aes si
als

sedi

fesbeconteactt
cars isons
faptaeesetrstttersy sss —

sesteen tay

towel ts
rote eater
z

a:
oa
ee

tH
saetees test hes
w

brig tasedores
Sone a) +1
shore geeetelysgtiit ct
jeanne
Ma
hy

Sy
‘o

Digitized by the Internet Archive
in 2010 with funding from
University of Toronto

http://www.archive.org/details/nemoirs41harv

af

al:
=

MEMOIRS

OF THE

MUSEUM OF COMPARATIVE ZOOLOGY

AT

HARVARD COLLEGE.

VOL. XLI.

/Ly Ke foe O
et Me) l4
CAMBRIDGE, U.S.A.

PRINTED FOR THE MUSEUM.
1910.

ae ee
Was
Ne)

CONTENTS.

TS ON THE SCIENTIFIC RESULTS OF THE EXPEDITION TO THE

STERN TROPICAL PACIFIC, in charge of ALEXANDER AGassiz, by the U.S.
Fish Commission Steamer “Albatross,” from October, 1904, to March, 1905, Lieut.

hy Commander L. M. Garrett, U. S. N., Commanding, and of other expeditions of
the “Albatross,” 1888-1904. XXI. Tue Sponges. By Roserr von LENDENFELD

1. The Geodidae. 260 pp., 48 Plates. August, 1910.
2. The Erylidae. 63 pp., 8 Plates. September, 1910.

is,"

CONTENTS:

No. 1. REPORTS on the Scientific Results of the Expedition to the Eastern Tropical
Pacific, in Charge of ALEXANDER AGassiz, by the U. S. Fish Commission Steamer
“ Albatross,” from October, 1904, to March, 1905, Lieut. Commander L. M. Garrett,
U.S. N., Commanding, and of other Expeditions of the “ Albatross,” 1888-1904. XXI.
THe Sproncres. 1. Tur Gropipar. By Rosperr von LeNDENFELD. 260 pp. 48
plates. August, 1910.

TABLE OF CONTENTS.

I. INTRODUCTION. :
II. DESCRIPTION OF THE SPECIES COLLECTED “BY THE ce ALB. \TROSS”
GEODIDAE
Srponops A
S. californica (Plate 0; ene. 1- _37)
5. angulata - : : : :
var. megana (Plate 12, figs. 1l- s, 16, 17, 19, 20; Plate 13, figs. 1-12, 22-25;
Plate 14, figs. 1-6, 16-22; Plate 15, figs. 1-4, 7-9, 11)
var. microana (Plate 12, figs. 11-15, 18, 21, 22; Plate 13, figs. 13-17, 21;
Plate 14, figs. 7-9; Plate 15, fig. 10)
var. orthotriaena (Plate 12, figs. 9, 10: Plate 13, figs. 18-20; Pl: ie M4, 4.
10-15, 23- 30; Plate 15, figs. 5, 6, 12) 5 :
S. oxyastra (Plate 6, figs. 1-23; Plate 7, figs. 1-20; Plate 8, figs. 1— 15)
S. bicolor (Plate 9, Cae 1-19; Plate 10, figs. 1-15; Plate 11, figs. 1-17)
GEODIA : ; : : , 5 - :
G. Spepenlos
var. intermedia (Plate 17, ie 23- 26, 34-40, 49: Pl: ate 18, figs 8, ‘10, 13- 20,
22, 27; Plate 19, figs. 9-11, 19, 20, 22, 24, 31) Z
var. micraster (Plate 17, figs. 27-33, 41-48, 50; Plate 18, figs. 1-7, 9, 11, 12,
21, 23-26; Plate 19, figs. 1-8, 12-18, 21, 23, 25-30, 32)
G. japonica (Plate 37, figs. 15-30; Plate 38, figs. 1-29; Plate 39, figs. 1-41)
G. ataxastra
var. mae (Plate 43, oe 9-25, 28-38: Plate 4, figs. “ae 12; 14-49)
var. latana (Plate 43, figs. 26, 27; Plate 44, fig. 13) 5
G. mesotriaena .
var. pachana (Plate 21, fig. ts Plate 23, Act, 3, 5, 6, 8, velar 24, fee 3, 5, 9)
var. microana (Plate 23, ASS 1-2; Plate 24, figs. 2, ae , 10-13. 16, 19, 21)
var. megana (Plate 21, figs. 2-6; Plate 22, figs. 1-10; Plate 23, figs. 4, 7, 10-25;
Plate 24, figs. 1, 4, 8, 14, 15, 17, 18, 20, 22-32; Plate 25,
figs. 1-11) : . é A 7 4 ;
G. agassizii (Plate 26, figs. 1-21; Plate 27, figs. 1-19; Plate 28, figs. 1-28; Plate 29,
figs. 1-17; Plate 30, figs. 1-17; Plate 31, figs. 1-10; Plate
32, figs. 146; Plate 33, figs. 1-14; Plate 34, figs. 1-17)

G. mesotriaenella (Plate 34, figs. 18-26; Plate 35, figs. 28-35)

G. breviana (Plate 35, figs. 1-27; Plate 36, figs. 1-12)

G. ovis (Plate 40, figs. 1-30; Plate 41, figs. 1-20; Plate 2, figs. rs 40: Plate 43,

figs. 1-8) . ns 3 : 4 :

G. micropora (Plate 36, figs. 13-36; Plate 37, Bap ce 14)

G. amphistrongyla (Plate 20, figs. 141) 5

G. lophotriaena (Plate 47, figs. 9-36; Plate 48, fice 1 34) : : : :

G. acanthtylastra (Plate 45, figs. 1-39; Plate 46, figs. 1-21; Plate 47, figs. 1-8)

G. media (Plate 16, figs. 1-21; Plate 17, figs. 1-22 3 5 5 5
GEODINELLA é : 2 : ;

G. robusta

var. median (Plate 1, a, 1 aa 16, ‘18- 24. Plate 2. figs. 1, 3; Plate 3,
figs. 3, 4, 7, 9; Plate 4, figs. 1, 4-7, 13, 21, 22)

var. carolae (Plate 1, figs. 5-12, 17; Plate 2, figs. 4-7, 9-11; Pk ate 3, a L:
Plate 4, figs. 2, 3, 8-12, 14-20) . : ;

var. megaclada (Plate 1, figs. 13-15; Plate 2, figs. 2-8; Plate 3, figs. 2, 5, 6, 8:
Plate 4, figs. 23-25) : : : : 3 :

7

or
bo ot

asa
oo

No}

8

Ill. GENERAL SYSTEMATIC ACCOUNT OF THE GENERA, SPECIES, AND VARIE-—
TIES OF GEODIDAE FROM THE PACIFIC OCEAN : : 2 5 0

GEODIDAE
CAMINELLA

C.

nigra .

CaMINUS

C.
Tsops

chinensis

I. contorta
I. imperfecta .
I. obseura

1. sollasi

SIDONO:
Ss.
Ss.
Ss.
5.

PS

lindgreni
picteti
californica .

alba

S. angulata

var. megana
var. microana

var. orthotriaena
S. oxyastra

S. reticulata
S. bicolor
S. nitida .

GEODIA .

G.

22

*

Q

G.
G.

’
G.
G.
G.

G.

nux

. eosaster

. globostella

1. distineta

. erinaceus

i. Variospiculosa

var. typica
var. clavigera
var. intermedia

var. micraster .

reniformis
japonica

. cooksoni

hilgendorfi
var. typica
var. granosa

3. ataxastra

var. angustana
var. latana
hirsuta
mesotriaena
var. pachana
var. megana

var. microana .

agassizil
mesotriaenella

. breviana

ovis A
sphaeroides
micropora .

. berryi

kiikenthali

CONTENTS.

Summary of the species found in the Pacifie Ocean

he ee 7S ~ 6 - a ee a 6 ee

be ptote Oly “tle Racue lt <A ere le a cae ee 2

OF THE PLATES”

”

GEODIDAE.

I. INTRODUCTION.

Tus monograph is the first. part of a report on the Siliceous Sponges col-
lected during the cruises of the ‘‘Albatross” in the Pacific Ocean. It deals
with the Geodidae and contains detailed descriptions of the species and varieties
of this family represented in the collection, a systematic account of the known
Pacific Geodidae, and a discussion on their distribution.

There are altogether eighty-nine specimens in the collection. Two of them
are dry, all the others are preserved in spirit. To make the identification of
these sponges positive and to be able to give fuller diagnoses of some species
which were insufficiently described, I have in addition to those collected by the
“‘ Albatross’ examined several type specimens in the British Museum and in
the collection of the Geological Survey of Canada.

In my report on the Tetraxonia of the Deutsche Tiefsee-Expedition I gave
an account ' of my method of fractional sedimentation for obtaining spicule-
preparations. This method has been employed also in preparing the material
for the present report.

For the graphic representation of sponges and of their parts, photographs
reproduced mechanically (phototypically) are, in my opinion, far superior to
drawings reproduced lithographically. It is not only that drawings never can
be so accurate as photographs, but in drawing such objects as sections or spicules
of sponges the author who draws them or has them drawn under his supervision,
and who sees that the points he considers important are correctly reproduced,

1R. ». Lendenfeld. Die Tetraxonia. Wissensch. ergebn. deutschen Tiefsee-Expedition, 1898-1899.
1907, 11, p. 82.

