east 19 black sand, shells....... 1 Do.
of Unimak Island). ‘
B46 occas. Straits of Fuca, Washington. ...- 114 gray sand, gravel........ 1 Do.
3486...... Bering Sea, west of Pribilofs...... | 150 green mud, fine sand.... 1 Do.
4205.....-. Vicinity of Port Townsend, Wash- 15-26 rocks, shells........----- 2 Albatross, 1903.
25 rocks, coarse sand, shells. 1 Do.
39 gray sand, broken shells. 4 Do.
4235...... Vicinity of Yes Bay, Behm Canal. 130-193 prayed esa oe cere 1 Do.
4285...... Chignik Bay, Alaskan Peninsula. . 31-59 gray sand, shells.......- 1 Do.
4289...... Uyak Bay, Kadiak............-.. 74-80 gray ude se. soscc eee 2 Do.
BUUiclnwnce « Near Semisopochnoi Island, Aleu- 52 fine gravel............-- 6 Albatross, 1906.
tians.
Citi Sep ce] hess QO bee ace odans ee cians neem epics 54 broken shells, pebbles, 4 Do.
sand.
4784...... Near Attu Island, Aleutians..... 135 coarse pebbles .......-..- 1 Do.

Remarks.—It may eventually be considered advisable to place this and the
following form in the genus Retaster. Practically all the current diagnoses of
Retaster read as if Pteraster never had a dorsal reticulum, or else as if deposits in the
supradorsal membrane characterized Pteraster. Many species of Pteraster entirely
lack the spicules. So far as the reticulum is concerned, the difference between the
two genera is one of degree rather than kind. Species that may be considered
typical members of Retaster are capensis, cribrosus, gibber, and insignis. These have
the muscle bands conspicuously developed, dividing the membrane into areas
sharply circumscribing the spiracula. Just this is found in tesselatus, but the muscle
bands are not usually conspicuous, except on the sides of the body (and elsewhere
in specimens in which the epidermis is more or less rubbed). The narrow actino-
lateral membrane is not characteristic of Retaster, since several undoubted Pteraster
also possess it. All known Retaster (excluding Diplopteraster), however, have the
narrow membrane.

In the following subspecies, or it may be, closely related species, the reticula-
tions are not evident superficially in the unique specimen. Neither are the spira-

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 363

cula in circumscribed areas, as they should be in Retaster. It is this circumstance
which has led me to keep both forms in Pteraster, although for several years I con-
sidered tesselatus a Retaster..

PTERASTER TESSELATUS ARCUATUS, new subspecies.
Pl. 103, fig. 1.

Diagnosis.—Similar to Pt. tesselatus, but arcuately pentagonal in form, with
thick supradorsal membrane, without signs of reticulation; spiracula sunken in
irregular creases. R=46 mm.; r=32 mm.; R=1.44 r.

Description.—Supradorsal membrane rather thicker than in typical tesselatus
and marked by fine, irregular, discontinuous, often branching, dark creases, into
which open the spiracula. These are more prominent on the dorsal than lateral
surfaces. Paxillar spines eighteen to twenty-four. Adambulacral spines five or
six, of which three are usually long and subequal and the two inner decidedly
shorter. Of the latter the innermost is about one-third and the second spine two-
thirds the length of the third spine. Mouth plates with six marginal spines similar
to those of typical tesselatus; suboral spine about as long as innermost marginal
spine, and with a transparent, tapering, sharp tip.

Color in life, abactinal surface mottled with fawn color on a ground of bluish-
gray or dull heliotrope purple; radiating from center are five narrow, dark, inter-
radial patches extending halfway to margin, and five radial larger and lighter areas;
outside of this is a stellate, irregular ring of dark, near ambitus. Actinal surface,
actinolateral membrane vinaceous buff; interradial areas dull heliotrope purple
(that is, outside of the narrow membrane); tube feet, yellowish.

Type.—Cat. No. 27794, U.S.N.M.

Type-locality.— Albatross station 4551, Monterey Bay, California, 56 to 46
fathoms, rocks.

Distribution.—Known only from the type, taken in Monterey Bay.

Remarks.—This form differs from northern tesselatus chiefly in the arcuately
pentagonal form, generally one less furrow spine, and in having the spiracula in
fine, dark furrows, which form disconnected, meandering, branched designs. The
reticulations of the supradorsal membrane of tesselatus are not evident in arcuatus.

PTERASTER OBSCURUS (Perrier).
Pl. 105, figs. 1-4; pl. 106, figs. 1, 2.

Hexaster obscurus Perrier, Comptes rendus, vol. 112, No. 21, May 1, 1891, p. 1227; Mém. soc. zool.
France, vol. 4, 1891] p. 267; Résult. camp. sci. du Prince de Monaco, fase. 11, 1896, p. 41,
pl. 3, figs. 1, la —Verrit1, Trans. Conn. Acad., vol. 10, 1899, p. 221—Lupwie, Arktische
Seesterne, 1900, p. 468 (distribution).

Pteraster (Temnaster) heractis Verrut, Proc. U. 8. Nat. Mus., vol. 17, 1894, p. 275.

Pteraster hexactis DépERLEIN, Zool. Anz., vol. 22, 1889, p. 338.

Temnaster hexactis Verruti, Amer. Journ. Sci., vol. 49, 1895, p. 202.

Pteraster obscurus DépEeRLEIN, Wiss. Meeresuntersuch., (N.S.) vol. 4, Abth. Helgoland, Heft 2,
1900, p. 213, pl. 8.—Morrensen, Meddelelser om Grpnland, vol. 29, 1903, p. 77.

Pteraster octaster Verrut, Amer. Journ. Sci., vol. 28, July, 1909, p. 61, fig. 1.

Diagnosis.—Rays six, seven, eight, or nine form®; robust, rays short, broad,
and usually rigid; disk high; R=1.4 to 1.65 r. Supradorsal membrane tough and

2 Only one example, the young of an eight-rayed specimen.

364 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

thick, without reticulation superficially, without spicules, and with very numerous
spiracula; sides of rays rounded; paxille stout with six to eight peripheral spines,
one being considerably enlarged, and from four to eight slenderer central ones;
peripheral spines give supradorsal membrane a bristling appearance ; pedicel slightly
higher than shorter spines; adambulacral spines four to seven, the innermost short-
est, the others subequal, flattened, truncate, bound together by a very thick skin
which is continued across actinolateral membrane as a low fold; actinolateral
membrane thick, wide; lateral fringe narrow, usually defining ambitus; oral spines
five or six, the two series webbed into a single scoop-shaped group at each mouth
angle; suboral spine pointed, hyaline, shorter than inner flattened mouth spine.

Notes on North Pacific specimens—Supplementing the description given by
Verrill of a specimen from Albatross station 2433, off Newfoundland Bank (No.
12004, U.S.N.M.), 57 fathoms, the following notes based on specimens from Bering
Sea and vicinity may be of value in showing the range of variation:

Perrier’s type measured R=20 mm.; Verrill’s, R=32 to 35 mm., r=22 mm.,
R=about 1.5 mm. The largest specimen from station 3285, with seven rays, has
the following dimensions: R=82 mm., r=50 mm.+, R=about 1.6 r. The largest
six-rayed specimen, station 3282, measures R=71 mm., r=43 mm., R=1.65 r}; a
small specimen, R=24 mm., r=17 mm., R=1.4 r. Twenty-one have six rays,
two have seven, four have eight, and one young extracted from the nidamental
cavity of an eight-rayed specimen has nine rays. The abactinal surface is well
arched, the rays usually swollen and the actinal surface slightly sunken. The general
appearance is shown by the figure.

The supradorsal membrane is very tough and opaque, with muscle fibers (when
viewed from the inner surface), but not regularly reticulated. From the outside
no muscle bands show at all. The surface is either uniformly papillated with the
stout peripheral pseudopaxillar spines, or a few paxilla: cause the surface to be
raised in warty humps (as shown by Perrier’s figure), these latter in addition to the
usual spiny appearance. Spiracula very numerous; and usually in large specimens
groups or lines of two or more lie in furrows or dark creases of the skin. The smooth
interradial grooves vary in width and length, but there are no slits such as Verrill
found; these are formed periodically as he suggests. (See ‘‘ Young,” on p. 365.)
Usually the area without spiracula extends about halfway from margin to osculum,
and is very narrow. (See also Déderlein, 1889, 1900.)

Pseudopapille stout, shorter abactinally than laterally, with robust cruciform
bases. Pedicel stout, abactinally a little higher than the spines. A specimen from

3251 has six to eight stout, often clavate, rough-tipped, blunt, slightly radiating,
peripheral spines and often a stout central spine to each pedicel. Surrounding the
central spine (between it and the peripheral six to eight) are from four to eight
very much slenderer spinules, which do not, as a rule, push up the supradorsal mem-
brane to an appreciable extent. One of the peripheral spines is usually much
enlarged with a blunt clavate tip which shows more prominently from the outside,
being thicker, heavier, and usually longer than the rest. The large seven-rayed
specimen has the same structure. The enlarged peripheral spine being well marked
and on some paxille greatly enlarged, especially on the dorsolateral surface of the
rays. Tips of spines minutely roughened.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 365

The largest six-rayed example has five spines in each of the adambulacral combs,
except on outer part of ray where there are four. The inner is shorter than the
outer three or four, which are subequal, or the outermost a trifle the longest as well
as stoutest. The spines are flattened and vary in thickness in different specimens,
being sometimes slightly tapered, sometimes not; tip of first spine bluntly pointed,
rough, the rest rough, obtuse, or truncate. The membrane is thick and obscures
the spines. The edge is emarginate between the spines and forms usually a rounded
tip to each; from the outermost spine the fold extends across the actinolateral
membrane as described by Verrill. Thus the space between each pair of combs
is continued as a narrow sulcus toward and even to the margin, as well shown in
the photographic figure and in Perrier’s figure of a little specimen. The adam-
bulacral structure does not differ in the large seven-rayed example except that
proximally there are six spines. The outermost is bent away from the rest as
described by Verrill in his type from off Newfoundland Bank. The tip of the spines,
as in typical examples, is armed with minute thornlets and is more or less com-
pressed and truncate. The number of adambulacral spines is variable in six-rayed
examples, however, for a medium-sized specimen has seven, others six, some five,
and small specimens four. The smaller seven-rayed examples have six spines
proximally. Aperture short and rather broad with a very convex aboral border.

Mouth spines of the two plates united by a thick continuous web forming a
spade-shaped group, five or six to a plate and decreasing very rapidly from the
inner to the outer; the larger spines, especially the innermost, considerably flattened;
the latter longer than interradial diameter of plate. Suboral spine stout, with a
hyaline pointed tip, about equal to interradial diameter of plate, and therefore
shorter than innermost oral spine.

Actinolateral membrane tough, and traversed by continuations of the adam-
bulacral fans. The extent of the free border is variable. If the supradorsal mem-
brane is swollen greatly the fringe is not apparent; otherwise it defines the ambitus,
as shown in Perrier’s figure. The actinolateral spines are stout, and hidden by
the thick membrane. The tips however are visible, being heavy and truncate.

Young.—tThe large seven-rayed example had a young one in the act of escaping
through a breach of the supradorsal membrane when killed. This opening is just
above the actinolateral border and midway along the side of ray. The membrane
was stretched and thin over that point. In one of the smooth interradial regions
the membrane was found to be thin also, and a cut revealed two young ones beneath,
actinal surface uppermost. The larger of these two has R=10.5 mm., r=7 mm.
The one which was escaping has R=7mm. Both have six rays, although the parent
hasseven. The membranes of the young are thin, all the actinal spines being clearly
visible. The lateral fringe is prominent. Adambulacral spines five, oral spines
five; prominent paxillar spines (as seen from without) four or five; spiracula four
or five over each paxilla, one spiracle between a pair of spines. The general appear-
ance of this young one is exactly like that of a Pteraster with thin membranes. The
six rays in the young shows that this is the fundamental number for the species,
despite the seven of the parent.

The slits observed by Verrill are temporary only, and may occur elsewhere
than in the interradial regions. They are probably formed periodically in other

366 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

species of Pteraster also, and are the normal means by which the large young escape
from the nidamental supradorsal cavity. The slits can not be regarded as a special
generic character. The young when they leave the parent are essentially like the
adult except in having more delicate membranes, fewer abactinal spines and spiracles,
They are also exceptionally large, the largest one having a diameter of 20 mm.

Type.—Musée océanographique, Monaco (%).

Type-locality.— Off Newfoundland, 155 meters.

Distribution Bering Sea, from Bering Straits south to Bering Island, Kam-
chatka, and Unalaska; Spitzbergen (Déderlein), Greenland (Mortensen), Newfound-
land (Perrier, Verrill). Evidently the species is circumpolar. The vertical dis-
tribution in Bering Sea is 17 to 85 fathoms; in the North Atlantic, 50 to 145 fathoms.
The specimens from Bering Island were evidently washed ashore as they are much
abraded.

Specimens examined.—Twenty-seven, from the following localities:

Specimens of Pteraster obscurus examined.

Station. | Locality. | Depth. | Nature of bottom. Number. Collection.
| a) Gamat Clan a
| | Fathoms.
Bering Island, (Commander) “s= |2222-0--0<| senevice ocr sce aeminm senieen sees 2a U.S. Nat. Mus., L. Stejneger
lands. |
| Bering Strait, 12 miles east of 7 belle as saingcaaeconeo tesa la U.S. Nat. Mus., W. H. Dall,
King’s Island. | 1210 (6101).
8225 ec o<e Near Unalaska..........--------- 58 black sand 5- 22 -s-nccccense 2 U.S. Nat. Mus.
3251...... | Bristol Bay, Alaskacsnccn-\sce=--- 37 fine gray sand............... 2 Do.
S200 seaiese Bering Sea. north of Unimak Is- 38 black sande -secesecc eee are 9 Do.
| Jand, Alaska.
3282...... | Bering Sea, north of Alaskan Pe- 53 fine sand, green mud........ 5 Do.
ninsula (near Point Moller).
3285...-.- Bering Sea, north of Alaskan Pen- 35 grayisand eo ecko 1d Do.
insula (Cape Kutuzof).
3505.....- East of Pribilof Islands, Bering Sea. 44 fine gray sand............... 1 Do.
3512...... Bering Sea,north of Pribilof Islands 38 fine sand, green mud........ 1 Do.
3557...-.. Pribilof Islands...........-------- 26 SANG esa eee eee la Do.
Acc:.INo: 16889 "0 BIN SMa. een anc oote seine s|vommiancics te one e a ee cece ee rae 1b Corwin, Alaska, 1881.
Ace. No. 15278, U-S.N.M..........|-...2.002+ [Eoye te uO st oa 1 Corwin, Alaska, 1884.
» Seven-rayed. a Right-rayed.

Remarks.—Perhaps the first specimen of this species ever taken was that col-
lected by Dr. W. H. Dall near Bering Strait in 1874 (Cat. No. 6101, U.S.N.M.).
The species was not described until seventeen years later, and not recognized by
zoologists generally until after 1894 when Verrill described a specimen from off
Newfoundland Bank (lat. 43° 05’ N.; long. 50° 43’ W.). Perrier’s preliminary
descriptions (1891) seem to have been overlooked, even by Perrier himself, for he did
not include his new genus in the keys of the Travailleur and Talisman report, 1894.

The work of the Albatross in Bering Sea, and the Dall specimen, greatly increase
the known range of this interesting form, which has hitherto been recorded only
from scattered localities in the North Atlantic. All the Atlantic specimens of
which I have seen any notices seem to be six- or seven-rayed, and smaller than the
largest Bering Sea example. There is little doubt that the seven- and eight-rayed
specimens belong to the same species as the six-rayed. The structural details are

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 367

just alike; and what is conclusive is the fact that young of a seven-rayed specimen
have six rays, while young of eight-rayed specimens may have seven rays.

The young grow to a large size before escaping, by rupture of the supradorsal
membrane. The membrane first becomes thin over the young, which are usually
found in the interradial region, but may occur also along the rays. The membrane
has the same appearance, before rupture, that the dorsal integument of Luidia
assumes before the autotomy of a ray. The fact that the young had its ventral
surface close to the membrane, about to rupture, suggests that possibly its digestive
juices may play a part in the process of escape. After the escape of the young the
membrane closes, and the slit heals, causing probably the scarlike appearance of
the interradial smooth areas.

This species stands quite alone, having no near relatives. Perrier, Verrill,
and Ludwig place it in the monotypic genus Hezaster; Déderlein, Mortensen, and
Hamann, consider that it belongs in Pteraster. Perrier and Verrill independently
made the species the type of a genus, so that it may be well to examine the original
descriptions. Omitting those characters of Perrier’s diagnosis which apply equally
well to Pteraster, the following remain: Six rays, actinolateral spines immersed in
the ventral membrane, not extending beyond the border of the ray. The second
character is nullified by Perrier’s figure of the type. As a matter of fact, the lateral
fringe is present, though not well developed. The supradorsal membrane has
muscle fibers which are evident enough from the inner side; if these were not
present how would the membrane assume so many different forms or attain such
‘an unusual toughness and thickness? Most of Perrier’s diagnosis applies equally
well to Pteraster.

Verrill (1894, p. 277) says: ‘‘This species not only differs from other known
forms in having six rays, but appears to be peculiar in the presence of naked inter-
radial grooves and genital slits. This last character may be sufficient to warrant
its separation as a distinct subgenus (Temnaster Verrill) or even as a genus. It
differs from our other species also in having fewer and stouter spines in the ventral
combs; in the broader and flatter ventral surface of the shorter rays; in the much
thicker skin of the ventral combs, and in the less evident comb of spines along the
margins of the rays. The dorsal membrane is also firmer and not at all granular;
the spinules over its surface are much more numerous, and the pores between them
are smaller and more numerous.”

The character of the adambulacral armature, the dimensions, the width of the
actinolateral membrane, and lateral fringe, the thickness of the supradorsal mem-
brane, the number of paxillar spines and spiracula are, as Verrill intended to show,
merely specific and not generic features. The generic characters mentioned in the
first sentence are: Six rays and interradial slits. The latter are temporary open-
ings for the escape of the young, not permanent apertures. Furthermore, as
probably holds true also for other species of Pteraster the slits may occur elsewhere
in the supradorsal membrane. In small specimens there is not even a smooth
interradial groove. We are really thrown back upon the number of rays which is
not invariably six, but is sometimes seven, eight, or nine. While it may be con-
venient to segregate Pteraster obscurus in a subgenus Hezaster I do not think the
number of rays should constitute a generic division, any more than in Anseropoda,

368 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Pedicellaster, Leptasterias, Asterias, or in other groups having constantly six-rayed
species. In the absence of any structural features separating this form from
Pteraster, I have classified the species under that genus.

Since the above was written Professor Verrill has described Pteraster octaster
from Bering Island.* I have listed two specimens with eight rays from Bering
Island, evidently a part of the same lot collected by Doctor Stejneger. These are in
rather poor condition as regards the actinal surface, but a small specimen from
station 3557 is in a perfect state of preservation.

I have gone over these specimens again and have compared them with seven-
rayed and six-rayed examples. The result is that I am unable to find sufficient
structural difference to warrant the segregation of the eight-rayed form as a sepa-
rate species. The highly characteristic paxille are the same in the eight as in the
six- and seven-rayed specimens, showing no more variation than can be found in
several six-rayed examples. So also with the actinal surface. In eight-rayed
specimens the interradial areas are naturally smaller than in examples with six
rays, but not much smaller. The structure and spinulation of the mouth and
adambulacral plates is the same.

Reference has already been made to the fact that the young of the large seven-
rayed specimen have six rays. I opened the nidamental cavity of the two eight-
rayed Bering Island examples and found six young, one of which has nine rays,
three have eight rays, and two have seven rays, all three kinds being found in the
same individual. This is enough to show that there is no stability in the number
of rays. This circumstance coupled with the structural similarity of six-, seven-,
and eight-rayed specimens has led me to place Pteraster octaster as a synonym of
obscurus.

It is of interest to note that Professor Verrill described the new form as a
Pteraster and not as Hezaster (Temnaster Verrill).

Two specimens were opened, but were too badly macerated to determine the
internal structure in detail. The anatomy seems to be essentially the same as
in typical Pteraster.

In “Some Japanese and East Indian Echinoderms’’’ Dr. Hubert Lyman
Clark has published a very useful key to the species of Pteraster, upon which the
following key is based. Pt. aporus has been omitted, and a number of changes, other
than the introduction of new species, instituted. As the two keys differ in some
details, both may be used to advantage for the identification of specimens. The
key has been brought up to January, 1910.

KEY TO THE KNOWN SPECIES OF PTERASTER.

a', Form more or less stellate; R greater than 1.8 r, usually 2 to 3.5 r.
b'. A stout spine (suboral) present on surface of each mouth plate.
c'. Tube feet in two series; supradorsal membrane with spiny deposits¢ and not conspicuously
reticulated.
d'. Ambulacral comb with more than five spines.
e'. Stalk of paxilla short, not much higher than thick; spinelets much longer, three to five;
supradorsal membrane crowded with deposits; oral spines six to eight. -.militaris, Miller.

@ Amer. Journ. Sci., July, 1909, p. 61, fig. 1.
5 Bull. Mus. Comp. Zodl., vol. 51, No. 11, April, 1908, p. 286.
¢ Not mentioned in caribbxus and affinis; probably present.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 369

KEY TO THE KNOWN SPECIES OF PTERASTER—continued.

e?. Stalk of paxilla high and slender with six or seven spines; oral spines five or six.
caribbaus Perrier.
@. Ambulacral comb with three to five spines.

e!. Oral spine very large and three-edged; paxilla stalk high with six spines; adambulacral
SPINGEL VOsa waves dos. Nac Meas Save ee enon ae Gat Chem y SUBUaR ages trigonodon Fisher.

e*. Oral spine not three-edged.
J‘. Oral spines six; supradorsal membrane thick and smooth with spiny calcareous bodies;
paxilla stalk low; three to five adambulacral spines.............-.-.... lebruni Perrier.
f°. Oral spines four; supradorsal membrane thin; paxille with numerous (five to ten)
spinelets;/ four adambilacral spines ees. <2: <2 sess ois nena seed affinis Smith,
c?, Tube feet in two series; supradorsal membrane thin, delicate, reticulated, and without deposits;
oral spines two, independently webbed............---.2.-.. 2-2-2 eee eeeeeeeee gracilis Clark.
c’. Tube feet in four series, at least proximally; supradorsal membrane either without deposits or

with a very few simple rods.
d', Adambulacral comb of three or four spines; oral spines four; no small lateral pockets back of
aperture papilla; no deposits in supradorsal membrane; spiracula......... jordani Fisher.
d@, Adambulacral comb of five spines proximally; oral spines five; small shallow pockets just
back of aperture papille; supradorsal membrane with a few scattered simple rods;

PACU PRG Gee eee nyaiscete one awa viewiivid ice iu aine suman t afew ae aie marsippus Fisher.

b?. No suboral spine present.
c!, Adambulacral spines five, in curved series; oral spines five.............- personatus Sladen.
c?. Adambulacral spines four, in straight series; oral spines six.................. sordidus Perrier.
©. Adambulacral spines six or five, oral spines six; fifteen to twenty spinelets to paxilla,
OGM Patties son ee ace mist eos. alate wa ean eis raka icin rea demas nea pee reductus® Keebler.

a?. Rays very short; R less than 1.8 r, usually 1.3 to 1.7 r; form often more or less pentagonal (that is
in pentamerous species); suboral spine present.
b'. Rays regularly five.
c!. Adambulacral armature five to seven spines.
d', Oral spines united by membrane.
e!. Oral spines five to seven, the two series joined by a single continuous membrane; abactinal
surface swollen and sometimes more or less warty in adult.
f'. Membranes thick and fleshy; size large; supradorsal membrane more or less reticulated,
or else creased by fine furrows into which open the spiracula; R=1.4 to 1.8 r.
q', With well-developed rays; R=1.6 to 1.8 r; with reticulated membrane... tesselatus Ives.
g?. Arcuately pentagonal in form; R=1.44r. No evident reticulations. ...arcuatus Fisher.
J?. Membranes not especially thick, but rather thin, often translucent; R=1.3 r.

g'. Paxille with high pedicel and five to fifteen spines; suboral spine slender cylindrical;
muscle bands connecting tips of paxillar spines; aperture slits small; oral spines
six to seven; no deposits in supradorsal membrane; lateral fringe not defining
BIND Wse eran can see eet a tet Gate emia aaaGe tenennwecaceecsoeenet pulvillus Sars.

g*. Paxille low, with low pedicel and five to seven short spines; suboral spine with three
sharp edges; no muscle bands connecting tips of paxillar spines; aperture slits
unusually large; oral spines five; in supradorsal membrane, deposits in the form of

branched rods; lateral fringe defining ambitus.................. temnochiton Fisher.
e. Oral spines three, the series independently webbed; abactinal surface not much elevated
and not at all warty; adambulacral spines, six...........-..-.---..-- multispinus Clark.

@. Oral spines free, not webbed.
ec’, Actinolateral membrane defining ambitus; first two (innermost) oral spines not flattened
and truncate; actinolateral spines much longer at middle of ray than near tip.
J’. R=1.7 r; suboral spine very heavy club-shaped, with an incipient three-sided point;
supradorsal membrane with reticulations; oral spines five to seven; spiracles not
GxHOSd NP IY NOMOrOUNTE shiv soos s ccatdGs mac danpweNenwasabak sees ee reticulatus Fisher.

4 Very close to personatus; according to Keehler, distinguished by its constantly smaller size, compact
paxillee, longer adambulacral and oral spines, with one more spine in each series.

7444°—Bull. 76—11——24

370 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

KEY TO THE KNOWN SPECIES OF PTERASTER—continued.

f*. R=1.4 1; suboral spines very slender and long, without trace of three-sided tip; no
heavy supradorsal reticulations; oral spines six; spiracles exceedingly numerous.
coscinopeplus Fisher.
e®. Actinolateral membrane very narrow, not defining ambitus; actinolateral spines of approxi-
mately equal length except at very tip of ray; no supradorsal reticulations; first two
oral'epines very flatjand trumcatesc. << stecneececls nlite eiseleeinte niet ieee ei obesus Clark.
c@, Adambulacral armature three to five spines.
d', Oral spines three to five, not united by a web.
e!, Adambulacral spines short, scarcely projecting beyond web.............-. stellifer Sladen.
e?. Adambulacral spines slender, projecting far beyond web..............--.--- danx Verrill.
@. Oral spines united wholly, or in groups, by a web.
e!. Oral spines of each mouth angle all united by a common membrane.
j'. Actinolateral membrane fairly broad, or very broad, the lateral fringe defining ambitus.
g'. Suboral spine three-edged, sharp; apertures unusually large; adambulacral spines
five;/oral/spinesifive:R—=1e3" fren ajaciteete es teeter ete eee temnochiton Fisher.
g*?. Suboral spine not three-edged; apertures not unusually large; adambulacral spines
three to five.

i'. Adambulacral spines usually three, sometimes four, short, scarcely projecting beyond
WED s (RAMA toi oios osielsniaps cys ata assis sia eialaja in oeles ae ee eee eS rugatus Sladen.

k?. Adambulacral spines three to five, usually four, projecting far beyond web.
i. R=1.75 r+; dorsal membrane thin, evidently reticulated. .semireticulatus Sladen.
. R=1.4 r+; dorsal membrane thick, not at all reticulate..........- ingolfi Perrier.
fj’. Actinolateral membrane narrow not defining ambitus; actinolateral spines short, little
longer than outermost adambulacral spine; oral spines six; adambulacral spines five;
spiracula very abundant; R=—!Gire ee cacesne acest eee acer multiporus Clark.
e’. Oral spines webbed so as to form at each mouth angle four groups; two innermost spines
long, four lateral short, each group united by web; adambulacral spines usually four;
nojspiracula... s222 fa: he dace es et sa see eee Ore RE CEE een Cees eee alveolatus Perrier.
b?. Rays regularly six, seven, eight, or nine, usually six; supradorsal membrane thick; spiracula
numerous; abactinal surface spiny; actinolateral membrane and furrow webs thick; adambula-
cral spines four to seven; oral spines five or six, completely webbed; R=1.4 to 1.65r.
obscurus Perrier.

Genus DIPLOPTERASTER Verrill.

Diplopteraster Verritt, Amer. Journ. Sci., vol. 20, 1880, p. 400. Type, Pteraster multipes Sars.

Retaster SLADEN, part, Challenger Asteroidea, 1889, p. 477.

Diagnosis.—Pterasteride agreeing with Retaster in having a strongly reticu-
lated supradorsal membrane, but differing in having two kinds of adambulacral
plates, less prominent alternating with more prominent, and correlated with them,
two sorts of furrow combs, one larger and set farther into furrow, and having some-
times one spine more, the other set back slightly; tube feet in four rows; mouth
spines all webbed together; actinolateral membrane (in contrast to Retaster) very
broad, and fleshy in fully adult specimens. Madreporic body not bearing a tuft of
spines on its summit.”