11

12 INTRODUCTION.

does not always pay sufficient attention to those other points which seem unim-
portant. This will often lead to the omission or inadequate representation in
drawings of such characters as seem unimportant, whereas, in the mechanically
reproduced photographs, they will be just as fully and correctly shown as all
other points. It is quite probable, that at some future time, when science
is more advanced, one or more of the points considered unimportant now, and
consequently not carefully represented in drawings, may become important.
These considerations induced me to represent the sponges and their parts
in my reports on the Tetraxonia of the Deutsche Tiefsee-Expedition (Lendenfeld,
loc. cit.), and the Deutsche Siidpolar-Expedition ' so far as possible photo-
graphically. During the years I was engaged in preparing these reports I
gained considerable experience in this photographic work. By the construction
of a diaphragm, placed just above the objective, I overcame the difficulties
» which had formerly prevented a photographic reproduction of the megascleres;
and though I was hampered by the quality of my microphotographic outfit
and the insufficiency of the light at my disposal I obtained quite satisfactory
photographs of these spicules. My attempts similarly to delineate the micro-
scleres were not nearly so successful and I was frequently obliged to draw them.
When [ began the study of the Sponges collected by the “Albatross” I ob-
tained in the first place a first-class microphotographic outfit, replaced the
Welsbach burner, formerly used as a source of light, by a Nernst lamp with
crossed rods and was thus enabled to improve considerably the microphoto-
graphic reproductions of the sections and megascleres. The difficulty of photo-
graphing the microscleres with high powers, however, still remained. Like the
megascleres the microscleres consist of colourless, transparent silica and are
rendered visible only by the difference in the refractive indices of the silica and
the surrounding medium. When, as is the case in the spines and other parts
of most microscleres, such colourless structures are very thin, less than a light
wave-length in diameter, no sharply defined image of them can be produced
on the photographic plate, no matter how excellent the lenses of the micro-
scope may be. The only way to get well-defined images of them is to make use
of light of shorter wave-length; the lower limit of the size (thickness) of
minute structures of this kind, still clearly reproducible microphotographically,
must, celeris paribus, be in inverse proportion to the wave-length of the light
employed. The shortest light-waves obtainable would accordingly be the best
for work of thiskind. Since, however, the employment of very short light-waves

'R. ». Lendenfeld. Tetraxonia. Deutsche Stidpolar-Expedition, 1901-1903, 1907, 9.

INTRODUCTION. 13

for microphotography is connected with great technical difficulties, we must
for the present content ourselves with light-waves 275-280 y» long, that is
about half of the length of the waves of ordinary light. Even this light is not
convenient to work with, because it is quite invisible to the human eye and does
not pass through glass. All the optical parts of the apparatus used for its
employment must be made of some material like quartz, and fluorescent sub-
stances like uranium-glass must be utilized for making the image obtained
visible and focusable. An apparatus for microphotography with light of these
wave-lengths has recently been made by Koehler (Zeiss) of Jena.

To overcome the difficulties previously experienced in photographing the
microscleres I obtained this apparatus and succeeded, after a number of trials,
in producing satisfactory photographs also of these spicules.

To work this apparatus 4.8-5 amperes of the ordinary three phase-current
of 120-125 volts supplied by the Prag Municipal Works were employed. This
current is converted, in a transformer, to one of from 15,000 to 15,560 volts,
which is made to pass, in sparks, between cadmium- or magnesium-electrodes.
The sparks are strengthened by eight Leyden jars attached to the circuit, and
are very brilliant. For photographing the microscleres of sponges I soon found
magnesium-electrodes more suitable than cadmium-electrodes, and I have
worked with the former ever since. The magnesium spark-light produced
passes through a quartz-lens and two quartz-prisms. In the ultraviolet part of
the magnesium-spectrum thus obtained there is an exceedingly intense line
produced by light of a wave-length of 280 wy. This light was used. It passes
through another quartz-lens, is reflected upwards by a quartz-prism placed
below the microscope, and concentrated and thrown on the preparation by a
quartz-condenser.

After many trials I found the following the best way to make the micro-
sclere-preparations to be photographed with this light: to make a sterraster-
preparation, a quartz-slide is covered by a thin layer of gum and, before the
gum is dry, a number of these microscleres, previously obtained by sedimenta-
tion, are allowed to fall on the gummed quartz-slide. This is then placed in
the thermostat oven. When it is quite dry, a small drop of a concentrated
solution of chloral hydrate in glycerine, a liquid with sufficiently high refractive
index and permeable to these ultraviolet rays, is put on and the whole covered
with a quartz-cover. To make a preparation of the euasters the minute spic-
ules obtained by centrifuging are spread out on an object glass and dried,
whereupon they appear quite firmly attached to the glass. A portion of these

14 INTRODUCTION.

microscleres is then scraped off with a knife and immersed in a small drop of
water placed on the quartz-slide. In this water they disperse and when they
are scattered a little gum may be added and the whole allowed to dry. The
spicules then appear attached to the quartz-slide. A small drop of the chloral
hydrate glycerine is put on and this covered with a quartz-cover.

The slide thus prepared is placed on the stage of the microscope, and a
spicule to be photographed sought with ordinary light and ordinary lenses.
When a spicule of which a photograph is desired is found, the ordinary lenses
are replaced by a quartz-objective and a quartz-eyepiece, the spicule brought
to focus and centred, the ordinary light switched off and the magnesium-u.
v. light switched on. A dise of uranium-glass with a lens attached at a slightly
oblique angle is then placed over the eyepiece and the tube of the microscope
lowered until a sharp fluorescence-image of the spicule becomes visible on the

wiranium-glass. When the spicule is thus focused for the 280 py-light the
uranium-glass arrangement is removed and the camera placed over the micro-
scope.

As it is difficult to focus quite correctly with this fluorescent arrangement
and as in many cases the shape of the microscleres can be clearly revealed only
by a series of photographs obtained by focusing at different levels, I usually
took four differently focused photographs of each spicule. After having focused
the spicule and put the camera on, I usually raised the tube of the microscope
with the micrometer-screw 3 4. At this level I took the first photograph.
I then lowered the tube 2 » and took a second photograph, and so on, the third
4 » and the fourth 6 « lower than the first. The intervals between these levels
were sometimes more, sometimes less. Four such photographs 6 by 9 em. can
easily be taken rapidly in succession by means of a sliding-plate arrangement.

I found the most useful combination of lenses to be immersion-quartz
monochromat 1.7 mm., quartz-eyepiece 10. This gives, with a suitable length
of camera, a magnification of 1800. Most of the u. v. photographs of the micro-
seleres on the plates accompanying this report have been taken with this com-
bination.

I have taken some u. v. photographs without an eyepiece. In these cases
the focusing was done by means of a uranium-glass with attached lens, placed
in a ruler-like frame, which was laid across the (open) top of the camera. In
this way excellently defined photographs can be procured but, the magnification
even with the 1.7 mm. quartz monochromat and the camera drawn out long,
is insufficient.

INTRODUCTION. 15

The time of exposure, even with the aperture in the diaphragm of the
condenser quite small, is very short, for magnifications of about 100 only about
one second, for magnification of 1800, 15-80 seconds.

I must record my obligations to the late Mr. Alexander Agassiz and to the
Smithsonian Institution for placing this large and interesting collection of
sponges at my disposal; and to Mr. R. Kirkpatrick of the British Museum
and Mr. L. M. Lambe of the Geological Survey of Canada, for the types they
sent me for examination and comparison.

With all this material and with all my energy and experience, however,
this report could not have been produced had not Mr. Agassiz supplied the
costly apparatus necessary and in every way furthered the work with so
generous a liberality. ;

The method of microphotography with ultraviolet light, employed in this
monograph for the delineation of the microscleres, will, I do not doubt, when
utilized more generally for the representation of minute structures of this kind,
increase the exactitude, and thus raise the standard of descriptive zodlogy; and
am most grateful to Mr. Agassiz for enabling me to introduce the method into
this branch of science and thus to assist in its advancement.

«

II. DESCRIPTION OF THE SPECIES COLLECTED BY THE
“ ALBATROSS.”

Geodidae.

Tetraxonia with rhabd, teloclade, and usually also mesoclade, megascleres,
and a superficial armour composed of massive, spheroidal or ellipsoidal sterr-
asters. Kuasters are always, ataxasters or microrhabds sometimes, present.
Without desme megascleres and without thin, disc-shaped sterrasters.

This family was established by Gray ' under the name of Geodiadae. It
has been retained by all later authors under the same or a similar name. Sollas ”
included in the family the genera Erylus Gray, Caminus O. Schmidt, Pachy-
matisma Johnston, Cydonium Fleming, Geodia Lamarck, Synops Vosmaer, and
Isops Sollas. Later Sollas* substituted the name Sidonops for Synops. I 4
united Cydonium with Geodia and established the new genera Caminella *
and Geodinella.’ The examination of the sponges collected by the “‘ Albatross”
belonging to these genera shows that Erylus differs very considerably from
Geodia, the type genus of the Geodidae, and should form a separate family,
Erylidae.

In the extent I now give the Geodidae it comprises the genera Caminella
Lendenfeld, Pachymatisma Johnston, Caminus O. Schmidt, Isops Sollas, Sido-
nops Sollas, Geodia Lamarck, and Geodinella Lendenfeld.

Three of these genera, Sidonops, Geodia, and Geodinella, are represented in
the collection made by the ‘‘ Albatross.” There are eighteen different species,
five of which are further subdivided into thirteen varieties.

1 J. E. Gray. Notes on the arrangement of sponges.... Proc. Zool. soe. London, 1867, 492-558.

?W.J.Sollas. Tetractinellida. Rept. voy. “Challenger,” 1888, 25, p. exlvii—cxlix.
’W. J. Sollas. On the geodine genera Synops Vosmaer and Sidonops, a correction. Proc. Roy.
Dublin soe., 1889, 6, p. 277.

*R. v. Lendenfeld. Die Tetractinelliden der Adria. Denk. Akad. wissensch. Wien, 1894, 61,

p- 179.

5 R.v. Lendenfeld. Loc. cit., p. 150.
® R. v. Lendenfeld. Tetraxonia. Tierreich, 1903, 19, p. 117.

17

18 SIDONOPS CALIFORNICA.

Of the eighteen species, three have been previously described, and fifteen
are new. In one of the previously known species two new varieties are
described.

SIDONOPS So ttas.

Among the megascleres are regular triaenes. The tetraxon megascleres
are confined to the superficial part of the sponge and are arranged radially. The
dermal microscleres are asters. The afferents are cribriporal; the efferents
uniporal.

There are twenty-three specimens of Sidonops in the collection made by
the ‘‘Albatross.”” These belong to four species; all are new, and one is divided

into three varieties.
Sidonops californica, sp. nov.

Plate 5, figs. 1-37.

Shape and size. The two specimens in the collection were obtained off
Lower California, and to this locality the specific name refers. They are some-
what fragmentary. Both are elongate tuberous, somewhat finger shaped. The
larger (Plate 5, fig. 6) is 24 mm. long and 6-10 mm. broad, the other measures
187-9 mm. Both appear to be digitate processes, broken off from a larger
mass; the smaller one was attached at one side. The surface is undulating,
and no trace of a spicule-fur can be made out with the unassisted eye. The
microscopic investigation of radial sections, however, shows that minute dermal
styles protrude slightly beyond it. In several places circular efferent pores,
100-300 » wide, are observed. These pores are not numerous. They congre-
gate in groups, one of which is situated on the rounded tip of the larger speci-
men. Parts of the surface are covered with a thin desmacidonid sponge-crust.

The colour (in spirit) is yellowish white, slightly darker in the interior than
on the surface.

The cortex is about 500 « thick and composed of a sterraster-armour.

Canal-system. Radial canals, 100-800 » wide, traverse the sterraster-
armour. Most of these are covered distally with sieve-membranes, the pores
of which are oval, 30-100 » wide, and occasionally so close together as to be
separated only by slender threads. These threads exhibit, when observed with
higher powers, a longitudinal striation. The radial cortical canals not covered
by such sieves, which I consider as efferents, form the groups above described.
The remainder of the surface is occupied by the afferent, cribriporal, cortical

SIDONOPS CALIFORNICA. 19

canals. The radial cortical canals, both the afferent and the efferent, are
cylindrical and restricted below by chonal sphincters, which usually protrude
slightly into the choanosome. Below these chones subcortical cavities (Plate
6, fig. 27b) are met with.

The skeleton of the inner parts of the choanosome consists of rather irregu-
larly scattered amphioxes (Plate 5, fig. 27e), a few styles, and numerous large
oxyasters and sterrasters (Plate 5, fig. 27d). Orthoplagiotriaenes, similar to the
subcortical ones, are also occasionally found in the depth of the choanosome.
It seems doubtful however whether these are here in their natural postion;
they may very likely have been carried into the interior in cutting the sections.
The remarkable abundance of sterrasters in the choanosome on the other hand
(Plate 5, fig. 27) is without doubt natural. Towards the surface the mega-
scleres join to form radial strands which abut more or less vertically on the
cortex. These strands (Plate 5, fig. 27c) are composed chiefly of amphioxes
and orthoplagiotriaenes. Anatriaenes, anatriaenederivates, and mesoplagio-
clades often with reduced cladomes also occur in them, but in much smaller
numbers. The cladomes of the orthoplagiotriaenes lie at the limit between the
cortex and the choanosome. The cortex is occupied by dense masses of sterr-
asters. Small dermal styles are implanted more or less obliquely into its
superficial (distal) part. The rounded ends of the styles are situated proxi-
mally; their distal, pointed ends protrude freely beyond the surface. These
spicules are not very numerous and form tuft-like groups. Those near the
afferent pore-sieves and the efferent pores incline towards these apertures and
thus form protecting fringes. Groups of such spicules, and single ones, are also
met with in the distal part of the choanosome. Numerous minute strongylo-
sphaerasters, forming a dense single layer just below the outer surface, are
imbedded in the dermal membrane. A few small oxysphaerasters also occur.
Besides these numerous other forms of spicules are observed in the spicule-
preparations. Most of these probably belong to the incrusting desmacidonid;
there is one, however, an exceedingly minute and slender microamphiox, about
which I have my doubts. This may be proper to the sponge and possibly forms
dragmes within it. I have, however, not succeeded in finding any of these spic-
ules, either singly or in dragmes, in situ in the sections.

The large choanosomal amphiozes (Plate 5, figs. 11, 12, 27e) are straight or
slightly curved, fairly isoactine, and rather abruptly and quite sharply pointed.
They are 1.2-2, usually 1.6-1.8 mm. long, and 30-48, usually 38-42 » thick.

The rare large styles are shorter and thicker than the amphioxes; some

attain a transverse diameter of 55 » at the rounded end.

20 SIDONOPS CALIFORNICA.

The minute, dermal styles are usually slightly curved, 175-290 » long and in
the central parts 3-7 « thick. They taper slightly towards the proximal rounded
end. The distal end is sharp pointed.

The minute microamphioxes which may be foreign to the sponge, and possi-
bly form dragmes, are quite straight, about 50 » long and 1 thick.

The orthoplagiotriaenes (Plate 5, figs. 13-19) have a rather small clade-angle
and might therefore perhaps also be termed orthotriaenes. The rhabdome
is straight, conic, and usually sharp pointed, rarely blunt. It is 0.9-1.45 mm.
long and at the cladomal end 20-78, usually 35-55 » thick. I do not think that
the great differences in the thickness of the rhabdome, which are clearly notice-
able in comparing the spicules represented in figures 13 and 18 on Plate 5, can
be altogether ascribed to differences in their age. The clades are 160-400 # long,
conic, often rather blunt, and uniformly curved, concave to the rhabdome.
Their chords enclose angles of 104-120° with the axis of the rhabdome. The
three clades of the same cladome often differ in size and sometimes exhibit a
sagittal character.

The anatriaenes (Plate 5, figs. 1-4) have long and more or less curved
rhabdomes, which are 10-17 » thick at the cladomal end. The clades are rela-
tively stout and 22-45 » long. Their proximal part is quite strongly curved,
their distal part straight. Their chords enclose angles of 45-66° with the axis
of the rhabdome. A slight apical knob is usually discernible on the summit of
the cladome.

Anatriaene-derivates of similar dimensions with two clades (anadiaenes)
(Plate 5, fig. 5) and one clade (anamonaenes) also occur, but they are rare.
The branched end of a very peculiar spicule, which may be an anatriaene-
derivate, is represented in fig. 10 on Plate 6. This spicule is a rhabd with two
small, recurved, clade-like branches, arising a little below one of the pointed
ends, and a very large straight branch-ray, also pointing downwards and en-
closing an exceedingly small angle (about 4°) with the rhabdome, arising some
distance below the small branch-rays.

The mesoplagioclades (Plate 5, fig. 7-9) have a long rhabdome which is,
just below the clades, 6-15 » thick. The epirhabd is conic and 45-125 » long.
The number of clades is one (Plate 5, fig. 7), two (Plate 5, fig. 9) or three (Plate
5, fig. 8). The clades are nearly straight, pointed or rounded at the end, and
20-42 » long. They are directed obliquely upwards and their axes enclose
angles of 102-118° with the axis of the rhabdome. These plagioclade spicules
replace the mesoprotriaenes of other geodine sponges, and I am inelined to

SIDONOPS CALIFORNICA. 21

consider them as derivates of ordinary mesoproclades produced by a change in
the position of the clades.

The large choanosomal oxyasters (Plate 5, figs. 21-26c, 3lc, 28, 29) have
from six to fourteen, rarely as many as twenty straight, conic, and quite uni-
formly distributed rays. There is no central thickening. The rays vary from
sharp pointed (Plate 5, fig. 28) to blunt (Plate 5, fig. 29). Apart from the very
short, smooth, basal parts, the rays are entirely covered with spines, which
appear to be directed backwards, towards the centre of the spicule. On the
apex of the blunt rays a terminal spine is usually observed (Plate 5, fig. 29).
The rays are at the base 1.7-3 » thick. The whole aster is 22-48 y in diameter.
An inverse proportion between size and number of rays is clearly pronounced.
The oxyasters with six to eleven rays are 27-48 , those with twelve or more
rays 22-26 » in diameter.

The rare, small oxysphaerasters (Plate 5, fig. 30b) have stout, pointed,
conical, spined rays and measure 7-9 in total diameter.

The small strongylosphaerasters (Plate 5, figs. 24a, 26a, 30a, 31a, 32-35)
have a centrum 2-3.5 » in diameter, from which six to seventeen rays arise.
These are usually regularly, more rarely irregularly distributed and 1.6-2.8 »
long. They are cylindrical, 0.8-1.5 » thick, and rounded at the ends. They
always bear spines, which are either quite uniformly distributed or massed at
the ends, where a verticil of larger spines sometimes appears to be present.
A conspicuous terminal spine, arising from the end of the rays, is often ob-
served. The whole aster measures 4.5-9 «in diameter. A correlation between
number of the rays and size is not discernible.

The sterrasters (Plate 5, figs. 20, 36, 37) are flattened ellipsoids, 116-130 »
long, 97-105 # broad, and 70-90 » thick. The proportion of length to breadth
to thickness is usually about 100: 81:69. In the centre a nearly spherical
cluster, 4 » in diameter, composed of numerous small but conspicuous granules,
is met with. In the youngest sterrasters observed, which appear as spheres of
slender rays, 18 # in diameter, this central cluster of granules has the same
size and structure as in those fully developed. The umbilicus lies in the centre
of one of the flat faces of the ellipsoid. It is about 12 « deep and 12 to 15 p
broad. In the great majority of sterrasters the free distal ends of the rays are
uniformly distributed, 2-3 » thick and 1.5-2 » apart. In a small minority,
perhaps 2 % of all the sterrasters, these free ray-ends are irregularly distributed
and in some places much farther apart, more or less extensive parts of the sur-
face of such sterrasters being free from them. These altogether rayless parts of

22 SIDONOPS CALIFORNICA.

the sterraster surface are covered with spines. In the normal sterrasters each
ray bears a terminal verticil of from three to seven conic not very stout spines
about 1.5 « long (Plate 5, figs. 36, 37).