Three closely related species are known: besides the type, Diplopteraster
verrucosus (Sladen), Atlantic entrance to Strait of Magellan, and D. peregrinator
(Sladen), off Cumberland Bay, Kerguelen Island.

@ Tam not certain as to the value of thischaracter. Retaster and some, but apparently not all, species
of Pteraster have a paxilla on the madreporic body.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS——FISHER. 371

KEY TO THE KNOWN SPECIES OF DIPLOPTERASTER.

a‘, Central spine of pseudopaxille projecting far beyond the others, giving a very bristling or spiny
appearance to supradorsal membrane; prominent furrow combs with four or five spines; non-
prominent with three to five.

6'. Pseudopaxille with eight to ten spines; triangular meshes of supradorsal reticulum with eight to
twenty spiracula; usually one spine less in the nonprominent furrow comb; longest adam-
bulacral spines and sacculus at least three times, sometimes four times, as long as base line of
PRO DREN GOODE: Soe cnet cic atain'c. veh tdi w use aie woo Sea cnm Skiue ed multipes (Sara), p. 371.

6. Pseudopaxille with fifteen spines; triangular meshes with four or five spiracula; both kinds of
furrow combs with five spines; longest adambulacral spines and sacculus less than twice as long
as the base line of the prominent combs (figure)...........-......-2..-0000- verrucosus (Sladen).

a, Central spine of pseudopaxille not protruding far beyond the rest and supradorsal membrane not
conspicuously spiny; five to ten paxillar spinelets; two or three spiracula to a mesh; prominent
furrow combs with seven or eight spines; alternate combs with five.......... peregrinator (Sladen).
Remarks.—The species described by Sladen are evidently very close to multipes.

It is somewhat remarkable that no mention is made of this similarity. The differ-

ences outlined in the above synopsis may break down somewhat when an adequate

series of the southern forms is examined. It will not be surprising if D. verrucosus
becomes a synonym of multipes, for the two are much alike. However, if the charac-
ters outlined by Sladen hold constant, they can be kept distinct. Sladen’s specimens

(number not designated) came from a single station, at the Atlantic entrance to the

Strait of Magellan, 55 fathoms, sand.

As to the validity of this genus, opinions will, of course, differ. The characters
mentioned in the diagnosis are as important, surely, as those which serve to separate
Retaster from Pteraster. In fact, it is much easier to draw the line between Diplop-
teraster and Retaster, than between the latter and Pteraster.

DIPLOPTERASTER MULTIPES (Sars).
Pl. 107, figs. 1, 2.
Pteraster multipes M. Sars, Forh. Vid. Selsk. Christiania, 1865, p. 200; Fauna Littoralis Nor-
vegizw, pt. 3, 1877, p. 65, pl. 8, figs. 1-17.

Diplopteraster multipes Verritt, Amer. Journ. Sci., vol. 20, 1880, p. 400.
Retaster multipes SLADEN, Challenger Asteroidea, 1889, p. 477.

Diagnosis.—Rays five, rarely six; general form stellato-pentagonal or very
broadly stellate, with depressed disk; actinolateral membrane broad and fleshy
with free lateral border defining the ambitus; ambulacra very wide with large tube
feet in four rows; R=92 mm., r=71 mm., R=1.3+r; a small specimen, R=1.52 r.
Supradorsal membrane tough, bristling with the prominent central spines of each
paxilla; paxille with long pedicels surmounted by a prominent stout central spine
and seven to nine slender shorter radiating peripheral spines, the tip of each being
joined to that of central spine, and also together, by a muscular band, the thinner
membrane between being pierced by about eight to twenty spiracula, the spiracular
areas thus are radially disposed about the prominent central spine of each paxilla;
osculum large. Adambulacral plates with twokinds of combs, one projecting farther
into furrow than the other; prominent combs with four or five webbed spines;
nonprominent with three or four; web prolonged beyond tips of longer spines as a
long flap or sacculus. Oral spines four or five (rarely six on one side), slender, the

372 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

two series united by a continuous membrane; suboral spine as slender or slenderer
than innermost oral and usually slightly shorter.

Notes on Pacific specimens.—Most of the specimens are very large, the spines
being heavy and the integumentary developments of the actinal surface thick and
tough. In the specimen from off San Diego the spines are more delicate and the
whole animal more lax.

The specimens agree very well with Atlantic examples, with which comparison
has been made. Judging by the single Californian example from station 4353, and
one Japanese specimen, 475 fathoms, Suruga Gulf, variation has been toward aslight
decrease in the number of spines in the former, especially on the nonprominent
adambulacral plates, and a slight increase in the latter (seven spines in the promi-
nent combs, six in the nonprominent). In the Japanese example there are about
nine spines to the paxillee, which have very long pedicels relatively slightly longer
than in Atlantic examples.

Ajl the specimens are bristling with the stout central paxillar spines which push
up the membrane and occasionally protrude through. In small examples the
spiracles are disposed in seven regular series radiating from the central spine, but
later the arrangement becomes irregular (except in the Japanese specimen where
the series are fairly straight).

The ‘‘prominent”’ adambulacral combs usually have the innermost, or furrow,
spine long, while the ‘‘nonprominent”’ plate has a very small furrow spine, though
sometimes it is long. Always, however, the comb is set back. The nonprominent
plate may have as many spines as the prominent, though in that case the furrow
spine is usually small; there is ordinarily one spine less on the nonprominent plate.
Ina large specimen, beginning with a prominent plate at base of furrow, the number
of spines run as follows, the nonprominent being marked by a star: 5, 5*, 6, 4*, 5, 4*,
5, 4*, 4, 4*, 5, 4*, 5, 4*, 4, ete. The innermost spine, if fully developed, is the longest,
the rest decreasing slightly in length, outward. Sometimes the second is the longest,
or all are very nearly the same length, or at base of ray the spines are graduated
in length from the innermost outward. The saccular prolongations of the web at
tip of spines vary in length, being usually nearly as long as the spine itself. In small
specimens, however, they are not prominent. The first ambulacral web is continued
across to meet its fellow, just back of the mouth plates.

The actinolateral membrane is thick and fleshy in the large examples, in some of
which the free border forms a nearly straight edge to the side of the body, but is
arched inward in others. Small specimens have the membrane thinner, so that the
spines show through. The ambulacral webs are continued out upon the membrane
for a short and variable distance, farthest, apparently, in the very old specimens.

The madreporic body is large, very convex, marked by coarse radiating furrows.
There is no tuft of spinelets from the summit as in Pteraster tesselatus.

One of the specimens has six rays, but is otherwise like the rest.

Type-locality —Drébak, Norway; 60 fathoms.

Distribution.—Probably cireumpolar; known from the north Atlantic, on the
west side, from 35° to 45° north latitude (Verrill) and on the east side from Norway
and Barents Sea; in the north Pacific, from San Diego, California, and Suruga Gulf,
Japan, north to Bering Sea. Vertical distribution: in the Atlantic 50 to about 580
fathoms; in the Pacific, 81 to 640 fathoms.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 373

Specimens examined.—Eleven, from the following localities:

Specimens of Diplopteraster multipes examined.

Station. Locality. | Depth. Nature of bottom. Number. | Collection.
=e =] —_ Sa ———-
| Fathoms.

Stic... Near Unalaska, Alaska............. -| 121 | black sand, gravel.............----| 1 U. 8. Nat, Mus.
8226...... North of Unalaska..............--.... | 128 mud, sand, shells...............+.- 2 Do.

GSS lirascen (seen CO SB Ore pol een | 350 TERUG Gore's nies dk. npn Se a 1 | Do.
35000.....} South of Pribilof Islands.............. | 121 fine gray sand, gravel............. 2 | Do.
3602...... West of Pribilof Islands.............. } sl green mud, sand.................. 2 | Do.
4235...... Near Yes Bay, Behm Canal, Alaska. | 181 GTMy WOU. cca ads esac sdetnatessuuy 1 | Albatross, 1903.
A888 hes Off San Diego, California............. | 628-640 | green mUd.......-----eseerennenes 1 | Albatross, 1904.

1

4784...... Near Attu Island, Aleutians.......... 135 COMREO TIDDISR Ss cov custessetuses nas Albatross, 1906.

« Bottle broken by earthquake, 1906; some doubt as to label; probably this station.

Remarks. —The differences between this species and the two others of the genus
are noted in the foregoing key.

The explorations of the Albatross in the North Pacific have more than doubled
the known range of this species, which was heretofore recorded only from the North
Atlantic. Some of the Bering Sea examples are very large (R=over 90 mm.),
and therefore show a tendency shared by other North Atlantic forms found in this
general region; as, for example, Henricia sanguinolenta, Pteraster militaris, Pseud-
archaster parelii. These specimens, if not the largest known Pterasteride, are at
any rate the fleshiest and heaviest.

Genus HYMENASTER Wyville Thomson.

Hymenaster Wyvitt—E THomson, The Depths of the Sea, 1873, p. 120. Type, H. pellucidus
Wyville Thomson.

Diagnosis.—Pterasteride in which the armature of the adambulacral plates
does not form transverse webbed combs, the spinelets being independent; nida-
mental cavity spacious; spinelets of paxille not protruding through, but support-
ing the membrane, which contains numerous muscle fibers and few to many spiracula.

KEY TO THE SPECIES OF HYMENASTER HEREIN DESCRIBED.

a’. Adambulacral armature consisting of one spinelet; size very large; paxille forming special raised
areas; spiracula small in definite circumscribed areas .............---------+0+5- kehleri, p. 373.
a?, Adambulacral armature consisting of more than one spinelet; spiracula not in definite circum-
acribed areas.

b!. Median radial paxille very much reduced in size; with one to three basal lobes; five rows of paxil-
lw on ray; adambulacral spinelets, two or three................-..-.---- perissonotus, p. 376.
b?. Median radial paxille not greatly reduced in size, with four basal lobes; nine rows of paxille

along ray; very numerous small spiracula; adambulacral spinelets, four or five.
quadrispinosus, p. 380.

HYMENASTER KCHLERI Fisher.

Pl. 108, fig. 1; pl. 117, fig. 4.
Hymenaster kevhleri Fisner, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 170.

Diagnosis.—Very near to IH. nobilis Wyville Thomson,’ which it resembles in
general appearance, number, and arrangement of paxillar spines and adambulacral

4 For figures and description, Sladen 1889, p. 495, pl. 87, figs. 1-3.

874 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

spines, but from which it differs in the following details: spiracula in very definite
circumscribed areas, not confluent between the paxillar crowns; the regular trans-
verse spiracular areas of interbrachial web of nobilis absent or only indicated irregu-
larly toward ends of rays; replaced proximally by detached scattered small spiracu-
lar areas; furrow spinelet much shorter than segmental papilla; oral spines short
(one-half length of interradial suture, or less).

Description.—Size large; general form stellato-pentagonal; R= about 100 mm.,
r=about 70 mm., R= about 1.4 r. The second specimen has more evident rays
and deeper interbrachial arcs; R=82 mm., r=45 to 52 mm.; R varies from 1.8 to
1.5 r. The interbrachial webs well shown in Sladen’s figure of nobilis, and which
leave the paxillar area somewhat raised, and sharply differentiated in both nobilis
and the present form, are very muscular and contractile, hence in preservation are
likely to differ widely in extent. H. kehleri has deeper interbrachial arcs than
nobilis, but this may be due to preservation, as the two specimens differ so much.
Abactinal surface very similar to that of nobilis, the supradorsal membrane fairly
thick, and the less elevated interbrachial webs very fleshy. Paxille with three
rather long, slender spines (occasionally four), which raise the membrane exactly as
described by Sladen for nobilis; moreover the spinelets are oriented the same way;
two usually aboral to the third, which is opposite to their interspace. There are six
rows of paxille along each ray as in nobilis. Spiracula numerous and small, in very
definite irregular areas surrounded by a low fold of the integument, or the whole
area may be sunken; ten to one hundred spiracula to an area. On the lateral
fringe or web are scattered spiracula areas, some of which are elongated and are
homologous to the regular and more numerous parallel areas in H. nobilis. These
elongated areas are more evident in the large specimens and occur beyond the middle
of the ray and are not so long as in nobilis. Proximally the areas of the web are
small and without order. From Sladen’s description and figure of nobilis the spira-
cula of the paxillar region are not confined to sharply defined circumscribed areas, as
in kehleri, but are confluent among the paxille. Osculum large, similar to that of
nobilis, the fan-like valves being a little higher and slightly more acute; spines
twelve. The depression just back of the fans contains numerous spiracula, which
also occur in the web between the five fans; they also form linear areas between the
spines of the fans, but are variable in this respect.

Adambulacral plates with an angular furrow margin bearing a single short
spinelet (2 to 2.5 mm.) invested in membrane, which is prolonged beyond the tip
(as in nobilis) one to three times the length of the spine. Aperture papill rela-
tively somewhat larger than in nobilis, ovate, with a slight sacculus at tip (length,
3.5 mm., width 2 to 3 mm.). Tube feet large, in two rows; sucking disk small.

Actinolateral membrane similar to that of nobilis, thick and fleshy, traversed on
actinal surface by parallel grooves (between the actinolateral spines). In the type
these reach to the interradial line proximally but leave a smooth triangular area in
the interradius near the margin. The apex of this triangle is at the point where
the proximal actinolateral spines of neighboring rays meet in interradial lines. Near
margin the grooves are roughly parallel with border of disk, but, as just mentioned,
do not reach to the interradial line. Spines buried beneath the fleshy membrane,
the ninth the longest; adorally from that point the tips of the spines nearly or quite

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 375

touch in the interradial line, but distad they are widely separated. In the inter-
radial angle near border, crisscrossing muscle bands are very evident on actinal
surface, those nearest the margin being largest. In the cotype the grooves are not
nearly so conspicuous as in the type and the actinolateral spines are easily seen
through the membrane, which is not the case in the type. From the first to the
ninth or tenth the actinolateral spines gradually become longer; from about the
tenth to twentieth they are subequal, being between one-fourth and one-fifth R in
length; from the twentieth to the fortieth they decrease rapidly in length, those at
tip of ray being exceedingly short.

Mouth plates as shown in figure, with a broad, free margin. Marginal spines of
type, three or four, short (one-half as long as median suture or less), the inner being
spaced a little from the other three. Cotype with two or three marginal spines.
Suboral spine one, not far from median suture and slightly more than one-third
the distance from inner marginal spine to outer end of median suture. Spine
short and thick. (See Pl. 117, fig. 4.)

Color in alcohol, bleached pink or brownish purple above, dull reddish purple
below.

Anatomical notes.—Nidamental cavity large, extending beyond tips of actino-
lateral spines. Papule very large and saccular, one to each paxilla, except the pri-
mary radials, and rising higher than the pedicels, to which they are attached. They
sometimes are filled with folds from hepatic ceca. Base of paxille with three long,
slender processes; the adradial paxillee with two longer processes directed toward
margin and a short mesially directed arm. The abactinal skeleton is very open
and loose. There is no radial series of poxille, and the adradials do not join across
the median radial area. The periproct membrane between the primary radial (or
oscular) paxille and anal aperture contains thin, large, subcircular or polygonal,
closely juxtaposed plates in mosaic. There are about two concentric circles of
them. Intestinal cecum a rather spacious, simple, or short-pocketed sac between
the dorsal stomach and anal aperture. It looks like a dilation of the intestine.
The stomach and hepatic coca essentially as in Pteraster. First ambulacral
ossicle, as in Pteraster, greatly enlarged for the insertion of muscles controlling
stomach. Ampulle double; tube feet, with a small button on the end; apparently
no Polian vesicles. Madreporie body large, convex, with meandering strie.
Gonads interradial, attached to the membranous interradial septum.

Type.—Cat. No. 27795, U.S.N.M.

Type-locality.— Albatross station 3603, Bering Sea, between Pribilof Islands
and Unimak Island (Aleutians), 1,771 fathoms, blue ooze; two specimens.

Distribution —Known only from type-locality.

Remarks.—This large species is allied to H. nobilis, from which it differs, as
detailed in the foregoing description. It differs from H. carnosus Sladen, as does
H. nobilis, in having one adambulacral spinelet and specialized paxillar areas. The
same differences separate it from H. pentagonalis Fisher. Both of these species
have the fleshy habit of kahleri. H. violaceus Ludwig (southeast of Acapulco,
Mexico, 1,879 fathoms), which also has one adambulacral spine, differs in having
a central and about six peripheral paxillar spinelets one-third as long as the
central. The abactinal surface is like /7. carnosus in the form and arrangement

376 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

of the paxillar spines; therefore quite different from kehleri. There are seven
irregular longitudinal rows of these points, while in k@hlert (as in nobilis) there
are six regular series, a median radial being wanting. In violaceus the furrow
spinelet (without the membrane) is as long or a trifle longer than the segmental
papilla, while in kehleri the latter is one and one-half times as long as the furrow
spinelet.

Hymenaster nobilis was taken at Challenger station 158, south of Australia,
1,099 miles southwest of Cape Otway, 1,800 fathoms globigerina ooze, and by the
Investigator in the Bay of Bengal, 1,748 fathoms. It is of interest to find such a
closely related form in the extreme north Pacific.

I take pleasure in dedicating this handsome species to Dr. R. Keehler, of the
University of Lyon.

HYMENASTER PERISSONOTUS Fisher.
Pl. 109, figs. 1, 2; pl. 115, figs. 1, la-e; pl. 117, figs. 3, 3a.
Hymenaster perissonotus FisHer, Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 170.

Diagnosis.—General form depressed, nearly pentagonal, with subplane abac-
tinal surface and extensive interradial webs supported by actinolateral spines,
which are conspicuously heavier beyond middle of ray. Paxillar area not con-
spicuously raised. Marginal paxille conspicuously larger than the three series of
abactinal, the radial being very short and, distally, rudimentary. Supradorsal
membrane close to abactinal wall along median radial area. Spiracula large,
scattered, fairly, but not very numerous; membrane with very numerous criss-
crossing muscle fibers, which radiate from tips of delicate paxilla spinelets, usually
three or four of these to a paxilla; points of spinelets uniformly spaced on supra-
dorsal membrane, and individual paxille not distinguishable. Adambulacral
spinelets two, or less often three; oral spines five or six, rarely seven, suboral
spines one or two; aperture papille large, ovate.

Description.—General form pentagonal and very flat, the sides of the pentagon
being nearly straight, except that the tips of the rays are slightly prolonged; R=70
mm.,r=50mm.,R=1.4r. Lateral web extending beyond paxille and supported by
the stout actinolateral spines. There is only a very narrow interradial strip free from
the spines, of which those near middle of ray are conspicuously stouter than the
proximal ones. Supradorsal membrane thin, and when the epidermis is rubbed (as
it usually is), very translucent, showing thread-like slender muscle bands running
between the tips of the paxillar spines. These fibers are not of uniform thickness.
They crisscross in every direction and branch, and are more numerous in some
specimens than in others. Paxillar areas broadly petaloid, but not raised, the
abactinal surface being as near flat as possible. Adjacent areas are confluent
along the adcentral half of r. Paxillar spines forming rather abrupt low uniformly
spaced protuberances over the paxillar area, individual paxille being not at all
discernible. Spiracula large for this genus, larger and fewer than in quadrispi-
nosus, and scattered without order. The larger specimens have more than the
smaller. A few occur on the interradial area outside the paxillar region on finger-
like outpouchings of the nidamental cavity.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 377

The paxille of the median radial region are very small, with short spinelets
and pedicels beyond the proximal third of R, but on either side they rapidly increase
in length, the lateralmost or marginal being the largest. This striking discrep-
ancy in size results in the nidamental cavity being almost entirely interradial, as
the supradorsal membrane rests upon the ambulacral ridge beyond the proximal
third of R. The radial paxille at about the middle of R measure 2 mm. or less in
height (the pedicel being too short to measure). A corresponding inferomarginal
paxilla measures: pedicel, 4 mm., + spines, 6 mm. (station 2919). On the outer
third of the radial line the scattered paxille are rudimentary and simply anchor
the supradorsal membrane to the very thin abactinal integument. Small paxille
have three very slender spinelets; the larger adradial, three; and the marginal,
three to five spines, all very slender. Osculum large, the valves with about
eighteen spines on the side toward aperture and about five short ones back of
these. The lateral spinelets of the comb rapidly become smaller. (Pl. 115, figs.
la, 1b.)

The abactinal skeleton is exceedingly weak, the processes forming the base
of the paxille being unusually long and slender. The inferomarginal plates have
four, the superomarginal three proximally, two distally, and the abactinal one to
three. The interspaces are unusually large, especially along the medioradial area.
The two adoral processes of the inferomarginal paxille are longer than the two
aboral, and the inferior branch of the superomarginal plates is stouter and a trifle
longer than the one or two processes directed mesially. The true abactinal plates
have almost thread-like processes and are in three irregular series, a radial and on
either side an adradial, the latter with usually three processes proximally and two
on outer part of ray. Proximally the processes of the plates touch, but distally
the radial series of rudimentary paxille has often only a single long, slender process,
which may not be in connection with any other plate. The inconstant number
of lobes results in an irregular mesh, the integument of which on the proximal
part of the ray contains irregular, elongated scattered thin plates, consisting of a
single layer of calcareous reticulum. The membrane between the primary radial
or oscular valve plates and anal aperture contains numerous large, thin subcireular
or elongated plates. (Pl. 115, figs. 1, le, 1d, le.)

Papule large on the proximal part of the ray, small and confined to the lateral
parts of ray beyond the middle of R. The distal fourth of R is without papule.
The papule are saccular, and the surface is subdivided into many small outpock-
etings, giving the appearance of a miniature cauliflower. The papule stand alone
(not attached to paxille), and the proximal are often filled with the hepatic ceca.

Madreporic body very large and prominent, being about as high as wide,
subglobose, and marked with fine branching centrifugal strie. It is situated not
interradially, but next to a primary radial or oscular paxilla. The actual disk is
so small the plate has become pushed to one side.

Actual disk very small; R=85 mm., actual r=10 to 14 mm.; diameter of ray at
base about 15 mm.; diameter of madreporic body 8 to 10 mm. Thus it is evident
the actinolateral membrane is very wide and the interradial cavities spacious. (See
dimensions above.)

378 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Adambulacral furrows fairly wide; tube feet in two rows, with small disks or
buttons at the tip, except near the tip of ray where the feet are pointed. Armature
consists of typically two (but some plates with three) slender furrow spinelets in a
longitudinal series, the adoral slightly the longer (4.5 mm.) and about as long as the
distance between the base of two adoral spinelets of two plates. The spinelets have
a flange of tissue broadening distally and continued beyond the tip for a short dis-
tance. Aperture papilla situated just back of the aboral furrow spinelet, or the
interval between the two spinelets. It is very broad, ovate or obovate, and with
the fleshy part is about as long as the calcareous portion of the longer furrow spine-
let. The calcareous support of the papilla is variable, but is either ovate with a blunt
or truncate tip, or is even broader than long with a truncate ragged end. A fleshy
tip often extends into the aperture, plugging it up, but is not superficially apparent.
It is really a short sacculus and varies in size, apparently never long.

Free margin of each mouth plate nearly as long as interradial diameter of plate,
the combined plates, therefore, presenting a broad margin to actinostome. Marginal
spinelets slender, subequal, five or six, rarely seven, to each plate, similar to the
adambulacral spinelets, and nearly evenly spaced. Suboral spinelets two, some-
what thicker and a trifle shorter (about one-half length of interradial diameter of
plate). The inner one sometimes moves to the margin and becomes the innermost
marginal spine. When there are two suborals there are four or five marginals.

Actinolateral spines gradually increasing in length and thickness along ray, the
twelfth to eighteenth (about) being longest and heaviest. None of the spines quite
meet, across the interradial space, those of adjacent ray. From about the seven-
teenth or eighteenth the spines become rapidly shorter and the tips end at the free
border of the web instead of being embedded in the web. Most of the larger spines
are broken (some apparently before the death of the animal), forming a sort of joint.
Just above the articulating facet the base of each spine has an abrupt expansion on
either side. The membrane is thin, so that the spines are conspicuous. The inter-
radial portion between the tips of two series of spines is muscular.

Color in alcohol, bleached yellowish.

Variations.—The outward appearance of all the specimens is essentially the
same. Those from very deep water (station 3307) apparently have a more translucent
membrane, but this is in a large measure due to the epidermis having sloughed off
either from long immersion in very weak alcohol or from scouring of mud and fine
sand in the dredge. In such examples the muscle fibers are much more conspicuous
than in normal specimens. However, the supradorsal membrane is weaker in deep-
water specimens. There is variation in the number of muscle fibers. Mention has
already been made that a minority of the adambulacral plates may have three
spines; the number of such plates is highly variable, and they occur in both northern
and southern and in the shallower and deeper water specimens.

The contour of the disk varies. Sometimes the long actinolateral spines cause
the interradial web to extend outward, forming an interradial corner so that the
creature apparently has ten angles and ten sides instead of five.

Avery remarkable specimen from station 2859 off Prince of Wales Island, Alaska,
1,569 fathoms, was taken along with H. quadrispinosus. It exhibits all the charac-
ters of perissonotus except that the adambulacral plates have an oblique transverse

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 379

row of five flattened spinelets, the inner the shortest, the outer the longest, the rest
graduated between. Sometimes the next to outer is longest. The aperture papille
are larger as in perissonotus. The specimen has R. 72 mm. and the characters of
paxillz, abactinal plates, supradorsal membrane, spiracula, actinolateral spines and
membrane are absolutely typical of perissonotus. As this aberrant specimen was
taken with quadrispinosus, it is possible that it is a hybrid between that species and
perissonotus, although no typical examples of the latter were taken at station 2859.
A six-rayed specimen is present in the lot from station 3074.
Type.—Cat. No. 27796, U.S.N.M.
Type-locality.— Albatross station 2919, off Cortes Bank, west of San Diego,
California, 984 fathoms, gray mud.
Distribution.—F rom San Diego, California, to Bering Sea, 225 to 1,771 fathoms.
Specimens examined.—Seventy-eight, from the following localities:

Specimens of Hymenaster perissonotus examined.

Station. Locality. Depth, Nature of bottom. Number. | Collection.
| |
—_—_— > Oo Fh
| Fathoms. |
bo Off Cortes Bank, west of San Diego, | 984 | MUA TN) aoa csts< cease fobnekh« | 6 | U.S. Nat. Mus.
California.
Off Sea Lion Rock, Washington 877 1D ot Do.
Satan Obsa ces eee anc tan sosens << nls 859 a4 Do.
C7 ee Bering Sea (north of Unalaska) 22: eT Do.
3307...... | Bering Sea (northwest of Umnak 1,033 7 | Do.
Island). | |
3342 4..... | Off Queen Charlotte Islands, British 1,588 eh Do.
Columbia. |
3346...... | Off Tillamook Bay, Oregon......... | 786 | ll Do.
3601 ¢..... Bering Sea, between Pribilof | | 1,044 | green mud.... 1 | Do.
| and Unalaska.
3603...... | Bering Sea (east of PribilofIslands)...) 1,771 NDIA GORE. ees so ear. ose waa aes | 1 Do.
| No locality | 9 Do.
1 i '
a Bottles naken during earthquake,1¢ 1906, ‘but label seotmabie correct. 6 Bottle broken during earthquake 1906.

Remarks.—This species falls under section B, sub-b, b, of Sladen’s Challenger
Synopsis, although it does not quite agree because there may be three adambula-
cral spinelets on some plates, and only three paxillar spinelets. Of the four species
mentioned, perissonotus most resembles vicarius, from off Chile, 1,375 fathoms.
It differs from glaucus in lacking the conspicuously defined radial areas raised in
relief above the rest of the abactinal surface. From vicarius, perissonotus differs in
having the median radial paxillee very much smaller than the lateral, in having three
or four, not four, five, or six paxillar spinelets, in having a more pentagonal contour,
with the actinolateral spines increasing in thickness at about the middle of ray, and
in having sometimes three, as well as two, adambulacral spinelets. From pellucidus,
perissonotus differs in respect to the size of the paxille, and in having fewer paxilla
spinelets, more numerous muscles in the supradorsal membrane, a thicker less
transparent membrane, and more numerous and large spiracula, HH. pellucidus
has four conspicuous rows of paxillse along each ray. //. infernalis has a stellate
contour, eight to ten paxillar spinelets, and none of the more salient characteristics
of H. perissonotus. Among the species with three adambulacral spinelets, figured
by Sladen, perissonotus resembles most H. membranaceus Wyville Thomson, from

380 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

which it differs in having reduced median radial paxille (not mentioned, however,
in Sladen’s description), more often two than three adambulacral spinelets, more
numerous spiracula, and differently formed mouth spines.