The two specimens of this species were caught with the tangles at Station
2829 on May 1, 1888, off Lower California, in 22° 52’ N., 109° 55’ W., depth
56 m. (31 f.); they grew ona rocky bottom; the bottom temperature was 23.4°
(74.1° F.).

Should the uniporality of the efferent cortical canals observed be due merely
to a local rubbing off of the superficial parts after capture, and should sieve-
membranes cover them in the living state, this sponge would of course have to
be placed in Geodia. Since however no indication of the former presence of
sieve-membranes can be discovered at the mouths of the now uniporal efferent
cortical canals, I think that these efferents must by nature be uniporal and the
sponge accordingly placed in Sidonops.

Among the species of Geodia and Sidonops hitherto described there are only
three which at all resemble these sponges: Geodia ramodigitata Carter 1880;
the sponge described by Dendy * as Geodia ramodigitata Carter, which differs
however so considerably from Carter’s type that I do not think it specifically
identical with it; and Synops alba Kieschnick 1896 = Sydonops (recte Sidonops)
alba Thiele 1900. From Geodia ramodigitata Carter and also from the species
described under this name by Dendy the above mentioned Californian specimens
differ by possessing plagiomesoclades often with more or less reduced clades in-
stead of the normally developed protriaenes or promesotriaenes, a difference
which is in itself, apart from the difference in the superficial part of the canal
system, quite sufficient for specific distinction. Sidonops alba (Kieschnick)
Thiele* is obviously much more closely allied to them. Most of the spicules are
identical in shape and not very different in size. The differences between them
most important systematically appear to be that Sidonops alba (Kieschnick)
Thiele possesses small anaclades, which Thiele terms exotyles, whilst the Cali-
fornian specimens are destitute of such spicules; that the latter contain minute
microamphioxes which are absent in the former; and that the reduction of the
mesoclade-cladomes is carried considerably further in the former than in the
latter. As regards the minute microamphioxes I do not attach very much
systematic importance to their presence or absence because it is quite possible

‘A. Dendy. Report on the sponges. Rept. pearl oyster fisheries. 1905, pt. 3, p. 88.
* J. Thiele. Kieselschwamme von Ternate I. Abhandl. Senckenb. gesellsch., 1900, 25, p. 46, pl. 2,
fig. 16.

SIDONOPS CALIFORNICA. 23

“=

that these spicules are foreign to the sponge. The presence of minute anaclades
(exotyles) and the far greater reduction of the mesoclade-cladomes in Sidonops
alba (Kieschnick) Thiele seem, together with the differences in the dimensions
of the other spicules, sufficient for specific separation; and the more so as
Sidonops alba (Kieschnick) Thiele occurs near Ternate while the specimens
described above have been found on the opposite side of the Pacific Ocean, off

the coast of Lower California. The differences between these closely allied

species are given in the following table.

SIDONOPS ALBA

SIDONOPS CALIFORNICA

Large choanosomal
amphioxes.

2.5mm. and more long; 30 » thick.

1.22 mm. long; 30-48 » thick.

Large choanosomal
styles.

Small dermal styles.

similar to the amphioxes, but shorter

(2)

shorter than the amphioxes; up to
55 p thick.

250 «long; 5 » thick.

175-290 wlong; 3-7 » thick.

Minute microamphi- | absent. 50 «long; 1 thick; perhaps foreign.
oxes.
Plagiotriaenes. thabdome 2 mm. long; about 50 » | rhabdome 0.9-1.45 mm. long; 20—

thick; clades 450 y» long; clade-

angles a little over 90°.

78 y» thick; clades 160-400 y long,
clade-angles 104-120°.

Ordinary anatriaenes.

rhabdome 2.5 mm. long; 14 « thick;
clades about 20 » long; clade-angles
44°,

rhabdome long; 10-17 » thick; clades
22-45 » long; clade-angles 45-66°.

Mesoclades.

rhabdome 3 mm. long; 14 » thick;
epirhabd 80 » long; 1-2 clades 18-
21 « long; clade-angles 84-102°.

rhabdome long; 6-15 yp thick; epi-
rhabd 45-125 » long; 1-3 clades
20-42 » long; clade-angles 102—118°.

Minute anaclades
(exotyles).

170 » long and over; with 1-3 clades.

absent.

Large oxyasters.

length of each ray 15-30 4; without
centrum.

Small sphaerasters.

total diameter 8 pu.

6-20 rays; 22-43 » in total diameter;
without centrum.

total diameter 4.5—9 yt.

Sterrasters.

110 p long ; 90 » broad.

116-130 »# long; 97-105 » broad;

70-90 p thick.

24 SIDONOPS ANGULATA.

Sidonops angulata, sp. nov.
megana, var. nov.
Plate 12, figs. 1-8, 16, 17, 19, 20; Plate 13, figs. 1-12, 22-25; Plate 14, figs. 1-6, 16-22; Plate 15,
figs. 1-4, 7-9, 11.
microana, var. nov.

Plate 12, figs. 11-15, 18, 21, 22; Plate 13, figs. 13-17, 21; Plate 14, figs. 7-9; Plate 15, fig. 10.

orthotriaena, var. nov.

Plate 12, figs. 9, 10; Plate 13, figs. 18-20; Plate 14, figs. 10-15, 23-30; Plate 15, figs. 5, 6, 12.

I establish this species for four specimens obtained at three different sta-
tions off the coast of southern California, in the vicinity of Santa Barbara Island.
Some of the amphioxes and also a few of the rhabdomes and clades of the telo-
clades are angularly bent, and to this character the specific name refers. Two
specimens from Station 2975 are identical. The other two differ from these and
from each other so much that it is necessary to recognize three varieties. In
the specimen from Station 2945, var. orthotriaena, the subcortical triaenes are
orthoclade, in the three others plagioclade. In the two specimens from Station
2975, var. megana, some of the anaclade-cladomes are large, while in the speci-
men from Station 4417, var. microana, all the anaclade-cladomes are small.

Shape and size. One of the specimens of var. megana is more lobose, the
other more massive. The lobose specimen (Plate 12, fig. 19) has the shape of a
stout fan, 86 mm. broad, 75 mm. high, and 28-38 mm. thick. Rounded pro-
tuberances rise from its surface and give to the margin of the fan a somewhat
serrated appearance. On one side these protuberances attain a greater height
than on the other, and here the depressions between them in one place join,
leaving a part of the sponge, 11 mm. thick, suspended like a bridge between
them. The surface is rough, shagreened. The greater part of this roughness
is due to the presence of slight pit-like depressions which are about 1 mm. wide
and are close together. Apart from a few holes, about 1 mm. wide, which do
not seem to be oscules, no apertures visible to the unassisted eye occur. In a
few sheltered places remnants of a spicule-fur are observed. The massive speci-
men of this variety (Plate 12, fig. 20) has the shape of an inverted cone with a
strongly rounded margin. It is 77 mm. high. The largest and smallest trans-
verse horizontal diameters are 112 and 107 mm. respectively. There are a few
broad and low protuberances, chiefly on the margin of the upper, somewhat
concave face. From the base a digitate process, 25 mm. long and up to 14 mm.
thick, arises. The most exposed parts of the surface are smooth; the rest of it

SIDONOPS ANGULATA. 25

a

is rough and shagreened. On the sides, this roughness is due to the presence
of shallow pits, separated by a network of ridges somewhat raised at their junc-
tions; on the upper face it is due to numerous wart-like protuberances varying
in size and about 1 mm. apart. There are no larger oscules. On some of the
more protected parts of the surface a spicule-fur, up to 5 mm. high, is observed.
The specimen of var. microana (Plate 12, fig. 18) is irregularly spherical, 49 mm.
broad and 40 mm. high. The lower part of the body is somewhat drawn out to
form a peduncle, 30 mm. broad and 15 mm. thick, whichis attached to a coral.
The upper side is flattened. The greater part of the surface is covered by a
dense spicule-fur which is in places 5.5 mm. high (Plate 13, fig. 21b). The
specimen of var. orthotriaena is a fragment of an irregular lobose mass. It
measures 33 mm. in length, 23 mm. in breadth, and 14 mm. in thickness. The
surface is quite smooth and without larger oscules. In places there are rem-
nants of a spicule-fur. A few insignificant symbionts, chiefly small crusts of
calcareous and monoaxonid silicious sponges, are attached to the surface of all
the specimens. The specimen of var. microana bears a dense growth of diatoms
on its surface.

The colour, in spirit, is yellowish in the interior and white to reddish or
purplish brown on the surface. The lobose specimen of var. megana is quite
white on one side and has a reddish brown tinge on the other. The massive
specimen of this variety is partly yellowish, partly reddish white on the sides
(below) and reddish brown on the upper face (above). This colour is not uni-
form, some parts of the upper face being considerably darker than others.
Variety microana is dirty white below and purplish brown above, var. ortho-
triaena brownish white.

The superficial part of the body is differentiated to form a cortex composed
of a thin outer dermal layer (Plate 14, fig. 20a), a central sterraster-armour
layer (Plate 13, fig. 21a, 25a) 0.7-1 mm. thick, and an inner fibrous layer (Plate
14, fig. 22a) excavated by subcortical cavities.

Numerous granular cells, extended paratangentially and measuring 12-18
by about 7 », lie in the dermal membrane just below the surface (Plate 14,
fig. 20b). Below these, between them and the most distal sterrasters of the
armour, slender fibres extend paratangentially. Those adjacent to the pores
are circularly bent and surround the pores sphincter-fashion. These fibres are
strongly stained with haematoxylin but only slightly with azure. In the choano-
some of the massive specimen of var. megana numerous irregularly polyedric
spaces, 40-70 # in diameter, occupied by dense masses of small cells (Plate 15,

26 SIDONOPS ANGULATA.

fix. 4a), were observed. Each of these cells contains a small, strongly staining
nucleus. Perhaps these cells are young spermatozoa.

Canal-system. The sides of the massive specimen of var. megana are coy-
ered with afferent pore-sieves (Plate 13, figs. 23, 24), which coalesce to form
extensive, nearly continuous poral tracts. The pores are oval or, more rarely,
cireular, and generally measure 25-300 « in diameter. The strands of dermal
tissue separating them are as broad or broader than the pores themselves.
These pores lead into elongate subdermal cavities, from five to seven of which
join to form stellate groups 0.4-1 mm. in diameter. These groups of radiating
subdermal cavities are sunk in the sterraster-armour layer and the spaces be-
tween them in great part occupied by sterrasters. In consequence of this the
stellate cavity groups are very conspicuous in superficial paratangential sections
of appropriate thickness (Plate 13, fig. 23). The cavities of each group converge
to a common centre and here they join to form a radial cortical canal which
penetrates the cortex and leads down into the interior. These radial cortical
‘canals are 1-1.4 mm. apart, and in the sections examined are strongly con-
tracted, usually quite closed. They are surrounded by mantles of circular
fibres. These mantles increase in thickness proximally and form chones (Plate
13, fig. 25b), which protrude into the subcortical cavities. Around these chones
I have often noticed extensive dome-shaped excavations of the proximal (inner)
surface of the sterraster-armour. When such excavations are present the
chones hang down as it were from the apices of the domes and are thus situated
some distance above the general lower limit of the sterraster-armour layer.
Below this layer, in the inner zone of the cortex, a system of subcortical cavities
(Plate 13, fig. 25c, Plate 14, fig. 22b) extends. From these cavities the afferent
choanosomal canals take their rise. These canals, and also the choanosomal
efferents, are narrow, the choanosome appearing very solid in consequence.
The flagellate chambers (Plate 14, fig. 21a) are spherical or oval and small.
The efferent choanosomal canals lead up to and open out into extensive systems
of subcortical cavities which underlie the parts of the cortex bearing the uniporal,
efferent apertures. From these cavities radial cortical canals with chones,
similar to the afferent ones described above, arise. On the walls of some of the
efferent cortical canals of the lobose specimen of var. megana I observed a few
large, broad, and blunt conic spines (Plate 16, fig. 7a) which protrude into the
canal-lumen. As these structures are rare and as I failed to find them in thin
sections, where they could have been studied with higher powers, I was unable
to ascertain their nature. Each efferent cortical canal leads up to a single

SIDONOPS ANGULATA. 27

circular pore up to 100 » wide (Plate 13, fig. 22). In the massive specimen of
var. megana and also in var. microana the efferent pores occupy the upper
depressed or flattened side of the sponge. In the other two specimens they
appear to be distributed less regularly. As stated above the upper face of the
massive specimen of var. megana is covered with wart-like protuberances, une-
qual in size and on the average about 1 mm. apart. Many of these warts bear
on their summit an efferent pore, many however are without an apical aperture.
I presume that all these warts are pore-bearing elevations and that on those on
which no pore was seen, the pore had been quite closed by excessive contraction
of the sphincter surrounding it. In the specimen of var. microana, on the other
hand, nearly all the efferent pores seem to be open. In this sponge many of
them lie on the level of the surface and are not raised above it.

Skeleton. Spicule-bundles, extending radially and abutting vertically on
the surface, traverse the choanosome. In the interior these bundles are chiefly
composed of quite stout amphioxes, to which a few thick styles or branched
style-derivates may be added. Towards the surface also plagiotriaenes (in
var. megana and var. microana) or orthotriaenes (in var. orthotriaena), anaclades,
and long and slender amphioxes (much more numerous in var. microana than in
the other varieties) are added to the stout amphioxes (and styles and style-
derivates). In var. microana and var. orthotriaena the cladomes of the plagio- or
ortho-triaenes lie on the level of the lower limit of the sterraster-armour layer.
In the two specimens of var. megana they are situated a little higher up, within
this layer, and entirely enveloped in sterrasters. The anaclades are not numer-
ous. Most of them are anatriaenes. In var. microana also anadiaenes are
met with. The cladomes of some of the anaclades lie in the inner zone of the
cortex; by far the greater number, however, protrude freely beyond the surface.
Some long and slender amphioxes lie altogether within the sponge; numerous
spicules of this kind protrude beyond it. Together with the anaclades they
form the spicule-fur. In it the slender amphioxes are much more numerous
than the anaclades. Some of both the stout choanosomal and the slender
chiefly dermal amphioxes, and a few of the teloclade-rhabdomes and clades are
angularly bent.

In the spicule-preparations of var. microana two or three dichotriaenes and
some mesoprotriaenes were observed. Since however I failed to find such
spicules in situ in the sections, I do not believe that they belong to the sponge.

The microscleres are smooth oxyasters and oxysphaerasters, spined strongy-
losphaerasters and sterrasters. The oxyasters are confined to the choanosome

28 SIDONOPS ANGULATA.

and are not numerous. The oxysphaerasters are very abundant, particularly
in the inner layer of the cortex, where they line the walls of the subcortical
cavities in large numbers (Plate 14, fig. 22c). These asters also occur in the
walls of the chonal canals and they extend distally along the walls of the dermal
cavities, right up to within a short distance of the outer surface. The strongylo-
sphaerasters occupy the dermal membrane (Plate 13, figs. 22-24; Plate 14,
fig. 20a) in a single or in several layers and are met with also in the walls of the
cortical canals and subcortical cavities. On the whole the strongylosphaerasters
situated superficially appear to be larger than those in the interior of the cortex.
The sterrasters form dense masses in the middle armour layer of the cortex and
occur scattered also in the choanosome, where they are particularly abundant
in the lobose specimen of var. megana and in var. microana.

Besides these microscleres, I found, in the centrifugal spicule-preparations
of var. microana, two oxyasters with more slender, spined rays, and in those of
the lobose specimen of var. megana numerous minute rhabds. Both these kinds
of spicules I consider as foreign.

The stout choanosomal amphiozes (Plate 12, fig. 17c; Plate 13, figs. 1, 2, 17a,
b, 20b) are fairly isoactine and quite sharply pointed. They are usually straight
(Plate 13, fig. 2, 17b) or slightly curved (Plate 13, fig. 1), more rarely angularly
bent (Plate 13, fig. 17a) near the middle. The angular bend amounts to 8-12°,
so that the two actines of these angular amphioxes enclose angles of 168-172°.
These angularly bent choanosomal amphioxes are more frequent in the lobose
specimen of var. megana and in var. microana than in the other two specimens.

The stout choanosomal amphioxes are 1.6-3.7 mm. long and 20-72 4# thick.
Their thickness is on the whole proportional to their length. They are con-
siderably longer in var. megana than in the other two varieties.

DIMENSIONS OF STOUT CHOANOSOMAL AMPHIOXES OF SIDONOPS ANGULATA.

Length, mm. Maximum thickness, 2
Var. megana, lobose specimen 2-3.7 40-60
Var. megana, massive specimen 2.4-3.5 53-72
Var. microana 1.8-2.8 20-52
Var. orthotriaena 1.6-2.5 20-70

The slender dermal amphiozes (Plate 12, figs. 16a, b, 17a; Plate 13, fig. 21b)
are usually simply curved (Plate 12, figs. 16a, 17a), rarely angularly bent (Plate

SIDONOPS ANGULATA. 29

12, fig. 16b). In the angularly bent spicules the angular bend amounts to 7—
12°, and is usually, as in the spicule represented in Plate 12, fig. 16b, not in the
middle, but considerably nearer one end than the other, the two straight parts
of the spicule thus enclosing an angle of 168-173° and being unequal in length.
The slender dermal amphioxes are fairly isoactine, thickest in the middle,
gradually attenuated towards the ends, and terminally abruptly and sharply
pointed. They are 2.9-9.5 mm. long and 5-34 » thick. Those of var. ortho-
triaena are much smaller than those of the others.

DIMENSIONS OF SLENDER DERMAL AMPHIOXES OF SIDONOPS ANGULATA.

Length, mm. Maximum thickness, p
Var. megana, lobose specimen 4.9-7.1 18-34
Var. megana, massive specimen 4.2-9.5 25-32
Var. microana 5.5-7 10-22
Var. orthotriaena 2.9-4.5 5-17

Styles and style-derivates (Plate 13, fig. 20a) were not found in the massive
specimen of var. megana at all and in the three other specimens they are very
rare. The regular styles are straight, conic, sharp pointed at one end, and
considerably thickened and club shaped at the other. They are 2.1-2.5 mm. long
and 60-100 « thick at the rounded end. In var. microana I found a style-derivate
2.3 mm. long and 110 » thick at the rounded end, which bore, 400 » below the
rounded end, a terminally rounded branch-ray, 120 » long, directed obliquely
towards the rounded end of the main shaft.