Of the species described by Ludwig from the Panama-Galapagos region, the only
one with which the present needs comparison is H. gracilis, from which perissonotus
differs in the great disparity in size between radial and marginal paxille, in having
three or four, rarely five, paxilla spinelets (not as many as seven), in having the abac-
tinal plates (radial and adradial) with only one to three basal processes, not four.
The abactinal plates of gracilis are regularly disposed forming quadrate meshes, and
the superomarginal plates have five processes, not two or three as in perissonotus.
The abactinal integument of perissonotus contains calcareous platelets in the meshes
of the skeleton; gracilis has none. Gracilis has three or four oral spinelets, perisso-
notus has five to seven. The madreporic body of gracilis is small, that of perissonotus
very large.

This species is readily distinguished from others described in this report by the
remarkable structure of the abactinal paxille which with their supporting base are
greatly reduced on the radial areas, the marginal paxille being very much larger. The
nidamental cavity is chiefly between the rays, in consequence of the smallness of the
radial paxille. The number of adambulacral spines is usually two but sometimes
three.

HYMENASTER QUADRISPINOSUS Fisher.
Pl. 108, fig. 2; pl. 110, figs. 1-3; pl. 117, fig. 2.

Hymenaster quadrispinosus Fisner, Bull. Bur. Fisheries for 1904, vol. 24, June 10, 1905, p. 315;
Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., 1910, p. 170.

Diagnosis.—Adambulacral spines four or five; aperture papilla broad, some-
times with a sacculus at tip; paxillar spines three or four (rarely five), the paxille
being arranged in nine rows along ray and forming specialized areas; lateral web
when not contracted extending nearly to tip of ray, with transverse parallel linear
spiracular areas; spiracula variable with age, very numerous, small, not in definite
areas but everywhere among the spine tips; paxilla crowns of adjacent rows alter-
nating.

Description.—The specimens exhibit extreme contractility of the interbrachial
web and variability of contour and proportions. In most of the specimens this
muscular interbrachial membrane is quite contracted, the paxillar crowns extending
to edge of body, which has a very stellate form. In two specimens, however, the
membrane is not contracted and exhibits transverse parallel lines of spiracula as
described by Sladen in //. nobilis, and the form is more pentagonal with the paxillar
areas sharply differentiated. The largest specimen has the rays too much recurved
to measure. A slightly smaller one with contracted interbrachial webs has R=51
mm., r=27 mm., R=1.88 r; a specimen with partly expanded web measures,
R=42 mm., r=29 mm., R=1.48 r.

Supradorsal membrane thin but tough, of a satiny luster, containing very
numerous but inconspicuous crisscrossing muscle fibers (visible only in the larger
Alaskan examples). Paxillar area comprising whole dorsal surface, except a free
lateral muscular web which is usually so contracted as to be very inconspicuous.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 381

Paxille with usually four spines (sometimes three, rarely five), which push up
the membrane (no central spine), the individual paxillw being easily distinguishable
and arranged in nine rows along each ray, the paxille of adjacent rows alternating
in a regular manner. Spiracula varying greatly in number according to age.
-In the immature type from off San Diego they are not numerous and are very
inconspicuous, but in the adult Alaskan examples they are very numerous though
small. In the smallest of these examples (R=10 mm.) spiracula are very much
fewer than in the largest. (See paragraph on young, below.) The spiracula when
fully developed are exceedingly numerous all over the supradorsal membrane both
between the paxillar crests and amid the spines, from the tops of which they tend
to arrange themselves in radiating lines. They are thus not in definite areas,
except on the interbrachial web, where the spiracula form parallel linear areas
extending outward from the paxillar region. These lines occur from the inter-
brachial angle (the longest line being fourth or fifth from angle) nearly to tip of
ray. As this fringing membrane is usually contracted these areas are then invisible
and were entirely overlooked in the type where they can not be seen. The spiracula
are not present between the spines of a paxilla crest in all the Alaskan specimens,
being absent in those from station 2859. Oscular orifice large; valves triangular with
a truncate summit, however; spinelets about eleven to sixteen to each valve, the
seven or eight central ones subequal, the laterals graduated; membrane of valves
rather thick; on base of each valve at either side a well-defined triangular spiracular
area about one-third the height of the longer spines, containing many spiracula;
in the membrane between adjacent valves numerous spiracula, which extend
nearly to the distal edge of the valves; supradorsal membrane slightly raised along
interradial lines in examples with reflexed rays.

Adambulacral spinelets flattened, tapering, sharp, in an oblique transverse,
slightly curved comb of four or five, the outermost longest and most adorally placed,
the rest graduated in length toward the inner, which is one-third to two-thirds
smaller. Just external to the outermost and in line with the series is the broadly
ovate, round-tipped aperture papilla, fitting tightly over the corresponding seg-
mental aperture like a valve, the base of this papilla abruptly narrowed for articu-
lation to the plate; the first papilla much larger than the rest and lying in the actino-
lateral membrane. In Bering Sea specimens the tip of some of the papillx,
especially proximally, is prolonged in a narrow membraneous sacculus or flap
one-third to one and one-half the length of papilla. The outermost spine of the comb
also has a fleshy tip of variable length. These sacculi are not present in the type,
but that is a young specimen and the sacculi are very deciduous. Moreover, the
Alaskan examples are not constant in this character as pointed out under “ Young,”
on page 382.

The combined mouth plates are higher and narrower than in typical species
of this genus. The medium suture is on the summit of a central eminence or keel
of the combined plates, which slope off steeply on either side to the margin next
to furrow. The actinostomial margin is slightly curved and slopes downward from
the median suture and forms a rounded angle at the mouth of furrow. At this
point is the greatest width of the combined plates. On this sloping free margin of

382 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

each plate are four or five slender, tapering, pointed, marginal spines, the inner
much the thickest and longest (about equaling interradial diameter of plate), and
a short distance back of this is a similar slender subequal suboral spine. The inner
marginal and suboral may have a sacculus; these are well developed in the largest
Alaskan example, which also has them on the aperture papille and adambulacral
spinelets.

Actinolateral membrane thin, the spines showing through, over thirty in
mature specimens, the seventh to eleventh the largest.

Color in alcohol rose pink; in life a deeper rose, madder, or scarlet.

Anatomical notes —Nidamental cavity spacious. Pedicels of paxille slender,
shorter than spines, about as 1.75:3 along midradial region, and as 3:5.5 on lateral-
most row; base of paxillze with four long slender processes, imbricating with similar
processes of four other paxille in such a way that very regular rhombic meshes
are formed, except between the two lateral (=marginal) rows of paxille (which
do not alternate as do the abactinal, but stand one above the other). Counting
directly across ray from one superomarginal to another there are three rhombic
meshes, or two complete ones and two halves; likewise omitting’ the two non-
alternating rows of paxille of either side, there are five longitudinal rows of true
abactinal paxille. Papule large, with many small divisions at the summit; there
is one to each paxilla except the primary radial, and none occur within the circle
formed of the primary radial (or oscular) paxille. The periproct membrane
contains several rows of thin, spaced, roundish plates, which consist of a single
layer of characteristic calcareous mesh. No calcareous deposits are discoverable
in the abactinal integument (other than the paxille) and none occur in the supra-
dorsal membrane. (Type, and specimen from station 3603 examined.)

Anal aperture a narrow slit; intestinal ceecum small, little more than a dila-
tion of the intestine. Gonad (female) interradial; interradial septa with thin
plates. Ampulle large, proximally single, but farther along the ray with two
parts (though essentially only two-lobed sacs). Tube feet in two rows, with a
small sucking disk, except near end of ray where the feet have an attenuate pointed
tip. First ambulacral ossicle enlarged and with the lateral wings of the large
odontophore forming a strong mouth frame; several large muscles to the stomach
are inserted on it.

Young.—The type is an immature specimen. The cruciform abactinal plates
overlap slightly by the lobes, which are not fully developed.

A series of nine specimens from station 2859 are graduated in size from R=
10 mm. to R=42 mm. and exhibit well the changes which take place from the
young to the mature form. It is this instructive series which has led me to place
the San Diego and Alaskan specimens in the same species, although adults of the
latter appear at first sight to be considerably different in the possession of many
more spiracula, as well as lappets or sacculi to the segmental papillae. A specimen
with R=15 mm. has the characteristic nine rows of alternating paxille (each
with four spinelets) along the ray. The spiracula are much fewer than in the
adult, are scattered between the paxillar crowns, and are apparently a trifle more
numerous than in the type (although on account of the whiteness of the membrane
it is difficult to see them in the latter). Adambulacral spines four, the aperture

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 3838

papilla as in the adult, but without any sacculus at the tip. The marginal mouth
spines are variable, ranging from three to five in number and differing also some-
what in relative position. The innermost is abruptly larger and usually somewhat
spaced from the laterals. It appears to be a suboral moved to a marginal position.
The longest adambulacral spinelet sometimes has a short sacculus. The smallest
specimens are essentially like the above except that it is increasingly difficult to
make out spiracula.

This series of young indicates that the number of adambulacral spines, number
of paxillar spines, number of rows of paxille along ray, and number of mouth
spines do not differ to any important extent with age, but that the spiracula
increase greatly with age, and integumentary appendages of spines also appear
in older specimens, when the young do not possess them. The interbrachial webs
increase with age and probably the special spiracular areas do not appear till the
animal is nearly adult.

Type.—Cat. No. 22344, U.S.N.M.

Type-locality.— Albatross station 4387, off San Diego, California, (longitude of
Point Conception), 1,059 fathoms, green mud; one specimen.

Distribution.—Bering Sea to San Diego, California, and probably south to the
vicinity of Panama, 1,059 to 1,771 fathoms, mud and ooze.

Specimens examined.—Fifteen; besides the type, nine from station 2859, off
Prince of Wales Island, Alaska, 1,569 fathoms, gray ooze; five from station 3603,
Bering Sea, between the Pribilof Islands and Unimak Island (Aleutians), 1,771
fathoms, blue ooze.

Remarks.—There is a possibility that the Alaskan specimens may constitute a
separate race. The lack of fully grown examples from the type-locality leaves the
question open. However, the only difference is the greater number of spiracula
in the Bering Sea examples. This character, as the specimens prove, varies greatly
with age, and immature Alaskan examples are not essentially different from the
type.

The foregoing description is an attempt to describe the species rather than a
single specimen. The description of the immature type was given when the species
was named, but this is incomplete for the species, and has accordingly been emended.

The very considerable number (four or five) of adambulacral spinelets sepa-
rates this species from most others of the genus, placing it in section D of Sladen’s
Synopsis (1889, p. 494). It differs from H. coccinatus in the form of the adam-
bulacral comb (compare figure with Challenger Asteroidea, pl. 91, fig. 12), lack of
distinct muscular reticulations in supradorsal membrane, in having many more
spiracula (as many as in H. porosissimus), in having a less widely expanded pair
of mouth plates (and different armature), and in having an interbrachial and
lateral muscular web. The abactinal muscular reticulations of coccinatus are very
conspicuous, although the type is probably young. J/. quadrispinosus differs
from H. precoquis * in having fewer paxillar spinelets, a paxillar area well differ-
entiated from a lateral web, more numerous spiracula, and large and differently
formed aperture papille. The aperture papillew of pracoquis are very like those of

@See Sladen Challenger Asteroidea, 1889, p. 524, pl. 90, figs. 5, 6; pl. 91, figs. 15-15.

384 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Pteraster, being free on one margin only. The mouth plates are also different.
Both species are from the Southern Ocean, west of the Crozet Islands, 1,375 fathoms.

In this section of Hymenaster must be ranged H. platyacanthus Ludwig and H.
purpureus Ludwig from the Panama-Galapagos region. Without a direct com-
parison of specimens it is not possible to determine the validity of these forms, as no
specific characters are figured. JI. platyacanthus differs apparently in having a
thin and tender supradorsal membrane, seven instead of nine rows of alternating
paxillee, deposits in the abactinal integument at base of papule, and in having the
papule in the middle instead of in the corner of the meshes (some of the lateral
papule are in the middle of the mesh in quadrispinosus). The deposits may be
present only in old specimens, and I am not sure that the position of the papule
may not vary. JH. purpureus differs from platyacanthus according to Ludwig
(1905, p. 196) in having a dark color, and a membraneous lappet or sacculus at the
tip of the aperture papilla. These differences are very trifling, as color is in star-
fishes an almost useless character, and the sacculus varies with age in quadrispinosus.
H. purpureus is, perhaps, the same species as H. quadrispinosus, but I hesitate to
place it in synonymy without specimens or figures for comparison.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 385

BIBLIOGRAPHY.

I. Seectan Papers AND GENERAL Works Rererrinc ro Norta Pacric Starrisues.
Agassiz, A.
1877. North American Starfishes.
Mem. Mus. Comp. Zool. Harvard College, vol. 5, No. 1. Quarto, with 20 plates.
Branopt, J. I’.
1835. Prodromus descriptionis Animalium ab H. Mertensio in orbis terrarum circumnavigatione
observatorum. Fasc. I, Petropoli, pp. 68 to 72.
1851. Bemerkungen iiber die Asteriden und Echiniden Ochotskischen, Kamtschatischen und Behr-
ingschen Meeres.
A. Th. y. Middendorff, Reise in den dussersten Norden and Osten Sibiriens, etc., vol. 2,
pt. 1; Echinodermen, Dr. F. Brandt und E. Grube, pp. 27 to 34.
Ciark, Hupert Lyman.
1901. Echinoderms from Puget Sound: Observations made on the Echinoderms collected by the
* parties from Columbia University, in Puget Sound in 1896 and 1897.
Proc. Bost. Soc. Nat. Hist., vol. 29, pp. 324-329, pl. 1-4.
Fisher, Water K.
1905. New Starfishes from Deep Water off California and Alaska.
Bull. Bur. Fisheries for 1904, vol. 4, June 10, pp. 291-320.
1906a. Two New Starfishes from Monterey Bay, California.
Zoologischer Anzeiger, vol. 30, No. 10, June 19, pp. 229-302.
1906. New Starfishes from the Pacific Coast of North America.
Proc. Washington Acad. Sci., vol. 8, Aug. 14, pp. 111-139.
1907. Note on Eremicaster, a Genus of Starfishes.
Zoologischer Anzeiger, vol. 32, No. 1, July 23, pp. 12-14.
1910a. New Pterasteridx from the North Pacific.
Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., pp. 167-170.
19106. New Genera of Starfishes.
Ann. and Mag. Nat. Hist., ser. 8, vol. 5, Feb., pp. 171-173.
1910c. New Starfishes from the North Pacific. I.—Phanerozonia.
Zoologischer Anzeiger, vol. 35, March 29, pp. 546-553.
1910d. New Starfishes from the North Pacific. II.—Spinulosa.
Zoologischer Anzeiger, vol. 35, March 29, pp. 568-574.
Gray, Jon Epwarp.
1840. A synopsis of the genera and species of the Class Hypostoma (Asterias Linn.).
Ann. and Mag. Nat. Hist., ser. 1, vol. 6, pp. 175-184.
Gruse, Ap. Epuarp.
1857a. Beschreibungen neuer oder weniger bekannten Seesterne und Seeigel.
Nova Acta Acad. Cees. Leop. Carol. Nat. Cur., vol. 27, 50 pp., 3 pls.
18576. Diagnosen einiger neuen Echinodermen.
Archiv f. Naturgesch., pp. 340 to 344.
1865. Ueber einige Seesterne des Breslauer Museums.
42. Jahresber. d. schlesischen Ges. f. vaterliindische Cultur, p. 52.
1866. Einige neue Seesterne des hiesigen Zoologischen Museums.

43. Jahresber. d. schlesischen Ges. f. vaterliindische Cultur, p. 59.
Ives, J. E.

1888. On two new species of starfishes.
Proc. Acad. Nat Sci. Philadelphia, p. 421.

1889. Catalogue of the Asteroidea and Ophiuroidea in the collection of the Academy of Natural

Sciences of Philadelphia.

Prov. Acad. Nat. Sci. Philadelphia, pp. 169-179.

(1892.) List of the Echinoderms and Crustacea in the Cabinet of Frederick Stearns, Detroit, Mich.
(Privately published, not dated; on p. iia list of west coast Asteroidea.)

57444°—Bull. 76—11——_-25

386 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

JENNINGS, H. 8S.

1907. Behavior of the Starfish Asterias forreri de Loriol.

Univ. Cal. Pub., Zoology, vol. 4, No. 2, pp. 53-185, 19 text figs.
Loriou, P. DE.

1887. Notes pour servir 4 l’étude des Echinodermes, II.

Recueil zoologique Suisse, vol. 4, No. 3, June 23, pp. 401-407, pl. 18.

1897. Notes pour servir a l’étude des Echinodermes, V.

Mém. soc. phys. et d’ hist. nat. Genéve, vol. 32, pt. 2, No. 9, pp. 12 to 24, pls. 1 to 3.
Lupwic, HuBert.
1886. Echinodermen des Beringsmeeres.
Zool. Jahrb., Abth. Syst., vol. 1, pp. 287 to 295, pl. 6.

1905. Asteroidea, in: Reports on an Exploration off the West Coasts of Mexico, Central and South
America, and off the Galapagos Islands, in charge of Alexander Agassiz, by the
U.S. Fish Commission steamer Albatross, during 1891, etc.

Mem. Mus. Comp. Zodl., vol. 32, July 17, pp. 1-292, 35 plates.

1907. Diagnosen neuer Tiefsee Seesterne aus der Familie der Porcellanasteriden.

Zoologischer Anzeiger, vol. 31, No. 9, March 5, pp. 317-318.
Monks, Sarau P.
1903. Regeneration of the Body of a Starfish.
Proc. Acad. Nat. Sci. Philadelphia, p. 351.
1904. Variability and Autotomy in Phataria.
Proc. Acad. Nat. Sci. Philadelphia, pp. 596-600, pl. 42.
Miituer, J., and Troscuet, F. H.
1842. System der Asteriden, Braunschweig.
Murpocu, JOHN.
1885. Marine Invertebrates.
Ray’s Report of the International Polar Expedition to Point Barrow, Alaska, Ex. Doc.
No. 44, pp. 158 to 160.
PeERRIER, EDMOND.
1875. Révision de la collection de Stellérides du museum d’histoire naturelle de Paris, pp. 64-71.
1876. Diagnoses of New Species of Asteriidze and Linckiide in the British Museum.
Ann. and Mag. Nat. Hist., ser. 4, vol. 17, pp. 34-36.
1878. Etude sur la répartition géographique des Asterides.
Nouv. arch. du mus. d’hist. nat. Paris, sér. 2, vol. 1, pp. 35, 91.
Stimpson, WILLIAM.
1857. On the Crustacea and Echinodermata of the Pacific Shores of North America,
Boston Journ. Nat. Hist., vol. 6, pp. 527-531, pl. 23.
1862. On new Genera and Species of Starfishes of the Family Pycnopodide.
Proc. Boston Soc. Nat. Hist., vol. 8, pp. 261-273.
VERRILL, A. E.
1867. On the Geographical Distribution of the Echinoderms of the West Coast of America.
Trans. Conn. Acad., vol. 1, pt. 2, pp. 323-339.

1880. Appendix C. On some Marine Invertebrata from the Queen Charlotte Islands. By J. F.
Whiteaves.

Report of Progress; Geol. Survey of Canada for 1878-79, pp. 192b-194b.

1894. Descriptions of New Species of Starfishes and Ophiurans, with a revision of certain species
formerly described, mostly from the collections made by the United States Com-
mission of Fish and Fisheries.

Proc. U. S. Nat. Mus., vol. 17, p. 268.
1909a. Descriptions of New Genera and Species of Starfishes from the North Pacific Coast of America.
Amer. Jour. Sci., vol. 28, July, pp. 59-70.

1909b. Remarkable Development of Starfishes on the Northwest American Coast; Hybridism; Mul-

tiplicity of Rays; Teratology; Problems in Evolution; Geographical Distribution.
Amer. Naturalist, vol. 43, Sept., pp. 542-555.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 387

Waurreaves, J. F.
1878. On Some Marine Invertebrata from the West Coast of North America.
Canadian Naturalist, vol. 8, Dec. 20, pp. 466, 467.

1887. On some Marine Invertebrata dredged or otherwise collected by Dr. G. M. Dawson in 1885,
on the coast of British Columbia; with a supplementary list of a few land and fresh-
water shells, fishes, birds, etc., from the same region.

Trans. Roy. Soc. Can., vol. 4, sec. 4, 1886, pp. 116-117.

II. GeneRAL Works AND Papers REFERRED TO IN THE ForEGOING Report.
(Papers in the above list are not repeated.)

Agassiz, L.
1835. Prodrome d’une monographie des radiares ou échinodermes,
Mém. soc. sci. nat. Neuchitel, vol. 1, pp. 168-199.
Axcock, A.
1893. Account of the collection of deep-sea Asteroidea (Natural History Notes from H. M. Indian
Marine Survey Steamer Investigator).
Ann. and Mag. Nat. Hist., ser. 6, vol. 11, pp. 73-121, pls. 4-6.
Batuer, F. A.
1906. (Ortmannia,)
Revue Critique de paléozoologie, Paris, vol. 10, April, p. 131.
Be, F. Jerrrey.
1890. On the generic name of Asterias sanguinolenta (Henricia).
Ann, and Mag. Nat. Hist., ser. 6, vol. 6, p. 472.
1892. Catalogue of the British Echinoderms in the British Museum, London, 202 p., 16 pls.
1893. On Odontaster and the Allied or Synonymous Genera of Asteroid Echinoderms.
Proc. Zool. Soc. London, pp. 259-262.
BurMEISTER, HERMANN.
1837. Handbuch der Naturgeschichte. 2 Abth., Zoologie, octavo.
Crark, Huspert LYMAN.
1904. The Echinoderms of the Woods Hole Region.
Bull. U. S. Fish Comm., 1902, pp. 551-557, pls. 3, 4.
1908. Some Japanese and East Indian Echinoderms.
Bull. Mus. Comp. Zoél., vol. 51, No. 11, April, pp. 279-289.
1909. Notes on some Australian and Indo-Pacific Echinoderms.
Bull. Mus. Comp. Zodl., vol. 52, No. 7, p. 110.
DANIELSSEN, D. C., and Koren, J.
1884, Asteroidea (Norwegian North Atlantic Expedition, 1876-1878), 118 pp., 15 plates, map.
Déveruei, L.
1899. Einige Beobachtungen an arktischen Seesternen.
Zoologischer Anzeiger, vol. 22, pp. 337-339.
1900. Die Echinodermen der Olga-Expedition.
Wiss. Meeresuntersuch. (N. S.) 1V, Abtheilung Helgoland, Heft 2, pp. 197-224, pls. 6-9.
Dusarpin, F., et Hurst, H.
1862. Histoire naturelle des zoophytes Echinodermes, 10 pls.
Fasricrius, O.
1780. Fauna greenlandica. Hafnis et Lipsise, pp. 367-372.
Fisuer, WALTER K.
1906. The Starfishes of the Hawaiian Islands.
Bull. U.S, Fish Comm. for 1903, pt. 3, pp. 987-1130, 49 pls.
1908a, Necessary Changes in the Nomenclature of Starfishes.
Smithsonian Miscellaneous Collections (Quarterly Issue), vol, 52, May 27, pp. 87593,
1908b. Some Necessary Changes in the Generic Names of Starfishes.
Zoologischer Anzeiger, vol, 33, No. 11, Aug. 18, pp. 356-359.

388 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Forses, Epwarp.
1839. On the Asteriade of the Irish Sea.
Mem. Wernerian Soc. Edinburgh, vol. 8, pt. I, pp. 114-129, 2 pls.
1841. A History of British Starfishes and other animals of the class Echinodermata, London, octavo.
Gray, Jonn Epwarp.
1848. List of the specimens of British animals in the collection of the British Museum. Part 1.
Centroniz or radiated animals.
1866. Synopsis of the Species of Starfishes in the British Museum, quarto, 16 pls.
1871. Description of Platasterias, a new Genus of Astropectinidee from Mexico.
Proc. Zool. Soc. London, pp. 136-187, pl. 9.
Greaory, J. W.
1900. The Stelleroidea.
A Treatise on Zoology, edited by E. Ray Lankester, pt. 3, pp. 237-259.
Griea, JAMES A.
1905. Goniaster nidarosiensis Storm og dens synonymer.
Bergens Museums Aarbog, 1905, No. 3, pp. 18, 5 figs.
1906. Echinodermen von dem norwegischen Fishereidampfer Michael Sars in den Jahren 1900-1903
gesammelt. III. Asteroidea.
Bergens Museums Aarbog, 1905, No. 13, 87 p., 2 pls.
1907. Echinodermata.
Report of the Second Norwegian Arctic Exp. in the Fram 1898-1902, No. 13, 26 p., 3 pls.
HAMANN, OTTO.
1899. Echinodermen.
Bronn’s Klassen und Ordnungen des Thier-reichs, Heft 2, Abth. 3, pp. 657-719,
Ka@aer, RENE.
1906. Echinodermes.
Expéd. antarct. frangaise (1903-1905) commandée par le Dr. Jean Charcot, Dec.,
pp. 1-27, pls. 1-4.
1907a. Note préliminaire sur quelques Astéries et Ophiures provenant des campagnes de la Princesse-
Altice.
Bull. l'Institut Océanographique, No. 99, Apfil, pp. 1-40.
1907b. Astéries, Ophiures et Echinides recueillis dans les mers australes par la Scotia (1902-1904).
Zoologischer Anzeiger, vol. 32, No. 6, Sept. 17, pp. 143-145.
1908. Astéries, Ophiures et Echinides de l’ Expédition antarctique national Ecossaise.
Trans. Roy. Soc. Edinburgh, vol. 46, pt. 3, No. 22, Dec. 6, pp. 529-581, 628-632, pls.

1-8, 12.
1909a. An Account of the Deep-Sea Asteroidea collected by the Royal Indian Marine Survey Ship
Investigator.
Echinoderma of the Indian Museum, Part V, Asteroidea (1). Calcutta, April, 143
pp., 13 pls.

1909. Echinodermes provenant des campagnes du yacht Princesse-Alice (Astéries, Ophiures, Echi-
nides, et Crinoides).
Résult. campag. sci. du Princ > de Monaco, fasc. XXXIV, April 30, pp. 1-136, pls. 1-23.
1910. An account of the Shallow-water Asteroidea.
Echinoderma of the Indian Museum, Part VI, Asteriodea (II). Calcutta, June, 191 pp.,
20 pls.
Lamarck, J. B. P. A. DE.
1816. Histoire naturelle des animaux sans vertébres. First ed., vol. 2, Paris.
LE1reoLpt, F.
1895. Asteroidea der Vettor-Ptsani Expedition (1882-1885).
Zeitschr. f. wiss. Zool., vol.°59, pp. 545-654, pls. 31-32.
Loriot, P. per.
1891. Notes pour servir & l'étude des Echinodermes.
Mém. soc. phys. et hist. nat. Genéve, vol. suppl., 1890, No. 8, p. 22, pl. 3.
1899. Notes pour servir 4l’étude des Echinodermes VII.
Mém. soc. phys. et hist. nat. Geneve, vol. 33, pt. 2, No. 1, pp. 12-29, pls. 2, 8.