The plagio- and ortho-triaenes (Plate 12, figs. 16d, 17d, 21, 22; Plate 13,
figs. 3-16, 18, 19) have a straight or slightly curved rhabdome, which is usually
on the whole conical, and attenuated towards the pointed acladomal end, more
rapidly in the acladomal terminal part than in the cladomal and central parts.
These normal plagio- and ortho-triaene-rhabdomes are 1.5-2.8 mm. long and
47-82 » thick at the cladomal end. In var. orthotriaena I found two abnormal
orthotriaenes, one with a cylindrical, terminally rounded, and considerably
shortened and thickened rhabdome only 550 » long but 105 » thick, and another
with an angularly bent rhabdome (Plate 13, fig. 19). The angle enclosed
between the cladomal and acladomal parts of the latter is 109°. The clades
are conic and usually pointed, rarely (Plate 13, fig. 8) rounded at the end.
They are 330-700 » long, those of the plagiotriaenes of var. megana being a

30 SIDONOPS ANGULATA.

eood deal longer than those of the plagio- and ortho-triaenes of the other two
varieties. The clades of the same cladome are equal (Plate 18, figs. 9, 11, 15, 18),
or slightly (Plate 13, figs. 5, 6, 8, 10, 13, 14, 19) or considerably (Plate 13, figs.
7, 12, 16) unequal in length. Generally the clades extend in a longitudinal plane
passing through the axis of the rhabdome; sometimes, however, (Plate 13, fig. 4)
an angular bend in a transverse plane is observed. Short clades are often
(Plate 13, fig. 7, 12), long ones rarely (Plate 13, figs. 9, 12), nearly straight.
Generally the clades are markedly curved, concave to the rhabdome. This
curvature increases towards the ends of the clades. In some (Plate 13, figs. 3,
13-16, 19) the degree of this increment of curvature is slight, in others (Plate 13,
figs. 5-8) it is considerable. Sometimes the distal part of the clade is bent down
abruptly (angularly) (Plate 13, figs. 5, 10). The angles enclosed between the
clade-chords and the axis of the rhabdome are in the two specimens of var.
megana on an average 101.3 and 103°, in var. microana 105.6°, and in var.
orthotriaena 93.7°. These spicules are accordingly in the first two plagio-
triaenes, in the last orthotriaenes. The clade-angles of the three clades of the
same rhabdome are usually about equal, rarely (Plate 13, fig. 6) considerably
different.

Very rarely one of the clades becomes quite rudimentary. Such diaene

spicules were found only in var. orthotriaena.

DIMENSIONS OF NORMAL PLAGIO- AND ORTHOTRIAENES OF SIDONOPS ANGULATA.

Thickness an Deke Cate
WERETHOF 0 Léneth OF and axis of rhabdome
phandame aap dome ARES an s i
cladome
mm. » B limits average
Var. megana, lobose specimen 1.82.8 50-82 350-700 92-104 101.3
Var. megana, massive specimen 1.6-2.6 50-80 330-650 89-111 103
Var. microana 1.5-2.6 50-77 380-580 91-112 105.6
Var. orthotriaena 2-2 ..35 47-78 380-500 91-98 93.7

The rhabdomes of the anaclades (Plate 12, figs. 1-15; Plate 13, fig. 17¢) _
are for a great part of their length nearly cylindrical. Of the long ones none
were found intact in the spicule-preparations, all being broken. The longest
fragment measured was 9 mm., the longest rhabdome observed intact was 6 mm.
in length. The rhabdomes of the anaclades of var. microana appear to be

SIDONOPS ANGULATA. 31

considerably shorter than those of the anaclades of the other two varieties.
The acladomal end of the rhabdome is attenuated and pointed, or cylindrical
and terminally rounded (Plate 12, fig. 15). At the cladomal end the rhabdomes
are in var. megana 7-39 y, in the two other varieties 10-18 » thick. There is
always on the summit of the cladome an apical protuberance which, however,
does not contain a prolongation of the axial thread of the rhabdome, the latter
terminating at the point where the axial threads of the clades arise from it.
The axial threads of the clades are directed obliquely downwards in their basal
portion (Plate 12, figs. 1, 5, 13, 14). After extending a short distance in this
direction they bend outward angularly and then follow the axes of the clades.
This peculiarity of the axial threads of the clades is doubtlessly the cause of
the formation of the apical protuberance of the cladome. Most of the anaclades
have three fairly equal clades (Plate 12, figs. 1-6, 8, 9, 12-14). In some there
are two longer and one short clade (Plate 12, fig. 7). In not a few of the
anaclades of var. microana one clade has disappeared entirely, so that these
spicules are diaenes (Plate 12, figs. 10, 11). The clades are conical, pointed, and
when long, distinctly curved, concave towards the rhabdome (Plate 12, figs. 1,
3, 5-10). In short clades (Plate 12, figs. 2, 4, 11-14) this curvature is usually
slighter, often hardly perceptible. The distal part of the clade is usually curved
less than the basal. Sometimes (Plate 12, figs. 3, 5,6) an abrupt angular bend
is observed where the curved proximal part passes into the more straight distal
part. The chords of the clades are in the anaclades of var. megana 45-210 p,
in those of var. orthotriaena 33-80 y, and in those of var. microana only 30-50 yu
long. The angles between the clade-chords and the axis of the rhabdome are
27-66°, on an average 47°.

DIMENSIONS OF CLADES OF SIDONOPS ANGULATA.'

4 Angle besweeu pletion and
mhabdome at the qhength of axis 0 me ome
B BB limits average
Var. megana, lobose specimen 11-39 45-210 35-65 47
Var. megana, massive specimen 7=35 48-142 27-66 45
Var. microana 10-18 30-50 50-62 47
Var. orthotriaena 14-18 i 33-80 43-54 48

'Rhabdome length not known.

o2 SIDONOPS ANGULATA.

The various oryasters and oxysphaerasters (Plate 14, figs. 1-15a, b, e, 16,
23, 24) form a continuous series. They have from one to twenty-three rays.
Forms with more than thirteen rays are much more frequent than forms with
fewer, and among the latter those with from one to three rays much scarcer than
those with four or more. These asters form a fairly continuous series ranging
from large oxyasters without centrum and few rays (Plate 14, fig. 24) to small
oxysphaerasters with large centrum and numerous rays (Plate 14, figs. 16, 28).
The rays are straight, usually on the whole conical, attenuated towards the sharp-
pointed end in the basal and middle parts more gradually than in the terminal
part, and perfectly smooth (Plate 14, figs. 16, 23, 24). Their shape varies in
correlation to their size, the longest rays being the most slender and the shortest
the stoutest. Sometimes, chiefly in the two- to four-rayed oxyasters (Plate 14,
figs. 3b, 13b), short, rounded or truncate rudiments of reduced rays occur in
addition to the properly developed conical and pointed ones. These ray-
rudiments are generally smooth, rounded knobs not longer than broad (Plate
14, fig. 3b), more rarely longer and in this case sometimes crowned with a few
terminal spines. The rays are always concentric. Their arrangement is in
the few-rayed oxyasters often irregular, in the many-rayed’ ones nearly always
regular. In none of the two-rayed forms observed were the two rays regu-
larly arranged, that is, situated exactly opposite each other in a straight line,
the angle enclosed by them being always considerably less than 180°. In
several of these asters this angle was under 120° and in one even under 90°.
These diactine asters consequently look like more or less opened compasses.
Also in the three-rayed oxyasters (Plate 14, figs. 3b, 13b, 24) irregular ray-
arrangement is the rule. In the oxyasters with four (Plate 14, fig. 1b) or more
(Plate 14, figs. 4b, 16, 23) properly developed pointed rays, on the other hand,
the rays are usually regularly arranged.

The oxyasters and oxysphaerasters are 11-64 # in total diameter, the cen-
trum attaining a maximum diameter of 12 4. The normal conically pointed
rays are 2.5-40 » long, and at the base 1.6-5 « thick. Roughly speaking, the
size of the rays and of the whole aster is in inverse proportion to the number
of rays. Only the monactine oxyasters appear as an exception to this rule.
Since however these spicules are rare and I was able to measure but few, I
do not attach much importance to this fact. The one- to five-rayed oxyasters
are 25-64 » in diameter, their normal conical rays measuring 13-40 by 1.7-5 4.
The six- to ten-rayed oxyasters are 23-44 » in diameter, their rays measuring
8.5-25 by 1.6-4.5 ». The eleven- to twenty-three-rayed oxyasters are 11-25 p
in diameter, their rays measuring 2.5-13 by 1.7-3.3 p.

SIDONOPS ANGULATA. 33

The development (size) of the centrum is, roughly speaking, in true propor-
tion to the number of rays and in inverse proportion to the size of the aster.
In some of the large two- to five-rayed oxyasters the knob-like ray-rudiments
clustering round the centre form an irregular thickened mass, but none of these
asters have a true centrum. The largest oxyaster with such a centrum observed
was 38 « in diameter. Most of the six- and seven-rayed oxyasters are also
without centrum, but among the oxyasters with eight to ten rays a great many
are provided with one, and in the oxyasters with eleven or more rays a spherical
central thickening is invariably present. In the larger oxyasters with centrum
the diameter of the latter is always much less than the ray-length, while in the
smaliest oxyasters (oxysphaerasters), which also possess the greatest number
of rays, the diameter of the centrum considerably exceeds the ray-length.

The oxyasters and oxysphaerasters of the four specimens are very similar.
The differences observed, which are recorded in the appended table (page 34),
are well within the limits of the accidental inaccuracies due to the smallness
of the number (only about one hundred) of oxyasters measured.