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER, 389

Lorrot, P. pE—Continued.
1904. Notes pourserviral’ étude des Echinodermes, sér. 2, fase. II. Bale et Geneve, pp. 32-34, pl. 3.
1906. Rectifications de nomenclature.
Revue critique de paléozoologie, Paris, vol. 10, Jan., pp. 77-78.
Lupwiac, Huserr.
1897. Seesterne des Mittelmeeres.
Fauna und Flora des Golfes yon Neapel, vol. 24, 471 pp., 12 pls.
1900. Arktische Seesterne.
Fauna Arctica, vol. 1, Jena, pp. 447-502.
1903. Seesterne.
Résultats du voyage du S. Y. Belgica, Zoologie, p. 2, 7 pls.
1905. Asterien und Ophiuren der schwedischen Expedition nach den Magalhaenslinden, 1895-1897.
Zeitschr. f. wiss. Zool., vol. 82, pp. 39-73, pls. 5, 6.
1910. Notomyota, eine neue Ordnung der Seesterne.
Sitzungsber k. preuss. Akad. Wiss., vol. 23, April, pp. 435-466.
Lirxen, Cnr.
1857. Oversigt over Grénlands Echinodermata, Kjgbenhavn.
1871. Forsatte kritiske og beskrivende Bidrag til kundskab om Sgstjernerne (Asteriderne).
Vid. Medd. naturhist. Foren., Kjgbenhayn, pp. 227-304.
MorrTENSEN, TH. '
1903. Echinoderms from East Greenland. ;
Meddelelser om Grgnland, vol. 29, pp. 68-82, pl. 2.
Mixer, J., and Troscuet, F. H.
1840a. Monatsber k. preuss. Akad. Wiss., April, p. 102. (Not seen; title of article unknown to me.)
1840b. Ueber die Gattungen der Asterien.
Archiv fur Naturgesch., 6 Jahrg., vol. 1, Sept., pp. 318-326.
Miter, O. F.
1777. Zoologize Danicee Prodromus, Hainiz 1776.
Narpo, J. D.
1834. De Asteriis.
Oken’s Isis, Heft 3, pp. 716-717.
PENNANT, THOMAS.
1777. British Zoology, vol. 4.
Perrise, Epwonp.
1884. Mémoire sur les Etoiles de Mer.
Nouv. arch. du mus. d’hist. nat., sér. 2, vol. 6, pp. 127-276, pls. 1-9.
1891. Echinodermes de la mission scientifique du Cap Horn. 1. Stellérides.
Mission scientifique Cap Horn, Zoologie, vol. 6, 198 pp., 13 pls. Paris.
1894. Expéditions scientifiques du Travailleur et du Talisman, Echinodermes, 432 pp., 36 pls.
1896. Contribution A 1’ étude des Stellérides de ]’Atlantique Nord.
Résult. campag. sci. du Prince de Monaco, fase. XI, pp. 1-57, 4 pls.
Rerzavs, A. J.
1805. Dissertatio sistens species cognitas Asteriarum, Lunde, quarto.
Sars, M.
1856. D’une nouvelle Etoile de Mer, Astrepecten arcticus.
Fauna littoralis Norvegiew, Heft 2, pp. 61-62; pl. 9.
1861. Oversigt af Norges Echinodermer, 160 pp., 16 pls. Christiana.
1877. New Echinoderms.
Fauna littoralis Norvegiw, Heft 3, p. 65.
Scauzze, C. F.
1760. Betrachtung der versteinerten Seesterne und ihre Theile, v1 and 58 pp., 2 pls.
Stapen, W. Psrcy.
1879. On the Asteroidea and Echinoidea of the Korean Seas.
Journ. Linn. Soc, London, vol. 16, pp. 424-434, pl. 7.
1883. Description of Mimaster, a genus of Asteroidea from the Faerde Channel.
Trans. Roy. Soc, Edinburgh, vol. 30, pt. 2, pp. 579-584, pl. 34.

390 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

StapEN, W. Percy—Continued.
1885. Report of the Scientific Results of the Voyage of H. M. S. Challenger. Narrative of the
Cruise, I, pt. 2.
1889. Report of the Asteroidea collected by H. M.S. Challenger, Zoology, vol. 30, 893 pp., 117 pls.
Txomson, ©. WYVILLE.
1873. The Depths of the Sea, London, octavo (p. 120).
VerruL, A. E.
1867. Notes on the Echinoderms of Panama and the West Coast of America, with descriptions of
new genera and species.
Trans. Conn. Acad., vol. 1, pt. 2, pp. 271-293.
1868. Notice of a Collection of Echinoderms from La Paz, Lower California, with descriptions of a
new genus.
Trans. Conn. Acad., vol. 1, pt. 2, pp. 371-376, pl. 4.
1885. Result of the Explorations made by the Steamer Albatross off the Northern Coast of the
United States in 1883.
Rep. U.S. Fish Comm. for 1883, pp. 540-543, pls. 13-19.
1895. Distribution of the Echinoderms of Northeastern America.
Amer. Journ. Sci., vol. 49, pp. 127-141, 197-212.
1899. Revision of Certain Genera and Species of Starfishes with Descriptions of New Forms.

Trans. Conn. Acad., vol. 10, August, pp. 145-234, pls. 24-30.
VicuieEr, C.

1879. Anatomie comparée du squelette des Stellérides.
Arch. zool. expér. et génér, vol. 7, pp. 33-250, pls. 5-16.
Woop-Mason, J., and Atcock, A.
1891. Natural History Notes from H. M. Marine Survey Steamer Investigator.
Ann. and Mag. Nat. Hist., ser. 6, vol. 7, pp. 12-15, 427-433.

Fig.

Fig.

Fig.

_ aownr wh Bm whe BRON e Ch Se ol

ms onwre

oe, ONE

tr

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 391

EXPLANATION OF PLATES.

Puate 1. Eremicaster tenebrarius.

. Type; side view, showing cribriform organs.

Californian specimen with small disk.

. Alaskan specimen with few superomarginal spines.
. Specimen with recurved rays; California.

PLatE 2.

. Eremicaster pacificus; abactinal view.

. Same; actinal view.

. Same; ambulacral furrow, enlarged.

. Bremicaster tenebrarius; abactinal surface of disk enlarged; Californian specimen.

Pate 3. Ctenodiscus crispatus.

. Specimen from Alaska, station 4223.

. Specimen from the Sea of Japan.

. Narrow-rayed variety, Alaska, station 4235.

. Specimen from south coast of Hokushu, Japan, station 5039, 326 to 269 fathoms.

Prate 4. Ctenodiscus crispatus.

. Actinal view, Alaskan specimen.

. Four-rayed specimen.

. Actinal interradial area of dried specimen, enlarged.

. Paxille, enlarged, deep-sea variety, Bering Sea, station 3307, 1,033 fathoms.
. Specimen from station 4223.

. Specimen from Sea of Japan, station 4818.

Puate 5. Astropecten armatus.

. Specimen from San Diego, California.
. Same; specimen with smeller disk and fewer superomarginal spines (Beacon 3 Shoal).

PLate 6.

. Astropecten californicus; typical; abactinal surface.
. Same; actinal surface.

. Astropecten ornatissimus; actinal surface.

. Same; abactinal surface.

PLATE 7.

. Astropecten californicus; paxille and superomarginal plates of ray, enlarged.
. Astropecten ornatissimus; paxille of disk and basal portion of ray, enlarged.

Astropecten armatus; inferomarginal and adambulacral armature, base of ray.

. Leptychaster anomalus; abactinal paxille and superomarginal plates, enlarged.
. Psilaster pectinatus; side view of ray, enlarged, to show flattened infero- and superomarginal

spines; specimen from Bering Sea, station 3604.

. Astropecten armatus; abactinal paxille, enlarged, to show coalescent granules; from specimen

shown in plate 5, fig. 1.

. Psilaster pectinatus; type; side view of ray, enlarged; compare with fig. 5.

Puate 8.

. Leptychaster arcticus; abactinal surface, enlarged.
. Leptychaster pacificus; abactinal surface, enlarged; station 4194.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

one

one

_

— on eK bo

mem CO DD

mown re

or

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

PLateE 9.
Leptychaster anomalus, typical.

iL
2. Leptychaster pacificus; actinal surface, enlarged; station 4194.

35

4. Leptychaster arcticus; actinal surface enlarged, from off Hokushu, Japan, station 5047,

Leptychaster propinquus; type.

Puate 10. Psilaster pectinatus.

. Abactinal surface of type.
. Portion of same, enlarged, ray; superomarginal plates on right.
. Actinal surface, enlarged.

Puare 11. Dipsacaster eximius.

. Abactinal surface, typical specimen.
. Margin and adjacent abactinal paxille, enlarged; the upper end of inferomarginal plates on left.

Puate 12.

. Dipsacaster borealis; type.
. Same; marginal plates and paxille, enlarged.
. Dipsacaster letmophilus; marginal plates and paxille, enlarged.

PrateE 13.

. Dipsacaster borealis; actinal surface.
. Dipsacaster eximius; actinal surface.

Puate 14.

. Dipsacaster eximius; portion of actinal surface, enlarged; adambulacral plates on right, infero-

marginals on left.

. Dipsacaster borealis; same.

Puiate 15. Dipsacaster letmophilus.

. Abactinal surface of type.
. Actinal surface of type.

Puate 16.

. Dipsacaster anoplus; type, abactinal surface.
. Same; actinai surface, enlarged.
. Dipsacaster eximius; young, much enlarged.

PuateE 17.

. Dytaster gilberti; type.

. Same; two superomarginals, enlarged.

. Same; abactinal paxille, enlarged.

. Thrissacanthias penictllatus; young specimen, enlarged.

PrateE 18. Thrissacanthias penicillatus.

. Abactinal view of type.

. Portion of actinal surface, enlarged, type. ‘
. Inferomarginal spines, ray.

. Actinal surface of small specimen.

. Paxille of type, enlarged.

Prats 19. Luidia foliolata.

. Abactinal view of Californian specimen, considerably less than natural size.
. Actinal surface of one ray.
. Example from Queen Charlotte Sound, station 4203.

fe]
oe
i)

Fig.

Fig.

Fig.

Fig.

mow

em ooh ee One on

one

woe

one mone

sone

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 393

Pate 20.

. Luidia asthenosoma; type.
. Lwidia ludwigi; type.
. Same; upper end of inferomarginal plates, superomarginal and abactinal paxill#, enlarged.

Prats 21.

. Luidia asthenosoma; actinal surface of type, enlarged.

. Luidia ludwigi; same.

. Luidia foliolata; side view of ray of Californian specimen, enlarged.

- Same; similar view of specimen from station 4203 for comparison; outer inferomarginal spines

are shown.

. Same; portion of actinal surface of ray, enlarged; Californian example.

Puare 22.

. Benthopecten mutabilis; type.

. Benthopecten acanthonotus; type, abactinal view.
. Same; specimen with large disk.

. Benthopecten claviger; type.

Prate 23.

. Benthopecten acanthonotus; aactinal surface of type, enlarged.

. Same; specimen from station 2919; abactinal surface enlarged.

. Benthopecten mutabilis; actinal surface enlarged, type.

. Nearchaster pedicellaris; portion of abactinal surface of type, much enlarged.

Pate 24,

. Nearchaster aciculosus; cotype.
. Nearchaster variabilis; type; see also plate 119, fig. 1.
. Myonotus intermedius; type.

Puate 25,

. Nearchaster pedicellaris; type.
. Luidiaster dawsoni; Note: the abactinal primary spinules have been rubbed off the distal part

of ray; see plate 120.
PiaTE 26.

. Nearchaster aciculosus; abactinal surfaces enlarged.
. Same, actinal surface, enlarged.
. Nearchaster aciculosus, upper figure; Luidiaster dawsoni, lower figure, for comparison.

PuLaTE 27.

. Myonotus intermedius; type; actinal surface.

. Luidiaster dawsoni, for comparison; station 3320.
. Nearchaster pedicellaris; actinal surface.

. Nearchaster variabilis; same.

Pate 28.

. Pectinaster agassizi evoplus; type.
. Same; actinal surface.
. Benthopecten acanthonotus; young.

Puate 29. Odontaster crassus.

. Abactinal surface of type.

. Actinal surface.

. Portion of actinal surface, enlarged, to show teeth and adambulacral armature.

. Portion of abactinal surface, enlarged; the odd interradial superomarginal shown above, and a

primary basal plate below.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

mw dhe

mown re or, WO De me CO po

wm OO De

bo

bo

He co

BULLETIN 76, UNITED STATES NATIONAL MUSEUM,

PuatTE 30. Gephyreaster swifti.

. Abactinal view of largest specimen.

. Actinal surface of type.

. Abactinal paxille, enlarged.

. Actinal surface of ray, enlarged, showing adambulacral, actinal intermediate, and marginal

plates.

PuaTeE 31.

. Pseudarchaster parelii; specimen from station 3225.

. Same; superomarginal plates and paxillze, enlarged.

. Pseudarchaster parelii alascensis; typical specimen.

. Same; abactinal paxille, enlarged.

. Same; superomarginal plates and adjacent paxille, enlarged.

Puate 32. Pseudarchaster pusillus.

. Type; abactinal surface.

. Variety with shorter rays.

. Variety with large paxille.

. Type; actinal surface, enlarged.

PuateE 33.

. Pseudarchaster pusillus; abactinal surface of type, enlarged.

. Same; abactinal surface of specimen with large paxille, enlarged.
. Pseudarchaster parelii; typical, actinal surface, enlarged.

. Pseudarchaster parelii alascensis; actinal surface, enlarged.

PuateE 34. Pseudarchaster dissonus.

. Type; abactinal surface. 5
. Type; portion of abactinal surface, base of ray, enlarged.
. Portion of actinal surface of ray, enlarged, the adambulacral plates shown in lower right corner,

the inferomarginals in upper leit.

PuLate 35. Mediaster xqualis.

. Abactinal surface, specimen from station 4281.
. Portion of abactinal surface enlarged; Alaskan specimen.
. Actinal surface enlarged; Alaskan specimen.

PLATE 36. Mediaster tenellus.

. Type; abactinal surface, slightly enlarged.
. Same; portion of actinal surface, enlarged.

PLATE 37.

. Ceramaster japonicus; large specimen from station 3331; specimen B referred to in text.

. Same; specimen A from station 3330. :

- Same; portion of actinal surface for comparison with that of C. patagonicus below (fig. 4).
. Ceramaster patagonicus; portion of actinal surface for comparison with C. japonicus.

PLATE 38. °

. Ceramaster patagonicus; vicinity of Naha Bay, Behm Canal, Alaska, station 4228,.
. Same; specimen from Bering Sea, for comparison.

. Ceramaster clarki; type.

. Same; portion of actinal surface, enlarged.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

to

nue

moO he

mo he

1.

SIA. wr

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 895

Puiare 39. Ceramaster leptoceramus.

. Type; abactinal surface.
. Portion of abactinal surface enlarged, showing adradial secondary plates, smaller than the

primary plates. The same are shown in fig. 1.

. Actinal surface.

Puate 40,

. Ceramaster arcticus; large specimen from ‘‘Unalaska and Unga, Alaska,’’ (W. H. Dall), and a

small one, for comparison, from Round Island, Coal Harbor, Shumagin Islands. Note the
difference in superomarginal armature.

. Same; actinal and abactinal surface, enlarged, of two specimens from Shumagin Islands.
. Pseudarchaster parelii alascensis; very young specimen, much enlarged.

Puate 41. Cladaster validus.

. Type; abactinal surface, enlarged.
. Same; actinal surface, enlarged.

Puate 42. Hippasteria spinosa.

. Abactinal view of a typical specimen.
. Portion of abactinal surface of a dried specimen, much enlarged.
. Portion of actinal surface of the same specimen, enlarged.

PLATE 43.

. Hippasteria spinosa; abactinal surface of a specimen from southern California, station 3664,
. Hippasteria spinosa kurilensis; type, from Simushir, Kuril Island, station 4804. See also

plate 121, fig. 1.
Puate 44, Hippasteria heathi.

. Abactinal surface of type, enlarged.
. Portion of actinal surface, enlarged.

Puate 45. Hippasteria californica.

. Type; abactinal surface.

. Portion of abactinal surface of dried specimen, enlarged.

. Marginal plates, enlarged, the actinal surface uppermost, viewed from side.
. Actinal surface of a ray, enlarged.

Puiate 46. Hippasteria leiopelta.

. Actinal view of type, enlarged.

. Madreporic body, enlarged.

. Abactinal view of type.

. Portion of abactinal surface, much enlarged.

Puate 47. Cryptopeltaster lepidonotus.

. Type; abactinal surface.
. Same; actinal surface.

Puate 48. Linckia columbia.

Three stages of a ray developing a new disk. In the ray above the buds of the new raysare just
beginning to show; in the middle figure they are well started; the lower figure is a ‘‘comet”

form.

. Specimen from side to show the marginal plates.

. Actinal surface.

. Abactinal surface.

. Abactinal surface of portion of ray, enlarged, to show relative size of plates and papular areas,

. Actinal surface, enlarged.
. Marginal plates, denuded; also actinal intermediate plates to left, and abactinal on right.

396

Fig.

Fig.

Fig.

Fig.

bo

bo

bo

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Puiate 49. Dermasterias imbricata.

. Specimen from Barclay Sound, British Columbia.
. Left figure: portion of arm of specimen from Puget Sound; the integument has been stripped

from right half, showing plates; on left half, numerous pedicellariz, in the abactinal
integument.
Right figure: portion of abactinal wall of ray, dried and viewed from cclomic side showing
plates.
Puars 50.

. Leptychaster pacificus; two adambulacral plates, proximal third of ray, X 10. la. Same; abac-

tinal plates, papular area, at base of ray, from inside, 20.

. Leptychaster anomalus; two adambulacral plates from about middle of ray, X 10. 2a. Same;

abactinal plates at base of ray from inside, X 20; from station 3263.

. Psilaster pectinatus; fifth, sixth, and seventh adambulacral plates, medium-sized individual,

station 3307; spines have been removed from middle plate, X 10. 3a. Same; adme-
dian paxillee, base of ray, X 10. 3b and 3c. Same; two views of odontophore, < 10.

. Astropecten armatus; abactinal plates near base of ray, from inside, X 10. The large plates to

left (r) constitute the carinal or radial series of ray.

. Astropecten californicus; abactinal plates, from inside, > 10; 7, radial series.

Puate 51.

. Astropecten ornatissimus; sixth, seventh, and eighth adambulacral plates, type, X 12. Ila.

Same; abactinal plates near base of ray from within, < 20; r, radial series; ad, adradial
series; p, papula. 1b. Same; section of arm, X 6, showing the marginal, adambulacral
(ad), superambulacral (sa), and ambulacral (amb) plates with their dorsal serrate edges.
lc. Same; a large paxilla from base of ray of type, enlarged.

. Astropecten californicus; adambulacral plates, near base of ray, X 10; specimen from station

4559 having more pointed superadambulacral spines; ab, aboral enlarged subambulacral
spine. 2a. Same; adambulacral plate from near base of ray, type, X 10.

. Astropecten armatus; ninth and tenth adambulacral plates, x 10; specimen from Beacon 3

Shoal, San Diego, California.

PuateE 52.

. Dipsacaster eximius; ninth and tenth adambulacral plates, X 6; ad, adoral side of plate. la.

Same; paxilla from radial area at base of ray, X 10. 1b. Same; larger paxilla from near
center of disk, X 10. le. Same; abactinal plates from within, x 10; r-r, median radial
series. Note the more numerous papule about the radial plates.

. Dipsacaster borealis; eighth adambulacral plate, X 6. 2a. Same; one of the largest paxille

(except those on madreporic plate) from base of ray, X 10. 2b. Same; abactinal plates,
base of ray, X 10; r—r, median radial series, x —x, course of adradial muscle band.

. Dipsacaster letmophilus; ninth adambulacral plate, X 6. 3a. Same; one of the furrow spines,

6. 3b. Same; paxilla from radial region, base of ray, X 10.

. Dipsacaster anoplus; fifth and sixth adambulacral plates, X 10. 4a. Same; large paxilla

from base of ray, X 10.
PuateE 53.

. Thrissacanthias penicillatus; abactinal paxille from near base of ray, type, X 10. la. Same;

side view of a larger paxilla from disk, X 10. 1b. Same; eighth adambulacral plate,
type; < 6; the adoral side of plate is on the right. 1c. Same; eighteenth plate from
medium-sized specimen, X 6; ad, adoral side. 1d. Same; bifid marginal spines, type,
> 6; superomarginal to left, inferomarginal (,) to right. le. Same; abactinal plates at
base of ray, X 5; 71-7, radial series. j

. Dipsacaster leimophilus; abactinal plates, base of ray, X 10; r—r, radial series.

Fig. 3.

Fig.

Fig.

wn

soa

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 397

Dytaster gilberti; paxille from near base of ray, adjacent to marginal plates, showing pedicel-
larie, X 20. 3a. Same; fifth adambulacral plate and an isolated furrow spine from side
to show bend; adoral side of plate to left, « 6.

. Eremicaster tenebrarius; second to fourth adambulacral plates, % 10; specimen without supero-

marginal spines, from Southern Alaska, station 2859. 4a. Same; second to fourth adam-
bulacral plates, type, X 10.

Pate 54.

. Luidia asthenosoma; tenth adambulacral plate, and marginal, 10.
. Inuidia ludwigi; superomarginal paxille with pedicellarie and upper end of inferomarginal

plates showing pedicellarix, from near base of ray of type, ¥ 12. 2a. Same; adambulacral,
actinal intermediate plates (int) and portion of inferomarginal plate (inf) showing actinal
‘pedicellarie, < 10.

. Luidia foliolata; adambulacral plate (seventeenth) and an actinal intermediate plate int; bottom

of figure is adoral, * 5.

. Benthopecten claviger; third adambulacral and first marginal plates, showing also four actinal

intermediate plates; bottom of figure adoral.

. Benthopecten acanthonotus; fourth adambulacral and second paired inferomarginal, % 8. Jnt,

actinal intermediate plate, with spines forming part of pedicellaria; bottom of figure is
adoral.

. Benthopecten mutabilis; seventh adambulacral and fifth inferomarginal, x 8. 6a. Same; primary

basal spine and neighboring abactinal plates, typical, 6.

Puate 55.

. Nearchaster aciculosus; a primary plate on radial area of disk surrounded by smaller plates,

type, X 10. la. Same; mouth plates, specimens from station4402, * 5. 1b. Same; ninth
and tenth adambulacral plates, adoral side to right, & 8.

. Nearchaster variabilis; large specimen, station 3330; two primary plates from midradial region,

base of ray; a smaller secondary plate and three papule, 10. 2a. Same; tenth adam-
bulacral plate, type, X 10; ad, adoral side of plate.

. Invidiaster dawsoni; a primary and several secondary plates from near center of disk, X 10-

3a. Same; ninth adambulacral plate, specimen from station 4791, * 8; ad, adoral side of
plate.

. Pectinaster agassizi evoplus; abactinal plates from interradii between inner ends of two papularia

showing pedicellariw, enlarged. 4a. Same; adambulacral and inferomarginal plates show-
ing pedicellaris and spines, X 8; a, accessory marginal spinules; int, actinal intermediate
plate.

Puate 56.

. Benthopecten claviger; abactinal spine from base of ray of type, X 6. la. Same; end of spine,

enlarged. 16. Same; abactinal plates at base of ray, X 8; r-r, radial line.

. Benthopecten acanthonotus; abactinal plates from base of ray, X 8; r-r, radial line.

Nearchaster aciculosus; station 4353; abactinal plates showing the primary (pr) and secondary
(s) and interspersed papular pores; r-r, radial line, X 8.

Nearchaster variabilis; station 3330; abactinal plates from radial region, base of ray, & 8.

Luidiaster dawsoni; abactinal plates; r-r, radial line; pr, primary radial plate; p, papule, x 8.

Odontaster crassus; tenth, eleventh, and twelfth adambulacral plates, 10.

. Dermasterias imbricata; four adambulacral plates cleared of tissue, seen from above, the spines

bent backward; a, furrow spines; 5, actinal spines, which are always immersed in tissue.

. Asterina miniata; an actinal intermediate plate from near mouth plate showing serrate scar

(dotted), X 6. 8a. Same; abactinal plates from inside; median radial row in middle
(stippled), X 4.

bo

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

PuaTe 57.

. Pectinaster agassizi evoplus; papularium drawn from inner side, showing disposition of plates;
r-r, radial line; p, papule, X 8.

. Gephyreaster swifti; specimen from station 4784, mouth plates, showing the inner angle or face
toward actinostome, the peristomial membrane (per) below, the vertical suture between
the plates (s), the median tooth (d), and the peculiar marginal series of spines which
descend the sides of plate nearly to peristome; the spinelets nearest the actinostome are
the lightest; the strictly marginal series of spinelets bordering furrow are marked a, x 6.
2a. Same; median tooth of type. 2b. Same; side view of abactinal plates from radial
line near base of ray, type, X 10. 2c. Same; larger spines from inferomarginals, x 20.

. Mediaster tenellus; fourth adambulacral plate, * 10. 3a. Side view of furrow spine. 3b. Pedi-
cellaria on first adambulacral plate. 3c. Pedicellaria from an actinal intermediate plate
just back of mouth plates, X 10. 3d. A large paxilla from proximal radial region of dorsal
surface, X 10.

. Pseudarchaster pusillus; seventh adambulacral plate of type, X 10. 4a. Same; adambulacral
plates showing pectinate pedicellariz.

. Pseudarchaster parelii; specimen from station 3225, ninth adambulacral plate; specimen with
very coarse spinelets, < 8.

. Pseudarchaster parelit alascensis; twelfth adambulacral plate, type, X 9.

7. Pseudarchaster dissonus; third, fourth, and fifth adambulacral plates with adjacent pedicellariz;

ad, adoral side of plate; x, jaw of pedicellaria on adambulacral plate; int, intermediate
plate, the three small spinelets being on the actinal intermediate plate and bordering the
furrow, X 5. 7a. Same; second, third, and fourth adambulacral plates of specimen from
station 3601, to show slightly different pedicellariz; adoral side of plates to left, > 10.

Puate 58.

. Ceramaster arcticus; specimen from station 3212, two adambulacral, and two adjacent interme-
diate plates (int), small specimen, < 10.

2. Ceramaster clarki; eighth and ninth adambulacral plates, x 10. 2a. Same; actinal intermediate

~I

go

pedicellaria, near above plates, X 10. 26. Same; crown of radial plate showing pedicel-
laria, < 10.

. Ceramaster leptoceramus; adambulacral plate, X 10. 3a. Same; one of the radial plates with
three secondary paxillee from proximal radial region, 15.

. Cryptopeltaster lepidonotus; third and fourth adambulacral and an actinal intermediate plate;
adoral side to the left, * 5. 4a. Same; two primary plates from abactinal surface, radial
region, one with a tubercle, the other with characteristic bivalved pedicellaria; between
the two are numerous secondary ossicles, X 8.

. Hippasteria heathi; eleventh adambulacral and adjacent actinal intermediate plate, X 5. 5a.
Same; abactinal plates, proximal radial region; specimen partly dry and papulz shrunken,
X 5.

. Hippasteria leiopelta; two adambulacral and an actinal intermediate plate, partly dry, x 10.

. Poraniopsis inflata; three adambulacral plates from near middle of R, the furrow spines above,
and three views of actinal adambulacral spines, below; on the right the actinal spine is
bent so as to be parallel with furrow spines, of which the upper or grooved side is shown;
type, X 5. 7a. Same; from station 4362, similar view to above from a variety with slender
adambulacral spines, X 5.

. Poranionsis inflata flexilis; type; three adambulacral spines; f, furrow and a, subambulacral
spine, with flange of tissue; & 5.

PLATE 59.

. Mediaster xqualis; abactinal plates from proximal radial region, showing internal connecting
ossicles (dotted); p, papulse; r, radial series, X 8. la. Same; bottom of a radial plate;
1b. Side view of same. 1c. Same; eighth to tenth ambulacral plates showing the small
superambulacral ossicles dotted; am, ambulacrals; ad, adambulacrals; int, actinal in-
termediate plates; swp, superambulacrals, x 8.

Fig.

Fig.

Fig.

Fig.

Fig.

2.

—_

moe

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 399

Mediaster tenellus; abactinal plates, proximal radial region, showing internal connecting ossicles;
papul, lined; r-r radial series, 10.

. Ceramaster japonicus; abactinal plates from inside, proximal radial region, X 8.
. Pseudarchaster pusillus; abactinal plates from inside, near base of ray; r, radial series, % 10
. Gephyreaster swifti; specimen from station 4784, abactinal plates from inside; R, primary radial

plate; r, radial series, X 7. 5a. Same; abactinal plates of type from near base of ray;
r-r, radial series, X 5.

. Ceramaster clarki; abactinal plates from inside, showing the curious lobes; ad, adradial series;

r, radial series, X 8. 6a. Same; large adradial paxilla, showing the concavity at base
(dotted) and the pit of pedicellarian muscles (p.)

Puate 60.

All drawings are from the inner or ceelomic side.

. Ceramaster arcticus; abactinal plates from proximal radial region, showing radial series of plates

(r) and the differently formed dorsolateral plates on either side. These have the lobes
developed on the side toward radial line. Papule represented by small circles, % 8.

. Ceramaster leptoceramus; abactinal plates from proximal radial region, showing the radial series

(r) and on either side the small secondary plates, x 8.

. Ceramaster patagonicus; abactinal plates; pr, primary radial; r, radial series.
. Hippasteria spinosa; abactinal plates from proximal radial region, showing primary (pr) and

secondary (s) plates and papulw (p), X 5.

. Hippasteria californica; abactinal plates; r, radial plates; small secondary intermediate plates

surrounding them; ad, adradial plate; papulee shown as small dark circles, % 5.

. Hippasteria heathi; abactinal skeleton midradial region; the secondary plates are deeper than

wide and form a fenestrated reticulum; papulz cross-lined, * 5.

. Cryptopeltaster lepidonotus; abactinal plates; the larger are the primary plates; papulw, small

dark circles, X 5.
PuateE 61. Asterina miniata.

. Six-rayed specimen from Monterey Bay, California.
. Seven-rayed specimen, same locality.

Specimen from north of La Paz, Gulf of California, station 3001.

. Specimen from Clayoquot Sound, British Columbia, dried without first being hardened.

Puate 62. Asterina miniata.

. Apical region, denuded.

2. Actinal surface, specimen from Monterey Bay, California.

bo

oe co to

Puiate 63. Poraniopsis inflata.

. Type and two young; abactinal view.
. Type; actinal view.

Puate 64. Poraniopsis inflata flerilis, type, abactinal view.

Puate 65. Henricia sanguinolenta.

. Actinal view of a specimen from station 4803, and abactinal, station 4281. These are as near

typical as are any Alaskan specimens.

. One ray of a giant specimen, variety ©, from station 4778, and a specimen referable to the same

variety from station 4784.

Pirate 66. Henricia sanguinolenta, varieties.

. Giant specimen, variety C, station 4754.

Small portion of ventro-lateral surface of typical H. sanguinolenta for comparison with fig 3.

. Actinal surface of fig. 1, enlarged, showing marginal plates below; see also fig. 5.

Small portion of abactinal surface of fig. 5, enlarged (variety C.).

. Actinal view of aray from a giant specimen, variety ©, from station 4778; abactinal suriace

shown in plate 65, fig. 2. These are believed to be hybrids.

400

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

or km whe

bo

bo

bo

me whe

wn

aoe

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Puare 67. Henricia sanguinolenta eschrichtit.

. Specimen from Kadiak Island, enlarged.
. Single ray, specimen from Kadiak Island.
. Specimen from Bering Island.

Puate 68.

. Henricia sanguinolenta eschrichtii; specimen showing young grouped around mouth, enlarged.
. Same; a nearly pentagonal specimen from Attu Island.
. Henricia sanguinolenta, variety E. Three specimens.

Puate 69. Henricia leviuscula.

. Specimen from Puget Sound (variety A).
. Three specimens of the small variety found at Monterey Bay, California, and other places,

(variety F), enlarged.
PLATE 70.

. Henricia leviuscula; specimen from Puget Sound (variety A), enlarged.
. Portions of rays, enlarged for comparison a, Henricia leviuscula, variety A. 6, Same; variety

B, slender-rayed specimen from station 4205. c, Same; variety B, Monterey Bay, Cali-
fornia. d, Henricia leviuscula annectens. :

Puate 71.

. Henricia leviuscula annectens, for comparison with fig. 2.
. Henvricia leviuscula, variety A., Puget Sound.
. Henricia leviuscula annectens; actinal view, for comparison with portion of ray of H. leviuscula.

Prate 72. Henricia leviuscula multispina.

. Type; station 4779.

. Variety from Kadiak Island.
. Variety from station 4771.

. Specimen from station 2842.

Puate 73. Henricia leviuscula multispina.

. Type; portion of disk and ray enlarged to show paxille.
. Type; portion of actinal surface enlarged.

Puate 74. Henricia leviuscula dyscrita.

. Type; abactinal view.

. Actinal surface of specimen from off Wilmington, California, 47 fathoms.
. Aberrant variety from station 4420; abactinal view.

. Same specimen; actinal surface.

. Portion of abactinal surface of two rays, enlarged.

Puate 75. Henricia aspera.

. Cotype; actinal surface.

. Type; abactinal.

. Variety from California (station 4551).

. Cotype; portion of actinal surface, enlarged.

. Type; abactinal surface, enlarged to show sunken papular areas and small spinelets.

Piate 76. Henricia longispina.

. Type; abactinal surface, enlarged.
. Same; actinal surface, enlarged.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

mow he

iy
. Henricia longispina aleutica; type.

to

we oO be

to

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 401

PLATE 77.

Henricia asthenactis; abactinal view of type, and actinal view of cotype.

Puate 78.

. Henricia polyacantha; type.

. Henricia clarki; type.

. Same; type, actinal surface.

. Henricia polyacantha; type, actinal surface.

Piate 79. Lophaster furcilliger.

. Type; abactinal suriace.
. Type; actinal surface, enlarged

Puate 80. Lophaster furcilliger vexator.

. Type; abactinal surface.
. Type; actinal surface.

PuaTE 81. Solaster endeca, from off the Shumagin Islands, station 2851, Alaska.

PLATE 82.

. Solaster endeca; station 2851; portion of actinal surface.

. Same; station 4246 (S. galaxides Verrill); for comparison with fig. 1.
. Solaster stimpsoni; station 4205.

. Solaster endeca; station 4246 (S. galarides Verrill).

Pate 83. Solaster stimpsoni.

. Specimen from station 4205; actinal view.
. Specimen from Vancouver Island (Port Renfrew).

Specimen from Kadiak Island, Alaska.

. Portion of actinal surface, enlarged; Kadiak Island.
. Paxille of disk, enlarged; specimen shown in fig. 3.

PuatE 84. Solaster dawsoni.

. Specimen from Monterey Bay, California.
. Specimen from Cape Mendocino, California; paxille and portion of madreporic body.

PuLaTE 85. Solaster dawsoni.

. Specimen from Point Franklin, Alaska.
. Specimen from Straits of Juan de Fuca (Port Renfrew, British Columbia).

Puate 86. Solaster dawsoni.

. Actinal surface of specimen from Monterey Bay.
. Abactinal surface of variety from station 3466.

Puiate 87. Solaster pavillatus.

. Large specimen from station 3331.
. Paxille, enlarged.

Puate 88. Solaster pavillatus; actinal surface of specimen shown in plate 87, fig. 1.

57444°—Bull. 76—11——26

Pig.

Pig

g-

Fig.

Fig.

lig.

lig.

lig.

Pig.

Fig.

on rwnd

bo nw

one

onre

br

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Pate 89. Solaster parillatus, variety.

. Specimen from station 2853.
. Similar specimen from station 4792, showing actinal surface. Compare with plate 88.

Puate 90. Solaster exiguus.

. Type; actinal surface.
. Same; abactinal surface.

Puate 91. Solaster borealis.

. Type; abactinal surface.
. Seven-rayed specimen from station 3607.
. Abactinal paxille, enlarged.

PuatE 92. Solaster borealis.

. Two specimens of the Californian variety; abactinal surface of specimen from station 4528; actinal

surface of specimen from station 3343.

. Portion of actinal surface of type, enlarged.

PLATE 93. Solaster hypothrissus.

. Type; abactinal surface.
. Same; portion of actinal surface, enlarged.

Puate 94. Solaster papposus.

. Typical specimen; station 2843.

. Variety C from station 4788.

. Variety C from off Oregon, station 3051.

. Variety A from station 3518.

. Variety A from station 4795.

. Nearly typical specimen from station 2846.

Pate 95. Solaster japonicus.

. Type from station 4818.
. Specimen from station 4859.

Puate 96. Heterozonias alternatus.

. Actinal surface of specimen from station 4400.
. Abactinal surface; station 4400.
. Abactinal paxille, dried; station 3343.

Piate 97. Peribolaster biserialis.

. Portion of actinal surface of type, and abactinal surface of specimen from station 3317.
. Abactinal surface of type and portion of cotype, dried.

PLATE 98. Pteraster militaris.

. Abactinal surface of specimen from station 2846.
. Same specimen; actinal surface.

Piate 99. Pteraster trigonodon.

. Type; abactinal surface.
. Same; actinal surface.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

mow he

bo

oe Ne

wry

~

to

1
2

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 403

Piate 100.
. Pleraster marsippus; abactinal surface.
. Pteraster jordani; abactinal surface.

Puate 101.
. Pteraster jordani; actinal surface.
. Pteraster marsippus; actinal surface.

PLATE 102.

. Pteraster coscinopeplus; type; abactinal surface.
. Pteraster pulvillus; abactinal surface.

Pteraster coscinopeplus; type; actinal surface.

. Pteraster pulvillus; actinal surface.

: Puate 103.
. Pteraster tesselatus arcuatus; type.

. Pteraster temnochiton; abactinal view of type.
. Same; actinal surface.

Puate 104. Pleraster tesselatus.

. Actinal surface of small dried specimen from station 4205.
. Actinal surface of specimen from station 4222.

. Abactinal surface of specimen from station 3461.

. Abactinal surface of specimen from station 4235.

. Actinal surface of specimen from station 3222.

Puate 105. Pteraster obscurus.

. Large seven-rayed specimen, showing position of young escaping through supradorsal membrane
in interradius. This opening was made with a knife, and the young one placed in same
position as that referred to in text.

. Small six-rayed example; actinal surface.

and 4. Abactinal surface of two six-rayed specimens.

‘

Puate 106. Pteraster obscurus.

. Small eight-rayed specimen, enlarged (—/Pt. octaster Verrill).
. Actinal surface of a six-rayed specimen, enlarged.

Puate 107. Diplopteraster multipes.

. Large specimen from station 4784, abactinal view.
. Actinal surface of small specimen from station 4235.

Puate 108.

. Hymenaster kehleri; type.
. Hymenaster quadrispinosus; portion of supradorsal membrane, enlarged.

Puate 109. Hymenaster perissonotus.

. Abactinal surface of specimen from station 3307.
. Actinal surface, specimen from station 3074.

Piate 110, Hymenaster quadrispinosus.

. Specimen from station 2859.
and 3. Specimens from station 3603.

Ne

on

a

ro

~

un

BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

Puate 111.

. Dendrogaster arbusculus (new species); natural size; in specimen of Hippasteria californica.
. Henricia polyacantha; tenth adambulacral plate of cotype from furrow, showing six furrow spine-

lets; X 15. 2a. Same; seventh adambulacral plate of cotype, side view, X 15. 2b. Same;
dorsal skeleton, inner side, the individual plates not shown; the regular transverse rows of
papulz to left; radial (r) series, dotted, x 6.

. Henricia longispina; fourth adambulacral plate, from side, x 10. 3a. Same; inner view of

abactinal and lateral plates showing papular areas; swp, superomarginal, x 6.

. Henricia asthenactis; abactinal plates near base of ray to show papular areas and spinelets; r,

radial series, X 10. 4a. Same; adambulacral plates near base of ray, the spines being
turned adorad, showing the aboral side of each plate; /, furrow spinelet; inf, inferomarginal
spinelets; p, papulee, 10.

. Henricia clarki; ninth adambulacral plate, from side; three furrow spinelets, X 15.
. Henricia leviuscula; midradial (ray) portion of abactinal skeleton from the outside, x 6.

Puate 112.

. Poraniopsis inflata; abactinal skeleton from the inside of a specimen with unusually stout tra-

becule, showing two rows of papular areas between the midradial (car) and superomar-
ginal (swp) series of plates; ad, adradial plates; a, papular area of specimen from station
4471, where the granules are fairly numerous; b, c, showing the condition in specimen from
station 4362, where granules are less numerous, X 3.

. Poraniopsis inflata flexilis; abactinal skeleton, from inside, showing the very large adradial papu-

lar areas; ad, adradial series; car, midradial or carinal series; sup, superomarginals. The
primary plates are cross-lined. The dotted line shows the position of the adradial series of
intermediate ossicles of inflata, X 3.

. Henricia clarki; abactinal skeleton from within, radial region, base of ray. The individual

plates of primary mesh are not shown, but the lower secondary plates are dotted; papule not
shown, X 6.

. Solaster exiguus; two adambulacral, and adjacent marginal plates of type; inf, inferomarginal;

sup, superomarginal, 15. 4a. Same; two inferomarginal spines, enlarged. 4b. Same;
pseudopaxillze from proximal radial region, much enlarged.

. Heterozonias alternatus; cotype, from station 4400; adambulacral, actinal intermediate (int),

two prominent (infero-) marginal (inf), two superomarginals (swp), and four abactinal
plates (ab); p, papulee, X 8. 5a. Same; adambulacral plate of type, x 8. 5b. Same;
skeleton of dried specimen (station 4317), base of ray, showing the large papular areas
and small spaced paxillee; papules cross-lined, 8.

Prate 113.

. Solaster dawsoni; adambulacral, marginal and two dorsolateral paxille from specimen from

Monterey Bay, X 8; inf, inferomarginal plate; sup, superomarginal plate, paz, paxilla;
Pp, papula.

. Solaster borealis; station 3071; abactinal paxillz and papule, radial region, near base of ray, X 10.

2a. Same; adambulacral plate and adjacent marginal, type; int, intermediate spinelet, 8.

. Solaster paxillatus; station 3331; adambulacral, and marginal plates from basal third of ray, X 8;

inf, inferomarginal and sup, superomarginal plates.

- Solaster japonicus; station 4818; adambulacral plate, 8.
. Solaster hypothrissus; type; adambulacral and actinal intermediate (int) plate, X 8. 5a. Same;

abactinal paxille, type, x 6.

PuaTe 114.

. Lophaster furcilliger, radial paxilla, X10. la tole. Spinelets from abactinal paxille. 1/. Same;

superomarginal spine. 1g. Same; subambulacral spine, shallow-water variety, southern
California, & 35.

Fig. 2.

Fig.

Fig.

Fig.

oem © PD

ASTEROIDEA OF NORTH PACIFIC AND ADJACENT WATERS—FISHER. 405

Lophaster furcilliger vexator; two paxillar spines, enlarged. 2a. Same; two subambulacral spines,
type, X 20. 2b. Same; abactinal plates of ray, seen from inner side, the irregular plating
of the radial region shown at r, X 5.

. Peribolaster biserialis, cotype, dried; three adambulacral and adjacent inferomarginal plates,

X 15. 3a. Same; adambulacral and inferomarginal spines of undried specimen, station
3317, X 8. 36. Same; mouth plates, X 8. 3c. Same; two papule and fascicule of spine-
lets with enveloping tissue, station 3317, x 8.

Puate 115.

. Hymenaster perissonotus; station 3307; abactinal and marginal plates, to show the small size and

arrangement of former. The plates have lobes a trifle heavier than in typical southern
specimens, and in some specimens the plates are less joined along the radial region; ad,
adradial series; r, radial series; s, superomarginals; inf, inferomarginals, X 5. la. Same;
inferomarginal paxilla. 1b. Radial paxilla for comparison, taken on same transverse line
and drawn tosamescale, <6. lcetole. Same;superomarginal (c), adradial (d), and radial (e)
plates of topotype to show the extremely slender lobes, x 6.

. Pteraster trigonodon; spicules from supradorsal membrane, much enlarged.

. Pteraster temnochiton; spicules from supradorsal membrane, much enlarged.

. Pteraster marsippus; spicules from supradorsal membrane, much enlarged.

. Peribolaster biserialis; cotype; abactinal skeleton and fascicules of spinelets of dried specimen;

papulz shown in one mesh; 7, primary radial plate, x 8.

Puate 116.

. Pteraster marsippus; eleventh and twelfth adambulacral combs, the eleventh bent back to show

the small pocket external to aperture papilla; note the depressions in aboral surface of
adambulacral web; dotted line parallel to actinolateral spine shows where adambulacral
web joins membrane; depth of pouch shown by dotted line; ac, actinolateral spine; ap,
aperture papilla; p, pouch, 5.

. Pteraster trigonodon; adambulacral armature, type; ac, actinolateral spine; ap, aperture; p,

papilla, X 10. 2a. Same; mouth plates; complete armature of one plate only, x 6.

. Pteraster temnochiton; paxilla seen from side, showing also interior of nidamental cavity; ¢, embryo,

actinal side up; p, papula, x 10. 3a. Same; adambulacral armature; lettering as in fig. 2.
3b. Same; suboral spine, 10.

. Pteraster coscinopeplus; mouth plates, < 6.
. Lophaster furcilliger; fourth and fifth adambulacral, an intermediate (int), and second inferomar-

ginal (inf) plates, < 10.
Puate 117.

. Pteraster jordani; adambulacral armature and part of adjacent actinolateral spine, * 5. Ia.

Same; abactinal paxilla with papula; the longer spine s is broken, x 5. 15. Same; mouth
plates, specimen from station 3346, X 5.

. Hymenaster quadrispinosus; station 3603; adambulacral armature, large Alaskan specimen, X 6.
. Hymenaster perissonotus; topotype; armature of third adambulacral plate; spines have been -

removed from second; X 6. 3a. Same; topotype; mouth plates, X 5.

. Hymenaster kehleri; mouth and adjacent adambulacral plates, X 3.

Piate 118.

. Myonotus intermedius; mouth plates, X 6; some of the suboral spines removed from left plate.

la. Same; fifth and sixth adambulacral and third inferomarginal plates, X 6. 16. Same; abacti-
nal paxilla. 1c. Same; abactinal paxilla from side. 1d. Same; sixth superomarginal plate
with spines, from side, X 6. Je. Thirty-fifth adambulacral plate with three furrow spinelets.
1f. Fourth adambulacral plate, one furrow spine, X 6.

. Nearchaster pedicellaris; mouth plates of type, X 6; some of the suborals have been removed from

left side.

_ Nearchaster aciculosus; sixth superomarginal and inferomarginal from side, showing the armature,

x 6.

406 BULLETIN 76, UNITED STATES NATIONAL MUSEUM.

PuaTE 119.

Fig. 1. Nearchaster variabilis; abactinal surface of a specimen having the maximum number of papule.
Luidiaster dawsoni; abactinal view of a young specimen, enlarged.

bo

Puate 120. Lwidiaster dawson.

Fig. 1. An abactinal view of a medium-sized specimen, from station 4792, enlarged.
. Actinal surface of same specimen, enlarged.

bo

Priate 121.

Fig. 1. Hippusteria spinosa kurilensis; type, actinal surface.
Hippasteria leiopelta armata; type, actinal surface.

bo

PLATE 122. Hippasteria leiopelta armata.

Fig. 1. Abactinal surface of small example.
2. Abactinal surface of type.

MUSEUM

NATIONAL

S.

U

“ meaty cava
cs moty
y: | a

pcre fa :

Spe
4 r

eres
>>. » Nhs ' an
a . ig ps i. pir

Rand

ae

ne ee

eT -_ ye
fe ¢ A ial “%

RR fa

Wise

ES SETS
oie

Rate ig) ‘
Oe E
NS eS

eyvierwreonk
Sole doi has we 3
SL lt Rd ese ta ape
ets 4 po eg > hale
AO irk pee

>
.
>

%atas
Pe TAL Sry
4 4%
rr
a
Ort ’
Th J

ey hd Pee Tee
Yom :

Ks Ay z é
SA a Scan sien
a

, iT ee ey

) wi}

. vs ee

ar %
a Se

bs Sie
Baise RmEON eS t/a

4 +. ne
Ae ee he
OO) ae

ae

Pd Ah ea EP SL.

at ° Be

mans SP

=

ANN a Ses
er ae Re
aan ee
en re
Sat kn,
rs es ‘a - 5 ‘i Ay Ab

DLs
ae

383

Ristt i)

tEd 7,

tf? -
Aad hi
: oe

vr; #4
:

re
ail

aT RM)
us

I

IO POSELLT

oe DP

s
Pret ee st
eee) So Py

eh Pee. ia Smee Be ye oy ry

ove
Pa epee! ?
Le) Pare Chee ss
7 , A

yD
a
-
z,
co
A
~
a
é

oe ans

gd TT HT BS

on?

rey aS at —~—S

we

aati
Bae sata ey
~ 4

2 ae

—

eZ

BONS.

os

A
Ny '
*

of

wv

ae

es
a TOF A
7 oe iy MY r. Ores
argh. obs na SO ee
Pod ETS vires st .
tx ote Be (er
SY ERR
: es oa!

|. AAS
Nhe
see

4:

5

v

an
ate a ew oe)
SS Pa Rie Se

Le

b ‘7 . »
uve © a ae yt |
| \ ° i oe °

mh

aft

A

N

LS)

Lt Rd
LT hd
a 7

ee ee
-

os ef*aw*
td od rr

eo -
ud ded
Ae a

od Sd

WeSes re
OT a eres ett
Pa , es > .

.
we?

<a tg eeg hia Ts

kay eactdg

s
866% Say >
ee eA

3 ar
ee dat Ve

Se et |

—
~«
7

J A”
ete ae

bn die te di ti te
oe Oe ees My >
ke

_

ow <

.

gee fees a 8 et SS
LT SEP ete ee

oa

]

(Aad

Oe
)

‘
ahs:
ee

“**eaeQ*, .

" to
eee te ie

ro wre ay rs é
betel, Pelt, atta bef

|

n
oe “i
he . OP a a
See PT
thet iw haa

>
w
D
>

INA

NATIK

' <<

-
»
’

I

ie

=

WS Ph

we es
ED
és .

~
at rd
a
wr
f

wy
~ ©
vo
si
at x |

a

ae ed ty
_
ie eek ole ihe i

=
i)

> eo
gt EEN ~

U. S. NATIONAL MUSEUM

1. LEPTYCHASTER PaciFicus, 2. L. ANOMALUS. 3. PSILASTER PECTINATUS. 4. ASTROPECTEN ARMATUS
5. A. CALIFORNICUS,

aot 396.

U. S. NATIONAL MUSEUM BULLETIN 76 PL. 51

FOR CAPLANATION OF PLATE S08 Pace 396,

—S oe | eee

U. S, NATIONAL MUSEUM

1. DiPSACASTER Eximius, 2. D. BOREALIS. 3 D. LATMOPHI

5. 4. D. anopius
Fom CaP ANATION OF PLATE OOF ma 196,

7s

U. S. NATIONAL MUSEUM

1. LUIDIA ASTHENOSOMA, 2, L. LUDWIG! 3. L. FOLIOLATA. 4. BENTHOPECTEN CLAVIGER. 5. B. ACANTHONOTU
6. B. MUTABILIS

U. S. NATIONAL MUSEUM

BULLETIN 76 PL. 55

1. NEARCHASTER ACcICULOSUS. 2. N. VARIABILIS. 3. LUIDIASTER DAWSON! 4, PECTINASTER AGASSIZI EVOPLUS
; ‘ . Tt ett pace 397

U. S. NATIONAL MUSEUM BULLETIN 76 PL. 56

1. BENTHOPECTEN CLAVIGER. 2, B. ACANTHONOTUS. 3. NEARCHASTER AcicuLOSUS, 4. N, vaRiaBiLis. 5. LuIDIASTER
DAWSONI. 6. ODONTASTER CRASSUS. 7, DERMASTERIAS IMBRICATA. 8, ASTERINA MINIATA

FoR EXPLANATION OF PLATE ere Pace 397

U. S. NATIONAL MUSEUM

1. PECTINASTER AGASSIZI EVOPLUS. 2. GEPHYREASTER SWIFTI. 3. MEDIA TER TENE 3. 4, PSEUDARCHASTER
PUSILLUS. 5, Ps, PARELI. 6. PS. PARE ‘ ' 0 NUS

U. S. NATIONAL MUSEUM

1, MEDIASTER €QuALIS, 2. M. TENELLUS. 3. Ceram TER JAPONICUS, 4, PSEUDARCHASTER PUSILLUS,
5. GEPHYREASTER SWIFTI. 6. CERAMASTER CLARKI,

FOR EXPLANATION OF PLATE SEE Pacts 398 ano 399,

U. S. NATIONAL MUSEUM BULLETIN 76 PL. 60

1. CERAMASTER ARCTICUS. 2. C, LEPTOCERAMUS. 3. C. PATAGONICUS. 4, HippasTcRia sPINOSA. 5. H. CALIFORNICA
6. H. HEATH. 7. CRYPTOPELTASTER LEPIDONOTUS.

Fom CXPLAMATION OF PLATE 6 Pace 999

U. S. NATIONAL MUSEUM

I!

I}

U. S. NATIONA

=
D

i Y

U. S. NATIONAL MUSEUM

U. S. NATIONAL MUSEUM

i

Pi Io

UM

MUSE

S. NATIONAL

U

JSEUM

INAL M

NATK

Ss.

EUM

MI

NATIONAL

Ss.

j

U.

U. S. NATIONAL MUSEUM

¥ iad,
ehh sae ta > bee he”
wy eh eee * 2

st ete N ae

=

4 ree :

eee eek
; ¥ r a C2

ooo

U. S. NATIONAL MUSEUM

<
z
=
<
z

U

oo),

!

Me

we

ey ied

c
Li

NY ad ea

OL
Port

Fos
1 ; OR ies)
zt = Pla Seth new 4 ‘yr ts A rs Peery Yr’ Can
LE Pa DY ID Toes
\ . M4 ms a ROT .
; A
+” ano , ACS ya “4
Sa wer Aor . A aa
~ Pe a hae

ath NOE ad re ey
sd 4 eis 4 Ata

he a.

ave»? en 3
ete ee Altes
pen

Ad
2 ro]
Aa PP ns

fat oe Wide

NYRR Ae

hk! Ny ! ¥ Rt
a} Tota % ot nee ee . ’ ee : 2

SS: iii

eae -

4 SHS
F ~ «3 Tae RAN SOSH
oo * cpkomisiesdins anee SSS eee Ss
\ a

CE ae eae’
wish Nit 6 ataeere La ons EEE Lic La ag ie
2 wi gai ROC GT GEE

WL

a al

4

a
a
Ne "
. -
x .
r
NEE

Si, hi. KE ale
CARE ee —— ~
owegixe EL ey:
Tat a Sane ke
ie YL ti es SS
_ . ; he SN Se aA

aS
“4
a!

ts
7...
|
-

Ns.

om i!
a] i
a i
zx <
4
‘a =
j
a
,
a
62 ee
- te
»
. .
.
=
af
“a
-

;

ie

a

~

Bey.
et
.
<n

7
Ai
Aa

”

, Ma Re eee

. Peete’. 4 5e
Sea

U. S. NATIONAL MUSE M

.

Eee at
OM INT Cetin

~<
bY

wy

=
<
z
4
z
vi

rt

U. S. NATIONAL M EUM

U. S. NATIONAL MUSEUM

<
<
z
a
&
z
Ie
Sted
o
,_o
<2
=>
- w
xz ad
mote
<x
as
ae
ar,
<x
Ca
cu
zo
w.
Ir
No
o
>
os
>
3
o
=
o
ac
<
«
aw
-
”
<
=
oS
z
Ww
fa)
=

. S. NATIONAL M
U. §& ONAL MUSEUM BULLETIN 76 PL. 112

1. PORANIOPSIS INFLATA. 2. P. INFLATA FLEXILIS. 3. HENRICIA CLARKI. 4. SOLASTER ExiGuUS. 5, HETEROZONIAS
ALTERNATUS.

FOR EXPLANATION OF PLATE Off Pace 404,

U. S. NATIONAL MUSEUM BULLETIN 76 PL. 113

SONI. 2. S. BOREALIS. 3. S. PAXILLATUS. 4. S. saponicus, 5. S. HYPOTHRISSUS.

1, SOLASTER DAW

FOR EXPLANATION OF PLATE Set PAGE 404.

U. S. NATIONAL MUSEUM

BULLETIN 76 PL. 114

a

<
© S% co Cece ce 0 af

= 20m

a SSS a SD
SS SO Sa SS TIT na HE
- = — Sa SS ee
SIRRIIQ DEES SE SSM nn SEE
—

2

la

os * * Se 9q59000"5')
—= ssp

1. LOPHASTER FURCILLIGER. 2. L. FURCILLIGER VEXATOR, 3, PERIBOLASTER BISERIALIS.

PLATE ate paces 404, 405,

or

FoR EXPLANATION

U. S. NATIONAL MUSEUM

1. PTERASTER MARSIPPUS, 2, PT. TRIGONODON 3. PT. TEMNOCHITON, 4, Pr. «

F EXPLANATION OF PLATE SEE PA 405

U. S. NATIONAL MUSEUM

a

U. S. NATIONAL MUSEUM
BULLETIN 76 PL. 118

1. MYONOTUS INTERMEDIUS. 2. NEARCHASTER PEDICELLARIS, 3. N. ACICULOSUS

FOR EXPLANATION OF PLATE #8 PAGE 405.

INAL MUSEUM

U. S. NATK

U. S. NATIONAL MU

ee =
SSy
= is

>

re

U. S. NATIONAL MUSE

INDEX.

The following index contains the names of the orders, families, genera, species, and subspecies of

Asteroidea which occur in this memoir.

Two kinds of type are used for the names, roman and italic;
the former indicates valid names, the latter synonyms.

But when a species name follows a genus name

that is a synonym of another genus name, both the genus and species names of the combination are itali-
cized, although the species name may be valid. Two kinds of type are used in the figures referring to
the pages; the heavy-faced type indicates the pages on which descriptions may be found. The names
occurring in the explanation of plates have not been indexed, as the easiest method of finding all the
figures of a given species is to turn to the page on which the description occurs, where a list of figures is

indicated.
Page.
abyssicola, Plutonaster.........-...- 8, 38, 72, 76, 77
abyssorum, Pteraster...............-...... 344
BRPTRIERE eee Soc elec seceeee.. PAD
PILLAY CHOSE oo 28 SS. 7,121, 127, 140
Acantharchaster aciculosus..............- 2, 10, 133
Acantharchaster dawsoni........-..-.-.----- 128
Acantharchaster intermedius............----- 10, 141
Acantharchaster variabilis..............----- 10, 137
Acantharchaster variabilis ag Seeeee 10, 138
PE UMANCO Leena S220... . 252, 258, 259
ERTHARL CLIC SE See ss = soe. 2. ---- oe se 5, 252
PIRMIEUNECTAVER Sema see ee Siac le eee cece | DOL
acanthonotus, Benthopecten.-.............-- 8,
1], 143, 144, 148, 149
acanthostoma, Asterias...................... 10
(co hie, ZACH Ea 6
aciculosus, Acantharchaster..........-.- 10, 132, 133

aciculosus, Nearchaster...........