The strongylosphaerasters (Plate 14, figs. 1d, 2d, 5d, 7d, 9d, 10d, 17-19,
25-30) usually have a spherical centrum and about ten to twenty radial rays.
In most of the strongylosphaerasters all the rays are about equal in size (Plate
14, figs. 17, 25-28). In not a few strongylosphaerasters of var. megana and var.
orthotriaena however, some of the rays are reduced to insignificant protuberances
of the surface of the centrum and are much shorter than the others. On Plate 14
two strongylosphaerasters of this kind are represented, one (Figs. 18, 19) with
all but three, the other (Figs. 29, 30) with all but one ray thus reduced. The
properly developed rays are cylindrical or cylindroconical and truncate, very
rarely conical and pointed. In var. microana they are on the whole more
slender, longer, and distally more attenuated than in the other two varieties, and
the strongylosphaerasters with pointed, conical rays have been observed only
in this variety. The rays are 1-12 » long and at the base 2-6 » thick, the
dimensions of the fully developed ones being in inverse proportion to their
number. Strongylosphaerasters with only one fully developed ray are rare and
have been found only in var. orthotriaena. In these the single ray is 12 » long
and 6 «thick. In the strongylosphaerasters with two or three fully developed
rays, which are quite frequently met with in both specimens of var. megana and
in the specimen of var. orthotriaena, these rays are 8 » long and 44.5 thick.
In the strongylosphaerasters with from four to nine fully developed rays which
are still more abundant in var. megana and var. orthotriaena, these rays are

SIDONOPS ANGULATA.

34

Ste 8 rae &@ (Sits 9 6 61 z 8 b 6 € 8 ral €G ra 9 6 LT ez-FI
Lill eG ie =g S111 ne S15 2 el = -9 —c1 =i ||| eaiOull ea. jul Sant =S -9 salt
ts €1 OT KG LG €1 OL GG € Or G8 | GE GG L 6 ix eL-It
=Galiend) le eo) |p =e | -0r) =9 9b Or | Sieh) eae |) 3/0) ean | Ree 8%
F LY OI- ins ¥ LT OF FE € 9T 9- €€ F LT {i té 9S tI on L1G or-8
-g | -@ | euoa| -ez || 91 | -6 | euou| -ez || -e:z | -6 | suc) -7e || -% | -OL | cucu) ce || —@ | —IT | suou | —9z
oT GG Sa lag & SG" | onou OF z 61 aa ee oT &% jim tPF P 0@ 8- 8& 1-9
=L'T | =G'g | suo) —9¢ Jil | aali —G& = —G'g | su0u | —9Z —% |-G PI | euou | —66
g | 8 |e | seo |e) OB econ) 22) 8 We2e leorous net iis Oe al 2 ae Oa ake cca fae oe
—L1 |-G'ZT | euou | —¢G-9z || —Z'T | —8T GG alt= GiG|emOG —0F —8'% |—S-ZI | uo | —G'9G <% —G6 +P
GF OF g-¢- 9 oF OF 09 F 1€ ¢ Le 86 (ty cS r
| 9uou euou 9uou al
% —EI | euou | —cZ “GG | Sz —OF =GiGii| mG —GE =G —€1I | euou | —Se a ne fe eae
cy | 22 | g9| ee |
5 z ; = o:
-ce | -cz | -¢e | -2z | cs L1G cP €& oT GG co 1X6 IT
2 | ae | 8 eS 2 | 38 | 8 = & | sR | 8 eS @ | Se | 8 oe) et a See | Se &
= Ee a a = 5 > g a + eS g a — Fis | 2 ei) es 5 e g a
5 Sie) 5 s z Ze) sg = = Ze g = ZO Ze) sg |e ae) S eel
Bn elt ee | ee ee | eon ea Nhe eee cele alee wlll ite eal gee 1 eee alee le
B ee | ¢ 3 ® ed | « a 3 oz 3 = ® e383 | 6 s || B a 8 5 sXe
D 2a n & ma oe ess & Bid g a & eee) 3B ee ||| (eco ast g a E
o =e gq :— ey 5 emily oS ° So 3 = o See) Sh leeks Il ea Ss; = || padojaaap Aqrng
4 par E 4 bac ie be 5 a $2] § SW ey ||
é Be | = & Bisel & z=) = & Siem mes Wc teoms| tee ce as!
g Ss ~ a S3 |r | a SF .~ | 3 55 | = een Some as
aS Reo = Rs RS | w= {| eae ersecoy
| | |
| |
eye[nsur sdouopig Buse IOYWO “IRA | BUROIOIUL “IBA luaurtoads aarssemr ‘wuRsaut “IBA || Uatttoads asoqoy ‘vuLsaUr “IBA |)

VLIVTINONVY SdONOGIS

AO

SUALSVUAVHdSAXO ANV SUALSVAXO AO SNOISNAWIG

SIDONOPS ANGULATA. 35

5.5-6 » long and 4 thick. In the ordinary strongylosphaerasters with from
ten to twenty equal rays, which occur in great numbers in all the specimens,
the rays are 2-4 u thick, in var. microana 3-9 y, and in the other two varieties
1-7 long.

The size of the centrum and of the whole aster is, like the size of the rays,
in inverse proportion to the number of the number of the latter. In the strongy-
losphaerasters of var. megana and var. orthotriaena with from one to nine fully
developed rays, the centrum is 11-13 » and the whole aster 21—26 » in diameter.
In the strongylosphaerasters with from ten to twenty equal rays these dimensions
are 7-14 » and 14-24.5 » respectively in all the three varieties.

The distal parts of the fully developed rays bear numerous spines while
their basal part and the centrum are smooth. The spines are conic, usually
0.5-1 » long, rarely smaller, and not recurved. Those situated on the terminal
face of the ray appear to radiate from a common centre within the tip of the ray
and diverge accordingly; those on the sides of the ray are slightly oblique,
inclined towards the end of the ray. When the ray is reduced in length the
spines on its terminal face retain their full size. Consequently the low protuber-
ances representing greatly reduced, rudimentary rays of this kind, are covered
with tufts of spines (Plate 14, figs. 18, 19, 29, 30).

In the rays of some of the strongylosphaerasters of var. orthotriaena I ob-
served thick axial threads. These were joined in a regularly concentric manner
in the centre of the spicule and extended in straight lines along the axes of the
rays to within a short distance of their ends, where they appeared to terminate
with slight irregular thickenings. Occasionally it seemed that exceedingly
fine branches extending towards the spines arose from the distal parts of these
axial threads. These being near the limit of microscopic visibility, it is doubt-
ful whether such structures really exist, or whether the impression of them was
merely an optical illusion.

Besides the strongylosphaerasters described above I found in the centrifugal
spicule-preparations of the massive specimen of var. megana some small ones,
7.5-14 » in total diameter with a centrum measuring only 1.5-2.5 », and seven-
teen to nineteen minutely spined rays 0.5-1 » thick. These asters appear to be
young stages of the ordinary strongylosphaerasters.

The sterrasters (Plate 13, figs. 22-25; Plate 15, figs. 1-3, 5-12) are more or
less regular flattened ellipsoids. When seen from above, with the umbilicus
in the centre of the upper side, their contour generally appears as a regular

ellipse, sometimes nearly approaching a circle (Plate 15, figs. 9-11). In var.

36 SIDONOPS ANGULATA.
DIMENSIONS OF STRONGYLOSPHAERASTERS OF SIDONOPS ANGULATA.
Number of fully developed rays 1-3 4-9 10-20
var. megana, lobose specimen 24 23.5 14-22.5
Total diameter of | var. megana, massive specimen 26 2 16-24.5
asters yt var. microana 19-24
var. orthotriaena 24 17-22
var. megana, lobose specimen 12 8-14
Diameter of cen-| var. megana, massive specimen 12 7-13.7
trum 4 var. microana 7-13.5
var. orthotriaena 11 8.5-12
var. megana, lobose specimen 8 6 2-7
Length of rays = 5 ; S
: re var. megana, Massive specimen 8 5.5 1-6.5
(without the cen- :
Gey var, microana 3-9
P var. orthotriaena 12 3.5-5.5
var. megana, lobose specimen 1 4 2-4
Basal thickness of | var megana, massive specimen 4.5 4 2-4
rays ft var. microana 2-4
var. orthotriaena 6 2-4

orthotriaena, however, a good many sterrasters have, when seen in this position,
asomewhat rhomboidically distorted outline (Plate 15, fig. 12). The sterrasters
are 85-122 » long, 75-113 » broad, and 57-86 » thick. They are largest in

var. megana, smaller in var. orthotriaena, and still smaller in var. microana.

DIMENSIONS OF STERRASTERS OF SIDONOPS ANGULATA.

Length Breadth Thickness
He a ft
Var. megana, lobose specimen 105-122 90-105 73-86
a megana, massive specimen 110-120 100-114 70-83
Var. microana 85-97 75-90 57-65
Var. orthotriaena 90-111 80-94 65-79

The average proportion of length to breadth to thickness is in the sterrasters
of the lobose specimen of var. megana 100 :87 :68, in those of the massive
specimen of var. megana 100 :92 :63, in those of var. microana 100 :89 : 66,

and in those of var. orthotriaena 100 :88:73.

In a thin radial splinter of a very young sterraster of the massive specimen
of var. megana with long, slender, and pointed ray-ends, which lay opportunely

SIDONOPS ANGULATA. 37

«

for examination with high powers in a centrifugal spicule-preparation, I clearly
saw that a radial axial thread, extending right up to its end, is contained in each
ray. In the centre of a young sterraster of var. orthotriaena a little cluster of
a few very small granules, lying close together, was observed. In the centre
of a slightly heated, adult sterraster of the massive specimen of var. megana I
observed an apparently solid black sphere, 15 « in diameter, from which black
rays radiated to some distance. Such a blackening has been observed several
times. It seems to show that the central part of the spicule contains more
organic substance than the superficial part, and that the axial threads of the
rays are distally silicified to a greater extent than proximally.