Wesdontitter._.. . 154,158

exqualis, Merete! ie 6
sequalis, Mediaster. 6, 12, 15, 196, 197, 198, 204, 205
INO, 2 a S95)
npinasrentagonaster..<....-.--.-.--------- 165
BERTEWULETASLGI=-------2=----------.---- 368
Gpassizii, Mediaster...............2..-.....- 197
agassizi, Pectinaster 123, 127
RE EMENMC ICM CSer Ee e228 ele eee ee 71
alascensis, Pseudarchaster...........--.----- 8, 185
alascensis, Pseudarchaster parelii..... . 2,15, 186
alaskensis (var.), Asterias epichlora.......-- 10
PERE ROS PEL se eae en bok eee eee 23
alboverrucosa, Asterias..............---+--- 307
aleutica, Henricia longispina. ..... 13, 15, 268, 300
PER CErinie eee ee i Le Sl ee 266, 210

57444°—11—-27

12, 15, 188, 141 |

Page.
Alexandraster inflatus.................-....- 8,261
Alexandraster mirus.............----- 260, 261, 263

Allasterias rathbuni........ recon eee 9
aivernatus} Crossastercseace- a oe eee 9, 330
alternatus, Heterozonias.......... 13, 14, 321, 331
alveolatus, bterasters.. oss sescee cc es eee ee 370
ammophilus, Spheriodiscus................ 167
Amphiastenssso-52- S22 2s 159, 161, 162, 168, 170
Am phiasher jnSipnis: 2: <1. ssssa teases 168
(ATSB LORIE Wace. ~ See ceet as ta ceneee eee 9
rAnaslerds soe) OSS ae ee aoe 9
andromeda, Astropecten.. See oii 71
annectens, Henricia ievaiee hee Sree: 10,

13, 14, 267, 268, 291, 295

ATL OCA SOUR a aoe iaicate Aa ta nen a 31
anomalus® Glyphaster.<- 2-5. 0s ens ase sees 48
anomalus, Leptychaster..................- 9,
11, 12, 15, 43, 48, 53, 54, 100

anoplus, Dipsacaster.......... 10, 12, 14, 86, 90, 97
Amséeppodanerncs sos aidce shen wit a see ee
IATISELO POOLED Hesee nse son was aos die aee ee ere 5, 253
antarcticus, Benthopecten............-...-- 149
PANTHEON GR $A oaje a's oe ane sehen 21, 159, 161, 163, 174
AMLNENAGRS anno once cn senin 4, 158, 159, 160, 161, 223
Azitheneinse) 22 2:-... 35252 eee SD, ee, LO eee
WA mthenvaster seeks Sas feedak Cen 163, 169, 173
Antheniaster epixanthus.............++++++- 169
Anthenoides............ 160,161, 163, 168, 169, 173
‘Atphroditaster. 2.5. -c2-.cacns= =~ 160, 161, 163, 170
aporus, Pteraster.......-...... 7, 10, 344, 346, 347
Arbacia stellata. <-..<++4-+<.2-sucukemeeues 259
arbusculus, Dendrogaster................-- 404
Archaster. . 7, 20, 37, 121, 158
Archaster are hie. 43

407

408 INDEX.
Page. Page.
Archaster DatT Qs a= << erases l= laa clnte sine iets 1,96) |) -Aisterzasiqnantlariss]-2-2-2 2-2 oe eee 168, 204
Archaster dawsont.....-------------- 7, 8,127,128 | Astemas helianthemoides:—.-..2 22) eee 325
Archaster magnificus....-..----+----------- 4] | Asterias heltanthoides..----=---- +e eee 5
Archasten DareUve oor ners 2 anes = eerie 180 || Asterias hexactiss.- <2-- 5 -a---5se- eee 6
Archaster parelii longobrachialis ..-.-.------ 180) | Asteniasyanthing..- 22 S222 ses eee 5
Archasteridse......-- 4, 18, 20, 22, 121, 153, 159,163 | Asterias katherine.........-.-.--......22 57
SAT CROSLEV ICR Ren eee eee eee 372 38i|| _Astenas kellert:---2-----2---e eee 7
arctica, Leptasterias......-.-----------<---- i | -Astemas ithenttc. sos 4-2 ae =e ee 6
arctica; Tosia-2se2- 4. aa =- ee ee ane te elo 93/219) || Astemasimacnopora:ss-2.- sce = eee 10
arcticus, Archaster.........----2---++=+--+=- 43 | Astertas milttaris...---.--....-.------.--- 344, 346
arcticus, Astropecten......--------+-+------- 43),| “Astentasiminiata..- \se= =~) =- eee 5, 254
arcticus, Ceramaster.....----- 13:15, 205,206; 209))|| Astertasiminutascee 22 2252s ee eee 254
arcticus, Leptoptychaster.....-------+-++++++- 43) || Asterias obtusanguluss. S22 202-2 ose eee 163
arcticus, Leptychaster... 11, 42, 48, 45, 52, 53, 347 “Astentas OChTaceas 2222). +s225=2 2 ee eee 5,6
arcticus, Pseudarchaster tessellatus....------- 180))|'Asteniasioculatas.. 22. 32-5225 271
arcuatus, Mediaster ........--------- 197, 199, 205 | Astertas papposa-...-....2-22.22--2sesee 306, 325
arcuatus, Pteraster tesselatus. 13, 14, 345, 363, 369 Asterias (Pisaster) papulosa..........-.-... 9
articulatus, Astropecten.......------------- 67 || Astenias: paucispind-- --_---=-2-22-e eee 6
armata, Hippasteria leiopelta...-..-..-- 135280)\| Astemas pectinata &.o22 eee eee » So ee 5
armatus, Astropecten......--------- 1,14) (6567 | Astervas\pertusa.....- = 223. - ee eee 271
armatus, Astropecten....--..-.<+2-<->----=\-- 60))|| Aistervas pleyadella..-- 52-5. eee eee 162
aspera, Asters. - mnie reels iat 307) |\-Aistemas' polaris. .....--2--- 545-2 eee eee 31
aspera, PlCnrICla << .s<<2rs\-ceeee eee esas 9; || Asteriasipolythela.....------.--5--eeeeneeee 9
13, 15, 267, 270; 272,287, 289,293 | Asterias rubens-..-......-.----2- s-5---ee 6
Asteracanthion margaritifer......----+------- 5 | Asterias saanichensis............----.----- 7
Asteracanthium ochotense...----------------- 6 | Asterias sanguinolenta...........-.------ 266, 271
Agterias: <2 -2c-2-eds oso eee eee 368: ||, -Asterzas sepostia.- 2-1... 4-4-2 ee 271
WAlSteriOS ce acawee eee nee Sct Scere eee 55 | -Astemasispinosa.5...-.- 5-2-2 -he eee 261
Asterias acanthostoma.......-.------------- 10)'| -Astertas!spongiosa.... . === -)- 244-5 271
WA ane le nih mececeuceossonaces ae AcaacSac 6: | Astemas tessellata: +... - 55-2222) 159, 167
WAStertas: 2OUGIISSSasree ea ee ae ets 6:\|| Astemas troscheltt. <<. 25-2503 ee eee 6
Astemasiapinises- -.<-4e20=o)o2 eee eer 5,325 | Asterias troschelwi var. rudis......-..---.--- 10
Asterias alboverrucosa ...-.---------------- 307 ||, Asterias -vancouveri:-..-------=- eee oe 7
“Astemas G8 Pend sonec nce e sae ee es 307) | -Astemas vernicina.—<..22 2 = 92 ee 247
Asterias brachiata.-- -22- -2 2 asses 6’ |, Asterias victoriana..<..<.255 22-4465 seen 9
tAlsteriasDTeviS DUNG ae se elas aia eie aeieiete =F 6 | Asterina, 2.2... ---2-- s-- <2 4, 8, 158, 159, 253, 254
Agterias Camehsticar. 2222 telecine 5) |eAsterinaminiata: = 22 oss eee 11, 18, 14, 254, 258
AS tenias CO PILOtO asec. see eae aes eee 6 | Asterina pectinifera_............--.----- 257, 258
VASLEMGS COMMU END ans = sass eee 248))|| Asterimidse:- 24. 92-6 --13---e ees 5, 159, 251, 253
Asterias columbiana........--------------- 10 || -Asterininse: -.-<....s2:2--- = 5-- e eee 5, 253
ASLENUGS: CON] ELL etnies =e eee 6) |eAstemscus:—---- 2c == a 254
TAS LEPLOSTCRUON GNU hae a2 nein ae ir 6) || Asteniscus\ militants. ----. 3-2-0222 346
ASEETUAS CTUS DOLD. = acres = aia = 2 = eee ee 31 | Asteriscus miniatus. ...- --2-2--4-e eee 254
Astenas COUP asic. os osce-e= eee eee 7 \wAsterod ON: as -=22-.- oe eee 153, 154
AisteriagdUbiaas sss cca ere ee eer 10 | Asterodon singularis..........------------ 153, 157
Assterias endecds- ase sen ss2ee soe ae el 306;,307) |pAsteropess-2 eeseseee ee 4,8, 158, 159, 247, 248, 250°
Asterias endeca var. decemradiata .......-.-- 5,311 | Asteropide....-- 4,5, 18, 20, 22, 158, 159, 247, 251
Asterias eprChlOnds.« =.24-2seniso eee ae 56) |WAstenopsiss. oon 2 eae nee eee 8, 247, 248
Asterias epichlora var. alaskensis.-..-.----- 10 | Asteropsis imbricata. . .- - Slade eC 6, 248, 249
ARLEN ABN ISSTS DUN Oe oles aie ome oe eee ater 6 | Asteropsis vernicina..........-------------- 247
Atenas forcipulata. ss... 26-236 5-35-95 10) |\-Asthenactise----.-. 22. Genoese 253
Asterias fOrnentmsccp eae oan eee eee 7 | asthenactis, Henricia ...-. 10, 13, 15, 267, 270, 297
Asterias Qibb08G 22 sahing ses =2ts ee ee ree 254 | asthenosoma, Luidia......-.---- 9, 12, 14, 106, 116
Asterasiqigantea 22-2 oe eee eee a 6) |) Astroceramus:.-.<-- 2-4-6 -5eee= eee 162, 172

INDEX. 409

um Page. Page.

EM ERIIIIANTER mo swim nis SSS ae sid els aie 162, 163,179 | Benthopecten pedicifer.................. 149, 152
Astrogonium........-.----.++-+2seee eee 166,179 | Benthopecten semisquamatus. .. 150
Astrogonium granulare............-.----.- 214 | Benthopecten simplex. ....- ’ 143
Astrogonium meridionale...........--------. 153 | Benthopecten spinosus. . . . 142, 146, 147, 148, 149
Astrogonium miliaris..............2.22.22- 153 | Benthopecten spinuliger................... 152
Astrogonium parelit............--..------- 180 | Benthopectinide......... 4, 17, 20, 21, 22, 37, 120
-Astropecten.... 18, 20, 38, 39, 40, 41, 53, 55, 79,117 | bifascialis, Phataria (Linckia) unifascialis.. 242
Astropecten andromeda...........2.2-2.02+- 71 | biserialis, Peribolaster ............ w+. 8,13, 15, 841
Astropecten arcticus....-.......--.-....--- a5) Blakiaster=.- soc ~-.-~ eee ee a 17, 38, 39, 40, 84
Astropecten armatus............... 11, 12, 56, 67 | borealis, Crossaster........................ 9
RE ROMECLEMIATINREME. cc <ineiee wales aes cee soe 60 | borealis, Dipsacaster... 10, 12, 15, 86,91, 95, 96, 97
Astropecten articulatus.................... 67 | borealis, Solaster.... 13, 15, 307, 317, 318, 820, 329
Astropecten aurantiacus................... 55 | bourgeti, Sphzeriodiscus .................. 167
Astropecten californicus.... 8, 11,12, 14,56, 61, 71 | bourgeti, Stephanaster.................... 166, 167
ABurOpecten duplicatus............22.:--... 60 | bowersi, Nereidaster...................... 199
PASEO PECLEM CFINOCEUS. . = 22-2 se-- 222s D6: (Orachtatan Astentas 2. .<22-.2J15 ane saeeee 6
PARIMOVECUCMISTACTIS. «2.2. -.oeccce cee ee ee G77 || brenspina, Asternas--.......-...20.ceeseeeue 6
BRO MECLON ELEN < = = Saye wane ew des cece 43 | brevispina, Luidia................... 107, 109, 112
Astropecten ornatissimus ......... Ost2 1d. FONG. || WOMISLUPAOXING =e an oes ccul sue Seen 8
BER OPECLEN ONSLEUU << — o.-02 os 222s ee ecnc ees D607 || WBUDasteR. | nc. cdo coheed aaa ee
ANC De OG 1d 180518 || Bunodaster ~ 22.22. . hcec cons ees eee PLD
Astropecten polyacanthus...........-.----. GL) Bunodaster:ritlert << — < sonc <n = an eee ee en
PANINOVEGLEM TOPALINS 2 =<... 5... eee ee ese cee 67, 71 | californica, Hippasteria ...................-.- 8,
Astropecten rubidus................-..---- 67, 71 13, 14, 163, 224, 225, 238, 234, 264
PAREMOPECUOUGVEDUN <<... 22. cce cece sees ce G7. (Renlsjormtca MING. =. <<cn const vee eee 7
Astropectinidee ...... 4,17, 19, 22,37, 38, 105,175 | californicus, Astropecten.......-- 8, 11, 56, 61, 71
ISIE INTUTE Soe 2 a a wiz a o/c. sialic tate ne 38 | caltfornicus, Chetaster........---......---s 6, 276
attenuata, Cribrella leviuscula........- 8, 280, 281 | californicus, Rathbunaster................. 9
PROENIUNCH, ISTEWLOTUL === <2 clo e se oeeS- Oo) Callisstersscsssc<. <>. eeee eee 162, 171
aurantiacus, Astropecten .........--------- 50: | Callidermia--2 <2 acc. cc co poses ae ee OD Ly Lae
DIET (One e 323) || 'Galliderma grayir<.2<-- << 2. deere teen eueee 153
Matos GUNOGISCUS.. <5 <5 -2-5----% SIS33537) || Cal yptrasters. concen cten c= eee een Uire 344
UU RORY Cr 323 ||) (Camehatica, Asterias-..---.-<--2.s.cs=.eaen 5
MRIS MDM oe eae co sina ac cicleicisicce 1625166; |\‘capelle, 'Ogmaster. <<... ....-.- 2205s --ce aes Oe
DITLUPANTINISER noo =< <n se sec ioe sees 196) ||/(capensis;, Retaster:-. s.n--.< <2 -ees aw arena 362
paardin Mediaster.........--2-.s--.-.0.62 197.0 99)| (capitate, Astersgss..-.-1.--<1--<s-a4eese eae 6
PRAUUMPIARLED =: afarianns socio 22 52. . $04007207/7582 | carnbbraus, Pteraster <2... cee. sseewes 368
PERE TENIR ATE II oo Sco cit Sais 2 LaleiSiccte 71 | carinata, Dytaster exilis.................... 104
Bathybiaster pectinatus........-..+--.------- BAD) | (COMTI) CMe eA BLN Bae Gas woos cose ae
Bathybiaster robustus..............---- .-- 89,77 | carnosus, Hymenaster..................... 375
Bathybiaster vexillifer.................... Udo | CULAR CET Sw marae nser ors aie fens ae mole a ce eter 23
MCMFOMMITINIGINS aoe <2 < cacao o.0ssceee0sse TG) COATAMORLEL So ce seas ae ae eee ee 160,
RIE ae a. Sek cae cock es 344 162, 164, 167, 170, 172, 197, 199, 204, 205
ISEMUHOPECtEN W252. 2022 =s2->-~ 8, 121, 140, 142 | Ceramaster arcticus..............-......... 138,
Benthopecten acanthonotus ...........-.----- 3) 15, 205, 206, 219
Ue 14S)1445 148° 149) | ‘Ceramanter clarkic. .. Ucn c us cdsvuewesucves 10,

Benthopecten antarcticus...............--- 149 13, 15, 206, 217
Benthopecten claviger.... 10, 12, 15, 144, 146,150 | Ceramaster granularis..................... 205,
Benthopecten cognatus.............------- 143 209, 210, 218, 219
Benthopecten huddlestoni...............-. 147 | Ceramaster heesitans. ..........-----.- 205
Benthopecten incertus..................-.-- 143 | Ceramaster japonicus. 11, 12, 15, 205, 206, 210, 211,
Benthopecten mutabilis........... ae ee 10, | 212, 213, 214, 215, 217, 218, 219
12, 15, 144, 146, 147, 149,151 | Ceramaster leptoceramus. 12, 14, 171, 205, 210, 219

Benthopecten pectinifer................--- 147 | Ceramaster mammillatus................. 205

410 INDEX.
Page. Page.
Ceramaster patagonicus.....-...--.--------- 11, | Cribrella leviuscula crassa.........-----. 8, 280, 281
205, 206, 214, 217, 218, 219 | Cribrella leviuscula..........-..----.------ 6
Ceramaster placenta..........------------+- 205" | (Gribrella oculata.]- -2-..2.. eee 271, 276
Ceramaster pulvinus....-.---.-.----------< 205 | Cribrella oculata var. cylindrella............ 271
ceramoideus, Plinthaster ..........-------- 165 | Cribrella oculata var. murina.........-...--- 271
@heetastens 20 te -eee eee eee 4, 18, 21, 37, 240, 241 | Cribrella sanguinolenta........-.-.------- 271, 276
Cheetaster californicus......-.--------------- 6,280 | Cribrella spiculifera...........-..--.---+--- 8, 295
Cheetastervestitus.....----2-seeneee ee eee 21 || eribrosus,, Retaster: .2..-.52 esse ee eee 362
Cheetasteridee:.<. 2... cca seek See eee Cee 4,04 || ‘emspata, Astenias:...-.--->---e eae eee 31
GCheiraster...----cc--+-----+=- 80) 120; 121, 122/123) crispatus, Otenodiscus.- +... esses seeee 11, 14, 81
Cheiraster agassizi evoplus.....------------ 10, 123 | crispatus, Leptogonaster.........------..... 169
Gheirasterhorridustas.-e--22-se==- eee 1'32!)|| (Crossaster: = 22-222 J2-22 2 eee °.... 306, 329, 334
Cheirasterinops’<as------ cee eee eee eee 127 || (Crossaster alternatus-25.-=2525e2-2 eee 9, 330, 331
G@heiraster/eny dere. 22. -----) eee oe 127) || (Crossaster‘australis. 4. =.= 55-2 323
Chewrastetid# =: <2 5-2-5. oo eee eee 120")|(Grossaster borealis... 2-25-22 ee eee 9
Ghitonaster®=="ee-sece eee eee A N6I5163;L73) ||) (Crossasten neptunt---..- =. ee 323
Chitonastenntes. 2225-42 - een eee 5),1160;163 || (Crossasten papposus:..-.----.- -seeeen eee 325
Ghunaster! .- 22-20. ce eee ase eee eee 23 || (Crossaster squamatus. ....5--.--s-eseee eee 329
Ciliaris® Wuidia=2. 25-2 - see eee eee eee 105 | Crossaster vancouverensis........-.--------- 7, 311
Girceastet 22. ae eee ke see eee 162,173 | Cryaster.i5..-..<-0-.2-se-2 eee eee 252
Cladaster 25 --22----e ose 159, 160, 162, 172, 173, 221 | Cryasteride....=....-- -seccee eee ee 5, 252
@ladaster!rudis:--=.2<.--c2=ses-e eee 221.223 | Cryptaster ss. 4222... ee eee 344
Cladaster validus.2- =.=. -sose-e ee 10, 13, 15, 222 | Cryptopeltaster..........-. 8,12, 160, 163, 174, 287
clarki, Ceramaster .....-..----- 10, 13, 15, 206,217 | Cryptopeltaster lepidonotus.. 8, 11, 13, 14, 237, 238
clarki, Henricia! =. s-c1<--i25'- 10; 13) 14,268,303 | \Cryptozonia_--.-2.-2-22------- = eee 4
clathrata® Ibuidiaes..5--s-eaeee-e ee eee eee W2"| (Ctenastends pea sacssle 2 = Sac ee eee 8, 306, 340
claviger, Benthopecten.....-.- 1OMS, 1551445150 \\(Ctenodiscinee - ----.- == = - 2-2 eee eee 4,18, 31
Clypeaster rojundus...s5e-ees sss == 20259! 'Ctenodiscus: s...5.-\ee eee eee 20, 23, 31, 37
coccinatus, Hymenaster.........-.--.------ 383' | Ctenodiscus australis... 22. = Weer 31, 33, 37
coccinea, Patimias.*--222-- = see eee 254 | Ctenodiscus crispatus .............-.-- 11, 14, 31
cognatus, Benthopecten.........----------- 148) || Ctenodiscus krauset. - -- 5522552 7,31, 33
copnatus, Mimaster ss. =e sea ses 53,175 | Ctenodiscus procurator..........-.------ 31, 33, 37
columbie, Linckia.........--.------- 11,14, 242 | Ctenophoraster. . .---------------- 20, 39, 42, 105
columbiana, -Asterias... 27-2 -=\.- 22 eee oe 10: \Guletta.cs-2- 4. . = 2s. 5: = 2c cree oe eee 168
comptus, Plinthaster...-2-3----e-=---e - 165, 205 | cuspidatus, Goniaster....-..-------------- 167
COnferta:, Asteras mere sane ses eie =r 6| Cycethra= .:.--- o2.26 525.02
constellatus, Solaster......--.---- Sue 9,311 | cylindrella (var.), Cribrella oculata....-..-- 271
coscinopeplus, Pteraster........-..--------- 10! | danse, Pteraster.<..- 5-2 5- = =. -- eee 370
13, 14, 345, 354, 355,370 | dawsoni, Acantharchaster.......-..------- 128
Graspidaster?.. 22242 s-e= ese ees 37 || dawsont, Archaster:----2-2-.2 202s seeeeee 7, 8; 128
crassa, Cribrella leviuscula......------- 8, 280, 281 | dawsoni, Luidiaster ........- 12, 15, 128, 141, 142
Crassus, Hremicaster---.35---os2--- see 28 | dawsoni, Solaster.. 7, 13, 15,307, 310, 313, 317, 318
crassus, Odontaster.......-.----- 8, 12, 14, 154, 158 | decemradiata, Solaster endeca.....-------- 311
Grenaster: sccmse haces ees eee 8,41, 55,101 | decemradiata (var.), Solaster endeca..--..--- 5, 311
Gribellas See eyns Hose one ey See eee 266 | decemradiatus, Solaster........-.---------+-- 6, 311
Gribella eschrichtit'= 22-=22-222 oee eee eee 276 | demonstrans, Dytaster........----------- 104
Cribella oculata-s3- e252 eee 271 ||Dendrorasters...- ee s- eee eae eee 237, 264
Gribellosaz see seett- seer snd Ss we 22 | Dendrogaster arbusculus. ........--------- 404
cnibramiamA stenas-=--0 sees 2 cece eee 6 | dentatus, Pentagonaster .........----------- 165
Gribraster: .2-2..220.~ coset eo eee ee 259 | dentatus, Plinthaster.......-.....--..--- 165, 205
Cribrasterialadent. <= 22.25 Yee ee 259 | Dermaster imbricatus....-.--------+-++-++++- 249
Gribrellah- 2 Ae ee ees 242, 266, 306 | Dermasterias............------+-----=- 4,12, 248
Cribrella leviuscula......2.ss0-<ss222--+ 506 280 | Dermasterias imbricata. .......-.------ 13, 15, 249
Cribrella leviuscula attenuata......--.-- 8, 280, 281 | Dermasterias inermis....---------------- 248, 249

INDEX. 411
i : Page. | Page.
Diadema mexicanum.................... 259 | Echinometra OBloviga! << sactog de eee 259
DMA LCUA Matern = fo Foc na Ose coe aides 260).|| dleganayTuidia 0c... 2... ee oeeas See 119
diatatus, Diplodontias ........-.......... 153 | elegans, Mediaster....................2: 197, 204
dilatatus, Pentagonaster .............-..... 153 | elongatus, Gnathaster..................--.. 220
Duplasterias.................-.--.- cease Si | ePendectiee se. et oc ee,_ cae kA 306
diplaz, Tanckia-....-................... 242,244 | endeca, Asterias...........2220.--.-----. 306, 307
diplax, Linckia pacifica var................. 242 | endeca, Solaster........... 306, 307, 311, 313, 315
MAMPREKIOMUNAS 22 eile ao soos celebs ese 8, 153, 154 | endeca, Stellonia... ........-..'..... cee 307
Diplodontias dilatatus..................... 163 °|| eptchlora, Asterias.......2-...2..20-0ssee- 5,6
iplopteraster..-..-.-.... 344, 345, 353, 362,370 | epivanthus, Antheniaster.................. 169
Diplopteraster multipes............. dU sary eed | Mremicastersscc,~octuvdo eae do. eee 8, 9, 28,24
Diplopteraster peregrinator ............. 370, 871 | Eremicaster crassus.............--.--++-- 28
Diplopteraster verrucosus............... 370, 871 | Eremicaster gracilis. ................-.0-- 28
ypsacaster.....-...-............ 88, 39, 42, 50, 85 Eremicaster pacificus...............-sceeee 29
Dipsacaster anoplus..........- 10, 12, 14, 86, 90,97 | Eremicaster tenebrarius.......... 8, 11, 24, 29, 35
Dipsacaster borealis.... 10, 12, 15,86, 91, 95,96, 97 | Eremicaster waltharii....................-. 25
Dipsacaster eximius.-..- 8, 12, 14, 85, 86, 91, 94,97 | erinaceus, Astropecten..............+++++--- 56
Dipsacaster leetmophilus. .... 10,12, 15,86, 95,97 | eschrichtii, Cribella.................-2------ 76
TR PRACANLERMEKIOUES |. ~ --2. <2 Sao. Sees 95,96 | eschrichtit, Echinaster.............-.+---- 276, 279
ipreteanter Rader 2 = 2.2. e ek. ke 85 | eschrichtii, Henricia sanguinolenta.......... ll,
RnR eS cso cies cur trek wale Sie 253 259, 267, 271, 276, 282, 283, 289
discus, Pseudarchaster..............-.. L79NLSS|| Hugoniaster. 20. . cs .-seeesaeeee 162, 172, 205, 219
dissonus, Pseudarchaster . -..-- 10, 12, 15, 180, 192 | europxa, Hippasteria..................+.-- 223
PENTEL ECVESU SE te es Liaise kes e ccs 169 | evermanni, Zoroaster (Myxoderma)......... 8
anelaniwActenagiccee. (Sec aS.2 2.4. eee | EXVODIOBOINA Je 5ic\n..werao ac saa ace eee 163, 174
SPI APPRSELIAS SS eee 2s occa cess. lee cee 10 | evoplus, Chetraster agassizi..........-..--- 10, 128
duplicatus, Astropecten . ................. 60 | evoplus, Pectinaster agassizi.............. 123
dyscrita, Henricia leviuscula.............. 13, | exiguus, Solaster ............ 10,13, 14, 307,819
14, 267, 270, 282, 289 | exilis, Brisinga..........2.....20.22.2200- 8
ree EES... 8, 38, 39, 41, 42, 101 | exilis, Dytaster...............-ce0cc-eeee 104
Dytaster demonstrans.............-------- 104 | eximius, Dipsacaster..... 8, 12, 14, 85, 86, 91, 94, 97
MEIN ERBEKTLIR ema ee oom. Ao LSS es 104 | fallax, Pseudarchaster. ............-....-- 184
Dytaster exilis carinata................... 104 | fascialis, Phataria (Linckia)....-...-...-.-- 242
etantenealbertiacs-----+-.......-.. 8,12, 14, 101 || fecunda, Freyella:..............2.cseetece 8
Lian on 1O4;|| Rerdinnwen- seek. eo ee cw eee 159, 241, 258
Dytaster madreporifer. ..........-.+-.--++- 104 | filholi, Pectinaster...........-...--- 123, 126, 127
PIMIBEUGRPNODUISe 252252625 So's i eee ees. LOM | yessrewind, A sterias. - 5. -.-- 2 oe eee eee 6
Wiytaaterispimnosuss-.:......:......--..--- 104 | flexilis, Poraniopsis inflata......... 10, 11, 261, 265
PRESET E ASA oso Coosa DO 258, 260, 261 | florse, Psilaster..........-.----- saahd nome 7
EIS ere ee ee Ie te 266 | florifer; Mediaster-....;..-:J0scepeeew ee eee 197
Bichinaster eschrichtvi..................+-- DIGIQ79 | Worden Ada... .- 2+ 22 akon oe eee 6, 106
Echinaster eschrichtit var. macrodiscus....-. - 276 | foliolata, Luidia............ 6, 11, 12, 14, 106, 113
Echinaster eschrichtii var. microdiscus....-.-- 276 | folliculatus, Peribolaster...........-..... 341, 343
SPRCMUTLISLEN OCULALUS 2. awa wee ee ee ee 271 | forcipatus, Pectinaster.........---.+-+----- 126
Echinaster sanguinolentus.........-------- 271 | forcipifer, Lonchotaster .............--.--- 39
Echinaster sarsii.......22..02000.000ee eee 27])\| orcipulata . <2... .22.-42 sn eee es em 17,251
Wichinaster sentus. ........2..200ceecceeees 261 | forcipulata, Asterias. .. ..........- 252.055. 10
Echinaster serpentarius.......-.----.----+- 260 | formosus, Paragonaster. ..........--.-. 170
Uichinaster spinosus.......--.--22522----. 261 | forrert, Astertas......-......sees0ssececeens 7
PCAN Ler SPNULLOSUS- <=... a's Ho eee woe 261 | fragilis, Astropecten.............--..-..... 67,71
Echinaster tenwispinus ...........-.-+--+-- 8, 259 | fragilissima, Luidia..........-.-.+--+------ 106
echinaster, Poraniopsis. .......--..-- 260, 261, 265 | Freyella fecunda. .............--.-----+-- 8
Echinasteride ...... y150;160, 240, 201, S525268 | Wromin.2Ls. sl =.... .< o. ocadeeedenwant 159, 241
echinasteroides, Poraniopsis........-.------- 260 | furcifer, Lophaster...... 334, 335, 336, 338, 339, 340

412 INDEX.
Page. Page.
furcifer, Solaster. .....-------++-+--++--+-=- 344 | gnayi, Gnathaster. «2. <2. -/-/-e eee . 154
furcilliger, Lophaster......-...---- 8, 12, 384, 338 | grayi, Odontaster.................... 153, 154, 158
galacides,,.S Olsen a. a. mene oe 9, 308 | guildingii, Linckia.............. 242, 243, 244, 245
Ganerlae.ss-)-gasn ses asee eee ese 251, 253 | Gymnasteria.......-.--- seaports 8, 159, 247, 248
Ganertidgeia ns snscic os cee tee er ne eee 52M || Gymnasteriadse - 22. cece eee eee 247
Ganeniins ao oe cae eeoee ne eee 258; || Guymnvastertidee=(.)5 sel .< -(staie -1=)s) eee 4, 159, 247
Gaussasterss.2 2085 2enc-- 5 -smacecemeereeer 120: ||; Hacelia. 12... .22252-.55.. os: 2 0eeee ee eee 241
Gephyreaster.......-- 10, 12, 21, 160, 161, 170, 175 | heesitans, Ceramaster ..........-.....-.-.- 205
Gephyreaster swifti.......-..-..- 12,15; 156)075: | hardunckit, Petalaster. .—-. - sate 105
gerlachei, Luidiaster...........----------- 127 | heathi, Hippasteria.... 8, 13, 15, 224, 229, 231, 287
eibbermRetasters: saeete= eo eerie eee 862) || (hebes,Peltaster <--..0+---- eee eee 9, 162, 345
gubbosa; PASteMmase ens amie one 254 | helianthemoides, Asterias......-...---------- 325
grgaritea A Sterns 5 on ioalo laren esse eletatee=taie 6 | helianthoides, Asterias.......-...----------- 5
Gilbertasters--se-eeeoeeeeee 160, 162, 163, 172,173 | helianthoides, Pycnopodia..........-...-- 6
silbertiyDytaster= === 2-4-s- eee eee SiilOd || enicnemist see se- = 1-2-1) eee eee 105
glaucus, Hymenaster.........2..0.0--s--0 37.9) MELenTICia =f en ere ees aase sess 37, 84, 258, 259, 266
Glyphastensci-7.s 22-2 22sec 10;12539%495 53255, || Henricia aspera. J-. 2 -o= + --c aeeeeeee 9, 13, 15,
Glyphaster anomalus....---.----+-++++--+-- 48 267, 270, 272, 287, 289, 291, 292, 293,297
Grinthasten ae 32 2a262 2 see eee ee 153, 154, 158, 159 | Henricia asthenactis....... 10, 18, 15, 267, 270, 297
Gnathaster elongatus ........-.----+------- 220) Henriciarclarkis. sone. e-em 10, 13, 14, 268, 303
Giiathastenigrayisee sn eee eon eee eee 154.) tHlenriciagracilis---:-o. ..-- se == 270
Gnathaster meridionalis.......-------------- 154. || Henricia ryadesi-------52..-9----=ee see 293
Gnathaster penicillatus......-..----------- 154 | Henricia leviuscula.......--..- 6, 13, 15, 267, 271,
Gnathastentne =, 522 ona ee eeeeeee 153 272, 280, 286, 289, 290, 291, 292, 293, 295, 296
Goniaster*s. 22455206555 35-- atone eae 21, | Henricia leviuscula annectens......-....--- 10,
158, 159, 161, 162, 166, 167, 168, 170, 219 13, 14, 267, 268, 291, 295
Gontaster = 25 ose ssh eee eee Ce eee 159, 163 | Henricia leviuscula dyscrita........-....--- 13,
Goniaster cuspidatusas csc saa eee ee 167 14, 267, 270, 282, 287, 289
Goniaster nidarostensis). = 2-22-12 = tele tine 162 | Henricia leviuscula multispina......-..... 10, 13,
Goniaster obtusangulus.....-...----+-+-+---- 159 15, 267, 271, 272, 286, 289, 290, 292, 296, 297
Goniaster tessellatus:: -<....-.2<m--seeeie <1 167 | Henricia longispina......-. 10, 13, 15, 268, 299, 300
Goniasteridse. 2.22 ...asc = sleeeein eee 5, | Henricia longispina aleutica.....-. 10, 15, 268, 300
18, 153, 158, 160, 161, 162, 247,251 | Henricia minuta................---------- 270
Goniasterine...< ...----=.2--- 5, 159.160, 162; 196. | Ienricianana.---5- =... ee eee 270
Goniodiscastersscec se sess ee 162) 168,171,172 || Henricia oculata..—..------ -22-> = eee 266, 271
Goniodistides: oon. =< 22 = sass 168 | Henricia pauperrima......---.2---2-se-= 303, 305
Gontodiscides sede -7-2- a2) 2 = 2 Seen 168 | Henricia polyacantha..............----- <u o
Goniodiscides studeri.......----<0-----s0+- 168 10, 14, 268, 302, 304, 305
Gonrodiscus: soo sscccstOs2 eo. ee eee eeeee 168 | Henricia sanguinolenta..........----------- u,
Gonroduscusppleyadellaeeecice eee 168 259, 267, 271, 279, 283, 296, 300, 347, 373
Goniodiscus stn gUlans lee) -lo12 1.2 eel 153 | Henricia sanguinolenta eschrichtii.....-.-- 259,
Goniedonsh Gt ook See eke eeeeaes eee 8, 153, 154 267, 271, 276, 282, 283, 289
Goniopectinid = <.-<.)--e-e-- eee 4, 17,19, 22,37 | Henricia sanguinolenta var. pectinata....---- 10
gracilis, sBremicasters--- = a-eeeereeee eas 28 | Henricia spatulifera--:- 2-2. ----2-e-e eee 10, 271
gracilis, Henricia. eeeiecisc cesar ace 270 | Henricia spiculifera............--- 13, 15, 268, 295
gracilis, Hymenaster........-.....-.-2---- 380 | Henricia tumida..........-----+------9--"= 10, 276
gracilis, Pteraster........ ---- 13,15, 345, 849,369 | Heterozonias.......-.---.-------- 10, 12, 306, 330
Qracilis, VRetasten 22.0052 42 ese 8,349 | Heterozonias alternatus. . ...- 13, 14, 320, 321, 331
erandis: Dytasters.222-ssos2sseseee eee 104 | hexactis, Asterias..--... 222. --52-=see=ee== 6
grandis, Sid eriaster:... 22. = e-eeeeee eee 4] | heractis, Pteraster. =... ..- «2-2 eee eee 363
granulare, Astrogonium......---------------- 214 | hexactis, Pteraster (Temmaster)......------ 344, 363
granularis, Asteria3..........----2-=+<+= 168, 204 | hewactis, Temnaster.......--...----+++----= 363
granularis, Ceramaster........... 205,209,210, 219 | Hemaster.......-.---------+--2e++---+ 344, 367, 368
GrayisCalliderman eRe once cee 153 | Hexaster obscurus.........-.------------ 344, 363

INDEX,

Page. |

21, 159,
160, 161, 163, 173, 174, 221, 222, 202, 237
Hippasteria californica.................... 8,
13, 14, 163, 224, 225, 226, 288, 234, 264

PIG DUSIATUE EUTOD EO... = 22-2 wee een nnn csecce
Hippasteria heathi..... 8, 13, 15, 224, 229, 281, 237
Hippasteria leiopelta......... 10,18, 224, 227, 230

223 |

413

| Page.
INPOLAd Pharaster: 326 Acs te tet. cl 370
inops, Cheiraster.................. 127
insignis, Amphiaster........... 168
insignis, Retaster.......... ; 362
intermedia, Molpadia.......... iene earee 319
intermedius, Acantharchaster......... . 10,141
intermedius, Myonotus.............. 12, 14, 141
intermedius, Pentagonaster.......... 165
intermedius, Pseudarchaster..............- 180, 184
investigatoris, Pentagonaster................ 162

|
|

Hippasteria leiopelta armata............. 13, 230
Panpastera pacificd........-.-......-:-.- 287, 240
Hippasteria phrygiana....................- 223,
224, 229, 230, 234, 236, 237
IP PASterIA SPINOSA..--...-----~-6- ca ete ws os 9,
13, 15, 224, 226, 229, 230, 234, 236, 237 |
Hippasteria spinosa kurilensis. .......... 13, 226
Hippasteriine........-.. 5, 159, 160, 163, 173, 223
ETSEIUS SUUIGIASLEL: <. .. . - . 2 an eice ss nda 127 |
hispidus, Odontaster...............--.-.-- 154
ries (CGR i ee 132
emoreis wiauidiasters. <= 2 22-5 - = + 127, 132
huddlestoni, Benthopecten...............-- 147
irese SLCC A] 22-225 5.5 1. n= 22s neice 293 |
ITEMS LET ee ase hw arn aci2/o' v= oc O44) 878
PevIenaster CAINOSUS. ...-----. 2-22 sense 375
Hymenaster coccinatus................-.-. 383
eivamenasterelaucus.......---.--.------0-- 379
Enymenaster Pracilis:---<...-..----.-~-----. 380
Hymenaster infernalis...........-...-.-.-. 379
Hymenaster keehleri........-- 10, 13, 15, 253, 373
Hymenaster membranaceus......-.....--- 379 |
Hymenaster nobilis.....-..- 373, 374, 375, 376, 380
Hymenaster pellucidus.............-.--- 373, 379
Hymenaster pentagonalis...........--...-.- 375
Hymenaster perissonoties ...... 10, 13, 15, 373, 376
Hymenaster platyacanthus...........------ 384
Hymenaster porosissimus......-.-.--.----- 383 |
Hymenaster preecoquis....-..----- 383
Hymenaster purpureus.........-.-.-.----- 384
Hymenaster quadrispinosus. ......-. 8,
12, 253, 373, 378, 379, 380
Hymenaster vicarius..........---..--.----- 379 }
eee adacn 375
Re Eee eta). ici x/ aw wn/e\0\ a\s «2:5. s\= 23

hypothrissus, Solaster..........

10, 13, 15, 307, 824
162, 172

ambricata, Asteropsis..........--------- 6, 248, 249
imbricata, Dermasterias.........-.---- 13, 15, 249
imbricatus, Dermaster........-.-----+-+++-- 249
incertus, Benthopecten............--------- 143
tmermis, Dermasterias.....-.------------ 248, 249
tmermis, Parastropecten.............-------- 42,53
infernalis, Hymenaster..........---- ----- 379
BInAta we OrANiOpsIS.....--------------+--> 8, 261
inflatus, Alexandraster......----.----------- 8, 261

40 |

TIsaster .. Seebubaceuaanwereeee 196
| janthina, ace saat ASerela Ske are ners ete 5
japonicus, Ceramaster ll,

12, 15, 205, 206, 210, 211, 212, 213, 214, 219

| japonicus, Mediaster. 197, 206
japonicus, Pe niapOniate i 206
jsponicus; Solasten.- 5. ..c oc. aeaceees 307, 330
VONANNAStOD oe ok eee soe ge eee ee 162, 173
jordani, Pteraster. 8, 13, 14,345,350, 353, 354, 369
katherine, Asterias ....... = 5 5,7
kerguelenensis, Leptychaster............... 42.53
RD LOT A SLETUIS ee an ae ee ee 7
keehleri, ahaa Dolan saree 10, 13, 15, 253, 873
Korethraster. . . 340, 341
earetnrantenidis. Orr terre at 5, 252, 340
ROT ELAS OSIETANEE aca, 5 arms as in 340
ERQUSEY. | OLATIOCNSCUS: = soca a en nai om eee 7, 31, 33
kurilensis, Hippasteria spinosa........... 13, 226
Wee GMSSLOV cc csce tes Osawa ae tee a karen 8, 306, 340
lxetmophilus, Dipsacaster... -- 16, 12, 15, 86, 95, 97
leevigata, Linckia..........-..-.-- 242
ieviuscula, Cribrella........-..:--0+---<=0> 280
TOUS LLCO Se aston oe warty Satake cea, a ea 260, 261
GOLA TOT Oa a an n'a 3 nee 260, 265
TCS EA wry tase, Die mnie cist are ees
TU AS CBR > emt es 3s ee ck gs ia he 248
ee runastensecis = 4-5. sw esses fake hen aero 251
lebruni, Pteraster.........-....-<----+--s- 369
TelARCGn re aeons 6 ae a ow a et 241
leiopelta, Hippasteria....-... » 13, 224, 227, 230
lepidonotus, Cryptopeltaster. 8, 11, 13, 14, 237, 288
Leptasterias..........-.-----.s-0ceeeeneeee 368
Leptasterias arctica. 7
leptocerama, Tec. : See ate 8, 210
leptoceramus, Ceramaster. “1, 14, 171, 205, 210, 219

159, 160, 161, 163, 169, 173
169

Leptogonaster

Leptogonaster crispatus.....-.-.--.--.-----
Bn oe eton nace aa aces ae en A
Leptoptychaster. Pane 42
Leptoptychaster arclicus. : 43
Leptychaster........- 20, 38, 39, 40, 42 53, 97,175
9,

Leptychaster anomalus. ...---
10, 11, 12, 15, 42,43, 48, 53, 54, 100
Leptychaster arcttcus..... 11, 42, 48, 45, 52, 53, 347

414 INDEX.
Page, Page.
Leptychaster kerguelenensis .......-..---- 42553 || iaidia sarsls: =: -2225-0--0- ene 116, 119
Leptychaster millespina........-.---------- LO! Mbuidiasterseeste. posse eee eee 120, 123, 127, 141
Leptychaster pacificus.....-.--.----2----. 9, | Luidiaster dawsoni........-. 12, 15, 128, 141, 142
12, 15, 43, 45, 52, 53, 54, 100 | Luidiaster gerlachei. ....................- 127
Leptychaster propinquus. -.--. 10, 12, 14, 43, 58, 54 | Luidiaster hirsutus........................ 127
leviuscula; Oribrellas <2ccse-0 eo meee 6) || Duidiasterhorridtg!-252252---—s— eee 127, 132
leviuscula, Plenricias=2-- see - meee nee 6; || Lundiaster oxyacanthus:-----)----seeeee 127, 132
18,15, 267, 271, 272, 280,/293'| Luidiaster teress_ 25---- == eee 127
leviuscula, Linckia......-- Leto cane 6;.276 || Luidiaster! vincenti- =. --255-4e0 eee eee 128
limbata, Dnnidisz2ss2.s 2-5 -n 2 see eee ee 1AZ))|| Wouidiidees sass eee aos eee 4, 17,19, 22, 37,105
Minckiawss20 serch ace soe eC eee ee PAQ 241 ADT Auaiinseteescceee faee eee eee Lele "87, 105
Tanck oo so eae seise os eawes sae ae eee 266:\| Uiltkeni, “A'stertas!s2 -- = 2-2 eee ee 6
Ibinckia\columbisesa- =e eee ee eee WN, TAS 2425 245) | lattkent. A stropectens=--2= == = eee eee 43
Tinckia Giplacs. 2.22 --8-e=- hese 242,244 | Toydiaster! 225-2-5:2.52--22 = 20s oo 162, 173
Linckia guildingii. -......2:5.2-+ 2425243)24404) || Wiveasterias=: |. 2+ == 2-2-2442 9
Tinekia \levigata=-- 2222-222 setae see 242 ||| macropora; Astertas.. 2-222. 22- 222 esse 10
Tanckialeviuscula..2-- 2222 sense oe 6, 280 | madreporifer, Dytaster.........-.---------- 104
binckia marmoratai)-.o22-225 ee see sees 242'| Magdalenaster= 2... 2. .-3. 5-22 see eee 252
Tinckiaoculate@s 2 os oc=on=cee ee sees eeee 271 | magnijficus, Archaster. <= 2 22252-42200 eeeee 41
TAnckia ornuthopus=.s-- 22. =+-52-2642s0 242 | mammillatus, Ceramaster..............---- 205
TANCK Pacyficdins re. « = sea eene see eos 244')| Marcelasters)./22. 22252 5222 Cee 120, 122
Linckia pacifica var. diplar........-.------- 242 | margaritifer, Asteracanthion.......-..----.- 5
Drnekia pertusaias secs eeeeeeee 271) | Marginaster 22022. </s2 i=) 4523. eee eee 247, 248
TNC DUBS o-oo oon eee Oe eee Eno 249) |Smarmorata, Dinekia=- 22s: = --- eee eee eee 242
EANCRUU Se Ss Seize oe aaa)s eee 8 ae Chee EEE 240. Marsipaster. | J2.22.2.22-0 322s ee eee 344
Linckiidee. 4,5, 18,20, 22, 158, 159,161,240,251,258 | marsippus, Pteraster.. 10, 13, 15, 345, 352, 356, 369
ihitonotaster: 2: eas ceccesssceeser re 162,165,172 Mediaster: 2) 25522 2-.2: 525. -eee == eee 159,
Sionchotastersseaeee a= secee eee ee 38, 39, 41, 42 160, 162, 164, 170, 176, 196, 205, 207,, 219, 313
Lonchotaster forcipifer: <= 422.2: ss--seessne 39 | Mediaster eequalis.. 6, 12, 15, 196, 197, 198 204, 205
Lonchotaster tartareus......-...--.---..-- 42) | Mediaster agassizil- 2. -. 222222525 se eee 197
longispina, Henricia......- 10, 13, 15, 268, 299, 300 | Mediaster arcuatus.............-...- 197, 199, 205
longobrachialis, Archaster parelit........-.-- 180) || Mediaster bairdiis 252-2 .< oe eee 197, 199
Lophaster. :o- 2.2 a:ssc sae seene see teeeee 306, 384 | Mediaster elegans... -..-..---2---2-22=e 197, 204
Lophaster furcifer......- 334, 335, 336, 338, 339, 340 | Mediaster florifer_..-....-.-..) 222. 22eeeee 197
Lophaster furcilliger.............- 8, 12; 884, 338 | Mediaster japonicus....-.. =... . 2-2-2 197, 206
Lophaster furcilliger vexator...._. 10, 11, 334, 388 | Mediaster ornatus...............-------- 197, 199
Gopho plenastencteri rc es secon ee eee 344 | Mediaster patagonicus...........--------- 197,214
lorioli Suidias teas. s 2 eee oeree eee 116 | Mediaster pedicellaris. ..........-.------- 197
hudwigi*lnidiae---sse--ee sees 9,12, 14,106, 118 | Mediaster stellatus.........-.------------- 197
Tiidias oe seer ay ec a eee 20,105 | Mediaster tenellus. ..--.------ 8, 11, 197, 202, 204
Luidia asthenosoma............ 9, 12, 14, 106, 116 | Mediaster transfuga............--------- 197, 204
Pidiaibellonse--\- eases eee eee 116 | Mediasterine............-------- 159, 160, 162, 196
Luidia brevispina............--.---- 107, 109, 112 | membranaceus, Hymenaster....-.-.------- 379
Lwuidia californica......--...-.--- Asari baie 7 | meridionale, Astrogonium........---------- 153°
Tuidia, cilisnis' = 4925 ese ee eee ee eee ee 105 | meridionale, Odontaster..--....----------- 153
Tnidiaiclathratase= =--2essee sees Sah see 112 | meridionalis, Gnathaster....-.------------- 154
Tuidiarélegans*\o0.ce sass eee eee 19) || Metrodirays-2-.o-220 255 eee eee 159, 241
DUO OURO see eet eee 6; 106 || mexicanum, Diadema=: 2 22--222-5-eeeeeere 259
Luidia foliolata............ 6, 11, 12, 14, 106, 113 | micropelta, Tosta......-.---------+---s=-= 205
TAG OQUUSSINUO ase lactacin eee 105 | micropeltus, Peltaster....-.---.----------- 205
Luidia limbata....-..-. ris Sse SE eee 112 | miliaris, Astrogonium.-.....---------------- 153
Eidiatlorioli= secs. ase eee ee ee 116 || miliaris, Odontaster-----=--2------==-==ee= 153
Duidia ludwigi.2- 2s... -+---25--- 9, 12, 14, 106, 118 | militaris, Asterias. ......----------=-=--— 344, 346
Tbuidis; quinaria see sees see ee ee 112,116 | militaris, Asteriscus.......-...-------------- 346

INDEX. 415
Page. |

militaris, Pteraster ..........-000-000000--- i, || Nepanbiinde eae, ee 158, 253
; ‘ 345, 846, 349, 350, 353, 354, 356, 368, 373 | neptuni, Crossaster........................ 323
millespina, Leptychaster.................. 10 | Nereidaster......... 159, 160, 162, 170, 197, 205, 219
Milteliphaster ote ee eee ee ee eee ee eee eeeee 162,171 | Nereidaster bowersi........................ 199
Mimaster. . ..--- 4, 21, 38, 53, 159, 160, 161, 169,175 | nesiotes, Dipsacaster ..................... 95, 96
Mumaster:copnatus.........-...--.------- 53,175 | nidarosiensis, Goniaster...............-.-.- 162
MRIS ETESU LURE tere e te tiocie + cwcee secs Oy 27D | Nidorellia-....lcc.cs0 beck baccshacceecoche. SISO
eriaBtennnizardiveess scot escccces. ccc. cece | 175 | nitidus, Plinthaster:.-.<.-....22-..0..00 165, 205
BT LALET Ie eel ete naa = a/o/= @\ia's =\= === = = AGO) \inobilis, Dy taster. eee as coeur. eee eee 101
Mimasterine............... 4,37, 159, 160, 161,175 | nobilis} Hymenaster........ . 373, 374, 375, 376, 380
mimicus, Pectinaster............... 123, 126,127 | Notomyota...................- 17, 20, 22, 120, 121
PTTL AS LCLULSE Se Sania winjan/-/a)22 sine nies = 5, 254 | Nymphaster.... 153, 159, 161, 162, 164, 170, 197, 219
Mminiata, Asteria. .--..--.--2-..-5 11, 13, 14,254 | Nymphaster ternalis.....................- 199
BURNIE PEACE = cin <o ies ene sce ew cece -o oes 6y|"Gbesus, Pterasters----22ses soso see eee 357, 370
TTPATINTIES PA SLETISCUS 3 © \2 oe ono eee = = 254 | oblonga, Echinometra....................- 259
PINIVUMPASLCTULS cee oases osc ees e nesses DOA \"OUSCUTUS: HELOSTET =< - cou. ce eee nee emeeees 344, 363
RPE MEL CNICI Nema tees con ees = cise Je= oe 270 | obscurus, Pteraster............-. 11, 334, 368, 370
SEITE MTRENER LOU Seer See on wines = <= s = 260, 265 | obtusangulus, Asterias..............-.----- 163
PRERTEM CHT INTLON TO so sno Saino an <n «<= ois + 265 | obtusangulus, Goniaster.............-.----- 159
PIBELOLANIOPAIN eros se wee nec oc secs <= 265 | obtusangulus, Pseudoreaster............. 159, 163
PRONOUNS Soe nas-e ss s-6-------2-<-- 206) ocellifera, Nectria---.-- ---<-.. 5 cc seussece 163
Menus PALETONGNASteT. .-- ...--------- 260, 261, 263 | ochotense, Asteracanthium..........-------- 6
irae ern eee esse a= 202, 200; 209!) Ochraced, ASlevias. =~. .22-5-0-ncccesncasens 5,6
PORTE semen ere toe = cicis wee ween === 5,262 |"ochraceus, Pisaster--.. ..---.<<-sces-seaees 5
MNP REIHURT eee seas a hla cc cc eee sestees ee 252 NL OClaster. P LerOSler a ae 2 cere can cee ee eees 9, 363, 368
STH Eee race sone = arcloie ce cle cee in 30%41)|' Geulata, Atenas. <2.<.c-ne~ cst teseneeue ees 271
Molpadia intermedia...........-.---------- SIO oculata Gribrelia. sess. -sctee eee c ete seeee 271 1, 276
mortenseni, Pentagonaster (Philonaster) . ...- OES INOCRIALGs eM ICUs os nen =aeeaes se seen eeeee 271

multipes, Diplopteraster......-...... 11,347,871 | oculata, TANCKIG 5 <2 = ae nwo one ae eee aes 27
SIPEEIUDES RE LETOSIET 02 -.2= = ie vin enn ce cee STloculatus: ChiINGSter 222-- 2 eae ea een eee 271
SR IER PLRCLIREOT Se oi cieine 2 = on == = <'- <= Sia) | Odontastene=-s-sceses =e setae 18, 153, 154, 158
multiporus, Pteraster..........-.-.----- 355, 370 | Odontaster crassus.......------- 8, 12, 14, 154, 158
multispina, Henricia leviuscula............ 10, | Odontaster grayi ..----------------- 153, 154, 158
13, 15, 267, 271, 272, 286, 296, 297 | Odontaster hispidus .........--.----------- 154
multispinus, Pteraster.. 8, 13, 15, 345, 355, 359, 369 | Odontaster menidionalis'<. cc see sane anne 153
murina, Cribrella oculata, var ....---.------- O71" Odontaster MINatIS-c..c-—=2 nse eee ee 153
mutabilis, Benthopecten. - Bea 10, | Odontasteride.......--- 4, 18, 20, 22, 153, 159, 175
2s 15, 144, 146, 147, 149,151 | Ogmaster.........-----.----------- 163, 169, 173
PINES TIS tees ere fore cso enn ores wei 12, 121, 140 | Ogmaster capella..............------«----- 169
Myonotus intermedius...............- 12, 14,141 | Ophidiaster...-......-.---.------------- 241, 242
eR ARLEN tn c= oa wn ain n= = 253 | Ophidiasteridx.........----+-+-----++-+---- 240
DMpean eC eee anon ce ess scene eee = BaShes| | Oreasterc.--- <=. -4-2 sen as weaker 8, 158, 159, 168
BO ueRsas enh Me eee ore =) 2 ose nin'= 2 === 52253) Oresstenlds -<nina. ese 5, 18, 20, 158, 159, 161
PAtertNt enim ere Me 8 oon cio mana ssc vie 8 | ornatus, Mediaster..............--------- 197,199
PiTOMBERE TSCA ye io ee < ayasic ie = = =\=/='slainin'n m= 270 | ornatissimus, Astropecten......-...---.-- 9, 56, 67
Rr eee ten 241 | ornithopus, Linckia......-.-------+++------- 242
3 006. ee “240, 241 | orstedii, Astropecten ...........--+++-+++-++ 56, 57
Nearchaster.........---- 12, 21, 121, 182, 140, 142 | Ortmannia.....-.-.------~--+--++++0+002=* 260
Nearchaster aciculosus.......-... 12, 15, 183, 141 Orimannia MiG... 2... 0. 06scceccnee seen 265
Nearchaster pedicellaris. ...-..--- 12, 15, 133, 188 | Othilia. ....--..--------++++++-se00--5> 260, 261
Nearchaster variabilis. ....--- 12, 15, 132, 133, 187 Othilia tenuispina.......--..-.-.-----+-+- 259
Nectria....................-.-- 161, 168, 164,197 | oxyacanthus, Luidiaster.........-----.. 127, 182
iNeeimocaliiterd:]... =... <<..s2-26-2 sec. 163 | pacifica, Hippasteria.......-----++---->- 237, 240
Nectriinse.....................-------- 4,160,161 | pacifica, Linckia. ........--+++++++-++=- 244

416 INDEX.
Page. Page,
pacificus, Hremicaster-,. 6-242 ----s-5- see 29)\| “pedteijer, -Pararchaster,--------o- seen eee 142, 143
pacificus, Leptychaster.. <a: oo: caer 9, | pellucidus, Hymenaster.................. 373, 379
12515, 43,45; 0233.54.01 00)|uheltastersnc sans seen ae eee 162, 171, 205, 219
MACifiCUs, POT CEUANAS Cina 2st eee ee 8,29) ||) Peltaster hebes: <c.05 .< <1 stones 1 162
Palmipes.. settee wiser - 158, 159, 253 | Peltaster micropeltus .....................- 205
papposa, Veena ee coon eee ee eee 306, 325 | penicillatus, Gnathaster.................... 154
pap posa, Stellonias 25. <2 cee a anes 325 | penicillatus, Persephonaster..............- 8, 78, 79
papposus, CLOSSASLEN << = aia = = = atmninimin-ta\f-y=r= = 325 | penicillatus, Thrissacanthias ........... 12, 14, 79
papposus, Solaster.....- 11, 306, 307, 320, 325, 331 | Pentaceros.. sists ese tea since Se ee 8
papulosa, Asterias (Pisaster)..- - - Simm cigy=afayets 9 Paocniee seoeeES - 158, 159
Paragonaster......-.-.- 159, 160, 161, 163, 164, 169 | pentagonalis, Hemonasten Senden eee 375
Paragonaster formosus..< 0-625 20-0 spars 170) || Rentagonaster:....-22--- <2. -0- ssa e eee 153,
(Pararchasiefassneccee tener ene 8, 121, 142, 143, 152 156, 160, 162, 166, 167, 171, 205, 219
Paranchaster pegicifer<.- -- 2s sao 1425143) || Pentagonaster cess => en aes oe eee 196, 204
POF On CRASEEPUN SE wae ase seiale <<less eel zo) L205" | Rentagonaster afiniss=-.c cc 422 ee eee 165
ParastenDdeccec nice sence oe eee 8;,159:,253) || “Pentagonasterdentatus: -.--2- +--eeee eee 165
Parastropectenan-eeees ee ee eee 39, 42, 53, 55, 100 | Pentagonaster dilatatus..............-.---- 153
ParasinopecteniNermis elas eee eee 42,53 | Pentagonaster intermedius....-..-....--.---- 165
DATE RAR CRASLEN See ae sey ae ee 180 | Pentagonaster investigatoris. ...........-.-- 162
pare. -ASTrOGOnNvUiM cannes toe as ee eee 180 | Pentagonaster japonicus ......---.---------- 206
parelw, ‘Astropecten <<. c-nse sess 180, 184 | Pentogonaster (Philonaster) mortenseni.....- 204
parely, Plutonasten. <== <tes essen 180 | Pentagonaster patagonicus .........--.----- 214
parelii, Pseudarchaster....... 11, 180, 191, 347,373 | Pentagonaster perrieri..............-.------. 165
Patagiaster a. cee ccse = eee ees 20, 38, 39,42 | Pentagonaster pulchellus................... 166
patagonicus, Ceramaster...... 11, 205, 206, 214, 219 | Pentagonaster validus ..............-.----- 168
patagonicus, Mediaster..........--..----- 197, 214 | Pentagonasteride.........--- 158, 158, 159, 161, 163
patagonicus, Pentagonaster...0.2 5-0... ---- 214) Pentagonasterins: 22-2... 2522s 196
PE QUUIUO sores eo acielg se ines tee 8, 159, 253, 254 | peregrinator, Diplopteraster ...........- 370, 871
IP atiriar coccinea ecm sects at nee ee eee eee 254 | Peribolaster.. Lae Soa - 340, 341
POuTMa MIN Oe nasese nee seater eee ae 6 | Peribolaster pee Solace sce ene ae 8, 341
DAUCLS DUNG A SLCnIG8 Sas cia enn eee 6 | Peribolaster folliculatus..........-- 13, 15, 341, 343
Paula sono. Sjas nie eee Ohare eee eee 159 | perissonotus, Hymenaster..-..- 10, 13, 15,373,376
pauperrima,, Henriciae-.-c-24- see = sare 303,305) Perkmaster. . - 5... 6) 0-6 on.0s 44 eee 259
paxillatus, Solaster.... 11, 18, 15, 307, 311, 315,320 | perrieri, Pentagonaster...........-.-------- 165
POR OSO 8 seem asec eee SEE eee L7,,195:22: | perniert, Plinthaster=-.-.. -.--.2---eee ee eee
Pectinaster ecco meetin eee ee 120, 121, 122, 123 | Persephonaster..........--.-- 20, 38, 39, 40, 78, 84
Pectinaster agassizicen-.--e-- eee eee 123,127 | Persephonaster penicillatus ......---------- 8,78, 79
Pectinaster agassizi evoplus.........-- 12,14, 128'| personatus, Pteraster............-----2ese8 369
Pectinasterlholizsceste--eccense eee 123; 126; 127), |\ pentusa, -Asterias=- -2a2e< 24-2 => -ee eee eee 271
IP ectinasterfOncupatusesesee= eee en se oe eee 126) | pentusa, Tanckia...--2--2- eee eee 271
Pectinaster mimicus............-..- 1235 126127. |) Petalasters aces ccc. sec ce eee 105
DECUTALO PASLENIOS Sere eeee = ae ae eet eee 5 || Pétalaster hardwickit.............--2--2s-ee 105
pectinata, Henricia sanguinolenta, var ...... LOW Petriciien-ceecet meet ene tree corte 4, 8, 247, 250
pectinatus, Bathybiaster..................-- Siivi2! || Petmicial punctatds esses es. saeee = ee eee 247
pectinatls, Psilasterse-ceseses eo seeeee 871438; 72) || Betricia vernicinas..-.-- ->2-225-Seeeeeeee 247
IPECtINIGISC UB! sane ee ee eee 23) || Phanerozonia...--.---------.--4- =e) AL Oso
pectinifer, Benthopecten..................- 147) Phariai cscs ctssseee. sec cee eee eee eee 240, 241
pectinifer, Pseudarchaster= 2: -_.- ss. 192-194°1 95.9) Phatarias.sc2-ce- eee eee eee 240, 241
pectiniferasvAsterinas: se ees eee eee 257 | Phataria (Linckia) fascialis..........-..----- 242
pedicellaris, Acantharchaster variabilis..... . 10,138 | Phataria (Linckia) unifascialis bifascialis.... 242
pedicellaris, Mediaster..................-- 197 ||| Phidiaster <2. ¢--< <2. en cae ee eee eee 71, 78
pedicellaris, Nearchaster ......... 12,915,183; 188) Phidiasterlagassizi- - --- = 9-9 eee eee 71
Pedicellasterssssen eee eee eee one 367 |\\Pislonaster. 4... =2-2 << sess acne nee eee 162, 204
pedicifer, Benthopecten................-. 1495152" |) Phovosters\)o-o. = ean eee eee ee 40

4

“pleyadella, Asterias..............2.0.0.020-

INDEX.

Page
phrygiana, Hippasteria. 223, 224, 230, 234, 236, 237
AIST ERE ii SSP a2) Sms isiare cian «oy we 8

BAI et HSE DS eRe selel stato ceo: /a)aino = = Sniisispaiaie« 105
platyacanthus, Hymenaster............... 384
PR ERaL ABECLAE Eo e ayo! = = ta. a los ania ais «05 260

pleyadella, Goniodiscus..............2..-.-- 168
Plinthaster. .. 160, 162, 165, 167, 171, 172, 205, 219

Plinthaster ceramoideus................... 165
falmnminaster COmptus..-....<---2% --.2~ 165, 205
Plinthaster dentatus.................... 165, 205
PRU DASTOR MIGICUS =. <.5.05.52 ecieccceas 165, 205
Hlmnthaster permieri.......25.2<---<<---- 165, 205
Plutonaster.... 19, 37, 38, 39, 41, 71, 77, 121, 175, 185
Plutonaster abyssicola...........-.- 8, 38, 72, 76, 77
IBIUONGSLED PATEL... 2-2. soe een ene ee 180
OSLO Fon cs cc. 5 4, 19, 37, 38, 159
PM RONLASL ETUC tern = 33) )< ws aini aims winisinieieys es oe 37, 38
LAA PACLCRUIS Janta a= a= 2 = 2 nn nw co cinco 31
polyacantha, Henricia. 9, 13, 14, 268, 8302, 304, 305
polyacanthus, Astropecten............-..- 61
EERE TINA oe a=) hn e = cine cide eiensiaiss ¢ 306
MolpdhelavAsteriass...--.---se-.e-sececes: 9
RRontastenses. 06.) <2. es = 38, 120, 121, 122, 123
TEESE DU IDEM oe = 335 Spatacatein ic cine fos 37, 120
BOTTA ee = - 4, 158, 159, 247, 248, 253, 261
SERENA OR oo 2 Sno Soe 3 uaa wins 4, 251
IerarsTONiOLp Mave raiepaee -\<:- =.= sisiciecien- ss 247, 248
PELE SURE = fas) ~ 2 ae fh 2 a - 259, 260, 261
Poraniopsis echinaster..........----- 260, 261, 265
Porianopsis echinasteroides............----- 260
Paraniopsis niata..-.. 2.22.2... 8, 13, 14, 261
Poraniopsis inflata flexilis......... 10, 11, 261, 265
OSI RE ee 265, 266
SERPS ETIARG EL yoo a.com) <n =e ln'cie <iciciciecinte 23, 24, 35
IEATCPLLONUSIET CTOSSUS=..- . = «<= = -.-=--++0-<<-+ 9, 23
Porcellanaster pacificus......-.-.---------- 8, 29
Porcellanaster (Eremicaster) tenebrarius.... 8,23, 24
Porcellanaster waltharii......-..---------++- 8, 24
Porcellanasteridee ..........------ 4,17, 19, 22, 37
Porcellanasterine ............-.--..---- 4,17, 28
porosissimus, Hymenaster.........--.---- 383
preecoquis, Hymenaster.............------ 383

procurator, Ctenodiscus......---.....---- 31,33, 37
propinquus, Leptychaster.... 10, 12, 14, 43, 53, 54

lho SEG HTyELND Se a er 21,

160, 161, 163, 169, 175, 178, 179, 197
Pseudarchaster alascensis.......------------ 8, 185
Pseudarchaster discus. .....---.-------- 179, 184
Pseudarchaster dissonus -....-- - 10, 12, 15, 180, 192
seudarchaster fallax...........--.------.- 184

417

. Page

Pseudarchaster intermedius............... 180, 184
Pseudarchaster parelii....... 11, 180, 191, 347, 373
Pseudarchaster parelii alascensis....... 12,15, 185

Pseudarchaster pectinifer........ 192, 194, 195, 196
Pseudarchaster pulcher 184, 191
Pseudarchaster pusillus.. 8, 12, 14, 51, 180, 186, 187

Pseudarchaster tessellatus arcticus............ 180
Pseudarchaster verrilli.................-. 191
Resudarchasterma se oo ood. ua eo eee 160
Pseudarchasterine.... 4,37, 159, 160, 161, 162, 179
Pacudasters 5 prewotc 55 enue eet, 23
Peetidoreasters <5 <= ecu dove ceo. 159, 160, 163, 174
Pseudoreaster obtusangulus............- 159, 163
Psilaster......... 32, 38, 39, 40, 71, 77, 78, 79, 81,82
Psilaster floristic = xis onc orate op mes 78
Psilaster pectinatus...............--. 8, 11,38, 72
Pet lastero pRi8 es Sas ww esc aaea eae teee 40, 78, 84
Pteraster..... 252, 253, 344, 367, 368, 370, 375, 384
Pteraster abyssorum..---......--.----<s<- 344
Bterasteratinis 5 a5 .c50c- se. eee 368
Pteraster alveolatiss~<- -c...22..<e-<nceck 370
IEteraster OMOTUS Sa = 2h te ne 7, 10, 344, 346, 347
Pierastenicambbous=.....2..... 855 seanteae 368
Pteraster coscinopeplus. ...............-.- 10,
13, 14, 345, 854, 355, 370

PLEXRStON CSE a oo oa eee ee 370
Pteraster gracilis. ........... 13, 15, 345, 849, 369
Pteraster hebes............- Oe eee 9, 345
TP NELOBLET I ELOACIR tees oon eee a 363
Pteraster (Temnaster) hexactis......--.--- 344, 363
iPreraster anol) ste ee ae 370
BterastbersjOxd SN ai sean ate ea 8,
13, 14, 345, 346, 850, 353, 354, 369

(Rieraster ODrumw eee aoe aa eee 369
Pteraster marsippus... 10, 13, 15, 345, 352, 356, 369
pteraster OUUbanis se ea ama oe ee ll,
345, 346, 349, 350, 353, 354, 356, 368, 373

Pteraster Multipes...< «2 = .2=-02==0s0s= 0m 370, 371
Pteraster multiporus. ..........-------- 355,370
Pteraster multispinus.. 8, 13, 15, 345, 355, 359, 369
(Pieranter ObeSUS* A. «<.<ven ts Joe cans 357, 370
Pteraster obscurus.......-..---- 11, 344, 3638, 370
Pteraster personatus. .......--.-.------+-- 369

. 9,363, 368

Pteraster octaster......--.------+-+-
11, 345, 356, 357, 358, 369

Pteraster pulvillus. . . ..

Pteraster TeGQuCUUS. ac =o < on enema aoe 369
Pteraster reticulatus ...........--------- 355, 369
Pteraster rugatus......--.---.-------------. 370
Pteraster semireticulatus........-.....----. 370
Pteraster s0rdidUs-..-.<. 25+ .s<-0-n00005s 5) ee
Pteraster stellifer...............-------- 346, 37

Pteraster temnochiton...-.. sare 10,

13, 15, 345, 355, 357, 369, 370

418 INDEX.
: Page. Page,
Pteraster tesselatus.... 7,13, 15, 345, 859, 369, 372 | sanguinolenta, Henricia................... te
Pteraster tesselatus arcuatus......-..--.--- 13s 259, 267, 271, 279, 283, 300, 347, 373
14, 345, 863, 369 | sanguinolentus, Echinaster ........-......-.- 271
Pteraster trivonod ons. - 22-26. ee cee einae OSs | (Sarkasters.eseee sae a tee ee eee 306, 334, 337
13, 14, 345, 348, 354, 356, 357, 369: | Sarkaster validus...............-.-.-- 334, 337, 338
Pterasterid ce sa2ase sane eee eres 5/258, 340) 848) ||\sarel, Jouidia.-22 2). 2222-2 eee eae eee 116, 119
Pierastervdes? 228 eer ee eee 1012344: || -sarsv. Pchinaster.. 2) nce ene eee eee 271
pulchellus, Pentagonaster.........-.------ 166) | “Scaphaster esse steno ee sae eee 241
pulcher, Pseudarchaster........--------- 184. 19) | Scotiaster! 4452! 22> 2) 5 ee ee 251
pul villus) Pteraster: 2. 2c c-e see emir 1S | seb" Goniodiscidess:2- = saecee eee eee 168
pulyinus, Ceramaster!--.--2-2-46--22-s2-- 205 | semireticulatus, Pteraster..........-...-...- 370
345, 356, 357, 858, 369 | semisquamatus, Benthopecten.............. 150
PUT CLALO IP CUCU meat eee eee a ete 247 | semituberculatus, Toxopneustes...........- 259
purpureus, Hiymenaster--..-.2 2.2.2 3222-2 .. 884) | sentus: Mehinasten: 222 222525224 55ee eee 261
pusillus, Pseudarchaster. 8, 12, 14,51, 180, 186, 187 | seposita, Asterias..........-.-..-.---------- 271
Pycnopodia helianthoides........ FE ce 6' | senpentarvus, Echinaster... . - 255-2 542ee een 260
Pyrenastenswecnse eee ee cee re 162; 1655) 1716 2055253) | pideriaster=s5--- 2-5. 2 sees eae eee 39, 41
iPythonastenise: oc fecee eee oe pees 253) ||| Sideriaster prandig-. 2.4. 4- -.-2so-seeeeeeee 41
Py thonasterinses=-scer ss acece nee eee 5,.253)'| Sidonasters--2-a- 22-245 49 ee aaa eee 23
quadrispinosus, Hymenaster..........-.--- 8, | simplex, Benthopecten .............---..-- 143
12, 253, 373, 378, 379, 880 | singularis, Asterodon ...........--....-..-- 153
quinaria, "Duidia.- 55-422 --eeeeenececeee 112,116 || singularts, Goniodiscus. ....-2..---2-2 5-222 153
Radiasters 2 nese nese eee 251/253) || sladent)|Cribraster: 222-2 —-|-— --- esse eee 259
Randastas si: shes ee eee 168.|4sladeni-*Dipsacaster=.--2-c--- ses eee 85
Rathbunaster californicus..........-....-- 9) | sny.dert; Cheiraster!=2-22-2--e--- eee eee 127
rathbuni, Alllasteriass2s---- soem eerie neaee QNiWSocomian-ces-cemsncs seen ec soo eee eee 253
reductus-sbteraster-csssscss02 sone eee eeee S69 | solastersa=see serene see 37, 258, 306, 329, 334
MECVESTA, OTIGONGs se neeas aie Rae eee eee 169) |; Solasten/astraliig ess sse- ese see eee 323
regalis., Astropectem:2<-a:=-s-e2- = 9sses = 67,71 | Solaster borealis. . 11, 13, 15, 307, 317, 318, 320, 329
regularis, Solaster...:.....-.--------------- 323 | Solaster constellatus:....-..-.-...-2.2222- 9, 311
Retasteriset ssc. scene 344, 362, 370, 371 | Solaster dawsoni. 7, 11, 13, 15, 307, 310, 318,317, 318
Retaster/capenkish--. notes. ee) eeeeree eae 362 | Solaster decemradiatus.........----.------- 6, 311
Retaster cribrosus!-.--25-42 1 5=oe eee 362 | Solaster endeca.........- 11, 306, 307, 311, 313, 315
Retaster pibber:---.----csasee seca seeeeses 362 | Solaster endeca decemradiata.......---.----- 311
Retastergnacilisiiv-- = 00-2 - sas ee sees ee 8, 349 | Solaster exiguus............-.- 10, 13, 14, 307, 319
Retasterdnsignis:- 2-9-2222 esses sesame 362. || Solaster furcifer.os.- 2... =e 22 ne eee 334
Retasterimultipessa2c.5- 22 s59ee = ese S71 || Solaster,galaxides: = 2252 5-2 es--eeeeee 9, 308
reticulatus;-eterastet--+<5---22-- ee eee 355, 369 | Solaster hypothrissus.......--- 10, 13, 15, 307, 324
Rhegaster <2. scccck ose soos soe eee One ace 248) | \Solaster japonicus: -----5-----5-=--=eeeer 307, 330
Rhipidasters9.. vee oc.s cisco = aa 306 | Solaster papposus. . ----- 11, 306, 307, 320, 325, 331
Ripastersssse sean eee caene eee 38, 39,41 | Solaster paxillatus.... 11, 13, 15, 307, 311, 815, 320
Tien. DUNOUUSLER Ace o-e ee ee eee ence 9°40) || ‘Solaster regulaniso 2-22 - soe =e eee 323
TODUstUS, Bathy biaster: <5 --<---es eee 39) 77, || Solastersquamatus:-o-o2-.- see seer eee 325
rotundus: Clypeaster’s--- 6 seis ee 259 | Solaster stimpsoni-. ......-- 7, 18, 15, 307, 811, 315
ROSasterse- co -cesccesec ones 159, 160, 162, 164, 170 | Solaster syrtensis..........-..-..-.-------- 308
rubens: Asterias:.2.-5.22-..>4-0- = eens ee GullSolasteridses-seseereae- ees 5, 252, 258, 305, 340
rubidus, Astropecten-..--2--.-----sass-e- 67,,7) | sordidus; (Pteraster-—-..--2----- ~<a 369
rudis, Asterias troschelii, var.......-------- 10) | spatulifera, Henricia! << 2... 52 cos2-ne 10, 271
rudis;.Cladasters-2. sasesecesieeshe se eee se 221, 223 | Spheeriodiscus........-..--- 10, 162, 166, 167, 171
rupatus, Pterasters-.-i...0-e ssscee eee 370 | Spheeriodiscus ammophilus . .....-.------- 167
saanichensis, Asterias...........-.---..-.- 7 | Spheeriodiscus bourgeti--.........--------- 167
sacculatus, Zoroaster (Myxoderma)........-- 8 || spicultfera, Cribrella....... 2.2.2 .0.e---== 8, 295
sanguinolenta, Asterias.........------- ees 27/1\)\|\ispicuilifera, Henricia=--4-e ee eee 13, 15, 268, 295
sanguinolenta; \Cribrella 2. eae ais = ne ciel 27] || spinosa, Astenias: 22 cameo areinee eer 261

IN

Page.

gpinosa, Hippasteria....................... 9,

13, 15, 224, 230, 234, 236, 237
spinosus, Benthopecten.......-- 142, 147, 148, 149
spinosus, Dytaster..........- 104
BpInosis phchinaster-.....-.---...---22-05- 261
spinuliger, Benthopecten................... 152
Ie ORV eer =u2-}------=------- 4,5, 16, 20, 251
spinulosus, Echinaster ............-.------- 261
MONMEMMUSIMPASLETIES scoccccsccoscccceeeseoss5 201
DHT RIEN OLS RRR TE pS a a!s)cn1=\o\2[cialn'<)n\</e\s siete o's 2'eie 9
PIIUREUS I CTOSSOSLEY oom aoe awe ee ws = 329
RAMADIRUUE PS SOLSStEr.--..<-..cee ccs oeee--s- 820
PUPPMDACEIse se ccccs ccc cess ccceeseecescee ©6204
IR EIRI ER eos sos 55
PPI RUMUTONRACD oo. o.- «nis oss s+ oe = == 55
PI MNSNE CDi. = <ic12<iasi nice ~ sce. 160, 163, 169, 173
OPTIC. ICI ee 197
Picton bterasterss.--0-- 56. eels ss seecee 370
DURETIUAEYICEE aioe a on) a[5 = 2\= «o's <eo cin'n se == 307
Stellonia papposa.......--- 325
Bere TUES LET ie (arn iciwin == ss > 5 alee a om 166
Stephanaster bourgeti........-..--------- 166, 167
SHEN eKIOSOMeeess ccs acclcecis-ces----- 2ol
stimpsoni, Solaster......... 7, 18, 15, 307, 811, 315
studeri, Goniodiscides................------ 168
SNERRRE A EREE eee o a(n a aie oS See eS - 23
swifti, Gephyreaster............-. 12, 15, 156, 175
awifti, Mimaster.............-........-.-.-- 8,175
RUMUGUMISMOOIASLCK=-2.2.---5<3-2------------ 308
tartareus, Lonchotaster..............------- 42
PETER ETRE a oo 2 ocala se sa 344, 367, 368
Temnaster hexactis........-------- 363
wermnochiton, Pteraster...........-------- 10,

13, 15, 345, 355, 357, 369, 370
tenebrarius, Eremicaster-.----. 8, 11, 24, 25, 29, 35
tenebrarius, Porcellanaster (Eremicaster) .... 8, 24
tenellus, Mediaster............- 8, 11, 197, 202, 204
Renmanpinas Ota. so... <..0-----+2----- 259
MPTSTINUS ECHUNAStE?.. ~~ -- 22-2220 8, 259
RenCH UTIGNAS LOTS eas oc ace eee cne ces 127
RarmuligeNiymiphaster-...-...--.-.-+------- 199
DEMME MUIMPABIETIIS 55 c/s s.c <<.s=.--2--2ss--s- 159, 167
PEROUATIE GONIIStET...-.. 2.05 0-c2ee2-+-- 167
tesselatus, Pteraster.... 7,13, 15, 345, 359, 369, 372
Tethyaster........... Pee ec Scie 38, 39, 41, 185
SENSI OH EC TPR en ayn c.cin a cere a's vis eine oe = 23
Thrissacanthias............ 10,12, 38,39, 40, 78, 84
Thrissacanthias penicillatus.....-....... 12, 14, 79
Dison 5s So 260
Posia...... 156, 160, 162, 166, 167, 172, 205, 219, 222

DEX. 419
Se Ae SCC REE Re ee eet BS st "0k
LOSIG GF Chl = se she ae Ree ee ee eee 9, 219
Moais'australissc. 625 cos oe aes ee ee 162, 166
LOM TEP OCH OMA cree aceasta Le
Dosjaimteropeld - <as25 tea cme sas reese 205
Posiantuberculata- <n. csc sesenceeneees 165, 166
Toxopneustes semituberculatus...........- 259
transfuga, Mediaster...........0.000s--- 197, 204
TOMABtERE Seis ios Satta ee eee eemaae oe 254
ARPeMAsterinte + scone cscceeweeec es. aue ee Oe
trigonodon, Pteraster..............2...---- 10,

13, 14, 345, 848, 354, 356, 357, 369
Myntonaster:.e csi Socwe cosa Ren aenn DEOL
troschelit: A steriass- 2s0- cscs ee sa oe ee 6
lubercolate, Vodids-2sce- es eane eee ees 165, 166
TUMNIAGFENTICUL = <3 25. nee oak Re 10, 276
ADV EReRS pecans See eee a tee, ee aera 247, 248
bY DUS LANCKAD. «a= joan omaha sane ee ee eee 242
Wniophord =< so sa2. ttn tee coos 158
validus, Cladaster.........-.-...--- 10, 13, 15, 222
validus, Pentagonaster............-...------ 168
validus, Sarkaster.............-.-.-- 337, 338, 344
Valvaster: 227 = <.510ccmce sm acin~ sneer 258, 259
Walvasteridia 22. ccns00 se owaceeee eee 5, 252
VGWOSterntNe ooccee << soe ee ee 252
Valvata stone ea sacs enced saa ae 17, 22
Vatvulatane cess = sane 232 2 eel 17
vancouverensis, Crossaster........--.-.------ 7,3l1
Wancoulyerl. ASt@rlas > 30-225 sae ee ee 7
variabilis, Acantharchaster.......---------- 10, 137
variabilis, Nearchaster.......- 12, 15, 132, 133, 187
wernicina, Asterias.........-----------++---- 2AT
vernicina, Asteropsis.....--.-.-=---------- 247
VeImicinas,PewWiCiNaacs= seco een acinar ees 247
verrli; Astropecten.-.-:------<-c ose cecum 67, 71

verrilli, Pseudarchaster. ......-..--------- 191

| verrucosus, Diplopteraster.....-..----.-- 370, 371
vestituss Ohsetaster. . ssi ees <ncre «=e 21
vexator, Lophaster furcilliger. 10, 11, 334, 337, 388
vexillifer, Bathybiaster............-.------ 77
vicarius, Hymenaster.......-.-.---------- 37
victoriana, Asterias...........------------ 9
vincenti, Luidiaster..............--------- 27
violaceus, Hymenaster......-..----------- 375
waltharii, Eremicaster.....-------++------+++- 25
waltharii, Porcellanaster. ....-------------- 8, 24
Zoroaster ophiurus.....-.-.--------+--++-+- 8
Zoroaster (Myxoderma) evermanni .....---. 8
Zoroaster (Myxoderma) sacculatus....--.--- 8

| Zoroasteridse. ......-------- 22-2 eee eeeees 4, 251

O

RTS
ee Ngee ies aL ae

Cue ay
io u FSG Sa pe

De bes Ws Pies wae eA
i de oe Cas ae o ne es oe oe

ely ong OG aoe a i thy . a
se Bee AC hee hee Lee oe PR
ae poe a Toa
fe sab: : aS PEM Le ed a uz os ie
Coe co on Cs coe ie: a
Res
RS QT SEU oe. CNA
oe Gig Co pet
Mey ny ce one an ares or a7 ee :

a voces aes -— ot Z
Doe SnD Ree
a ss ee ee Me Ye Oe Ke

Wi AY oe Heme Cas A a?

sees Ae cae oon a coe ey era

sor,

Cy SAL | mL
4) ie

HH) a .
Se hers I be Se eer ie EN oe ee G
Ss Sx et ve Bue EG ee ae RE Wi nS
BN se en

x bit aS
star An PS YR) nae 5
a. () a Tc KY eo. Sait PRS
R \ SS Sige. eR S EU ee et
Oi AYRES 4
AO ag

ay YS QA
at (Se FN At i
\ Je Ww nS. Oe sy se Bee vo my pias ys
rk LS Ory ni ore pe
eS tke oo Yeo Ce Has ee Sy
Borat oa FAL es aa sen ene SNe 4 Savsits Pe
See SPSS: Sa see
vO ay RANG RE aes aoe Pak.
ike A $3 A ee |

= oe Not PAS ae
po HS Die Ae S tek tee
cox fg Fo UL So Je ee BEC =a s
. od m cone orAL are ‘Ges EK OS: AR i Sess: ::
5 ES ye Mae FD Oe SZ aS S ee :
ie oe seas Begs 2 ae on AE
ae eek ae Pee Seve

aed ge ee ERR ay SA

“WiLL