The distal, freely protruding parts of the rays are in normal sterrasters
everywhere, except in the vicinity of the umbilicus, 4-5 « thick and provided
with a terminal verticil of usually four to six stout, blunt, and often somewhat
curved, lateral spines (Plate 15, figs. 5, 6). The distal ends of the rays sur-
rounding the umbilicus have a transverse section, elongated in a direction
radial to the umbilicus, usually 4-5 « broad and 6-7 » long. They are generally
provided with from seven to nine lateral spines and also bear several spines on
their terminal face (Plate 15, figs. 1-3). The spines of these rays, which are
directed towards the centre of the umbilicus, are a little larger than the others
and often curved.

Besides these normal sterrasters some abnormal ones, for which I propose
the term sterroids, were observed, chiefly in var. orthotriaena and the massive
specimen of var. megana. The most frequent kinds of abnormities met with
are sterrasters in which the distal ray-ends are thicker, as much as 6-9 » in
transverse diameter, farther apart, and provided with a greater number of
spines than in the normal sterrasters. In some of these sterrasters single scat-
tered spines, similar to those forming the verticils on the rays, arise here and
there between the protruding distal ray-ends directly from the surface of the
solid centrum of the spicule. Much more rarely strongylosphaeraster-like sterr-
asters with relatively long, terminally rounded, protruding rays were observed.
In some of these the protruding ray-ends were smooth, in others densely
covered with small spines.

The two specimens of var. megana were trawled at Station 2975 on February
12, 1889, in 34° 1’ 30” N., 119° 29’ W., depth 66 m. (36 f.); they grew on a bottom
of gravel and broken shells; the bottom temperature was 73.9° (57° F.). The
specimen of var. microana was caught with the tangles at Station 4417 on April
12, 1904, near Santa Barbara Islands, S. W. rock Santa Barbara Island, N. 8° W.,

38 SIDONOPS ANGULATA.

11.7 km. (6.3 miles), drift S. 73° W.; depth 53 m. (29 f.); it grew on a bottom
of fine yellow sand and coralline rock. The specimen of var. orthotriaena was
trawled at Station 2945 on February 6, 1889, in 34° N., 119° 29’ 30” W., depth
55 m. (30 f.); it grew on a pebbly bottom.

The similarity of the rather peculiar smooth oxyasters and oxysphaerasters
and the smallness of the differences of most of the other skeletal elements and
the soft parts of these sponges show that they are nearly related to one another,
a conclusion which is corroborated by the fact that they all come from the same
region. The differences between the two specimens from Station 2975 are so
slight that I do not hesitate to place them in the same systematic unit. Several
of the differences between these and the other two and between the latter are,
on the other hand, considerable. Some of these differences, as for instance the
much smaller size of both kinds of amphioxes in the smaller specimen from Sta-
tion 2945, may be due merely to differences of age or growth and are therefore
systematically unimportant; other differences appear to be of greater signifi-
cance, and of these the following may be noted: the, subcortical triaenes of
the specimen from Station 2945 are orthotriaenes, while those of the others are
plagiotriaenes; the strongylosphaerasters of the specimen from Station 4417
have more slender and conical rays, those of the other specimens stouter and
more cylindrical ones. Many of the sterrasters of the specimen from Station
2945 are rhomboidically distorted, while all or nearly all the sterrasters of the
others have regularly elliptical contours. The sterrasters of the specimen from
Station 2945 are slightly, those of the specimen from Station 4417 very con-
siderably, smaller than those of the specimens from Station 2975. The anaclades
of the specimen from Station 4417 are all small; in the specimen from Station
2945 medium sized, and in the specimens from Station 2975 large anaclades
occur besides the small ones. In the specimen from Station 4417 the anaclades
are partly triaene and partly diaene, in the two latter all the anaclades observed
were triaene. There can, I think, be no doubt about these differences being due
to congenital particularities and not to mere individual (somatic) adaptations
or differences of chromatin-separation or mixture before and during fertiliza-
tion. For this qualitative reason, and also for the quantitative reason that
these differences are by no means inconsiderable in extent, I think that they
must find systematic expression. As these variations are due mostly to pecu-
liarities of parts lying either, like the strongylasters and sterrasters, close to
the surface, or, like the anaclades, even protruding beyond it; as structures
thus directly exposed to the influence of external forces are, a priori, liable to be

SIDONOPS ANGULATA.

39

somewhat different even in most closely related individuals; and finally as

the difference in the subcortical triaenes, which is the only valid difference of the

less exposed internal parts, is, as a comparison of the cladomes of these triaenes

(Plate 26) shows, not very great; it appears advisable to place these four

sponges in one and the same species, with three varieties within this species.

DIFFERENCES BETWEEN THE VARIETIES OF SIDONOPS ANGULATA.

Colour

Var. megana

Var. microana

Var. orthotriaena

white, yellowish, reddish
brown.

dirty white, purple-brown.

brownish white.

Stout, choanoso-
mal amphioxes

2-3.7 mm. long, 40-72
thick.

1.8-2.8 mm. long, 20-52
thick.

1.6—2.5 mm. long, 20-70
thick.

Slender, dermal
amphioxes

not numerous; 4.2-9.5
mm. long, 18-34 y thick.

very abundant; 5.5-7 mm.

long, 10-22 y» thick.

not numerous; 2.9-4.5
mm. long, 5-17 » thick.

Styles (style-deri-
vates)

2.1-2.5 mm. long, 60-100
yw thick (not found in the
massive specimen).

as in the lobose specimen
of var. megana.

as in the lobose specimen
of var. megana.

plagiotriaenes. plagiotriaenes. orthotriaenes.
trhabdome 1.6-2.8 mm. | rhabdomes 1.5-2.6 mm. |rthabdome 2-2.35 mm.
Plagio- or ortho-| long, 50-82 , thick;|long, 50-77 » thick; | long, 47-78 thick; clades
triaenes clades 330-700 » long; | clades 380-580 » long; | 380-500 y long; clade-
clade-angle 92-104°, av- | clade-angle 91-112°, ay- angle 91-98°, average
erage 102.15°. erage 105.6°. 93.7°.
anatriaenes. anatriaenes and anadi- | anatriaenes.
thabdome long, 7-39 | aenes. rhabdome long, 14-18
Pe rolades thick; clades 45-210 4 thabdome shorter, 10—| thick; clades 33-80 4
long; clade-angle 27-66°, | 18» thick; clades 30-50 » | long; clade-angle 43-54°,
average 46°. long; clade-angle 50-62°, | average 48°.
average 47°.
Oxyasters and oxy-| 1-23 rays; 11-64 » in|1-I8 rays; 15-54 m in| 2-20 rays; 13-60 p in
sphaerasters total diameter. total diameter. total diameter.
20 more cylindrical rays;| 10-17 more conic rays;|1-17 more — cylindrical
oe 14-26 p in total diameter. | 19-24 » in total diameter. | rays; 17-24 y in total
diameter.
all regularly ellipsoidal; | all regularly ellipsoidal; | besides the regular ellip-
105-122 » long; 90-114 »| 85-97 » long, 75-90 4 | soidal also rhomboidically
Sterrasters broad, 70-86 » thick. broad, 57-65 y thick. distorted ones, 90-111 y

long, 80-94 » broad, 65—
79 » thick.

The structure of the skeleton and the canal-system with its cribriporal
afferents and its uniporal efferents clearly show that this species belongs to
Sidonops. None of the previously described species either of Sidonops or

40 SIDONOPS OXYASTRA.

Geodia, which latter I have, for the reasons given in the description of Geodia
agassizii, also compared, at all resemble these sponges in their spiculation.
The only species similar to S. angulata is S. bicolor described in this Memoir,
and from this it is distinguished by the possession of anaclades and angularly
bent amphioxes, the smoothness of the oxyasters and oxysphaerasters, and

the smaller size of the sterrasters.

Sidonops oxyastra, sp. noy.

Plate 6, figs. 1-23; Plate 7, figs. 1-20; Plate 8, figs. 1-15.

I establish this species for two specimens from Duncan Island, Galapagos.

The asters of this species are all oxyasters (oxysphaerasters) and to this the
specific name refers.

The larger of the two specimens (Plate 6, fig. 5) forms a mass 94 mm. in
maximum diameter, attached to a flat, water-worn pebble, half of which it has
overgrown. Deep incisions partly divide this mass into lobes; the central
undivided part is 76 mm. broad; the lobes (Plate 6, fig. 4) taper distally and are
rounded. The surface appears undulating, smooth, and, to the unaided eye,
destitute of a spicule-fur. Large parts of the strongly convex and most exposed
portions of the surface are altogether without pores. In other exposed parts
a few minute pores are observed. The flat and the concave, more sheltered
parts of the surface are perforated by very numerous pores, two kinds of which
can be distinguished. The whole of the extensive flat surface of the central
mass and considerable parts of the surface of the lobes are occupied by sieves
containing small afferent pores. In some places, where the dermal membrane
forming these pore-sieves has been rubbed off the entrances to the radial afferent
cortical canals are exposed to view. There is a tract 4-14 mm. in extent
occupied by a group of large and conspicuous efferent pores (Plate 6, fig. 4) on
nearly every lobe.

The smaller specimen, which measures 56 mm. in maximum diameter,
resembles the larger one, described above, in every respect. It also grew, as
the impression in the detached base shows, on a flat pebble perhaps another
part of the one to which the larger specimen is attached.

A monaxonid sponge (Plate 6, figs. 19a, 20a; Plate 7, figs. 1b, 2b, 6b) and
composite ascidians incrust parts of the afferent areas of both specimens.

The colour of the surface of the large undivided central mass is in both
specimens (in spirit) purplish brown, that of the other parts much lighter,
brownish white. The interior is light brown.

SIDONOPS OXYASTRA. 4]

A cortex (Plate 7, figs. la, 2a, 6a), composed of a thin outer dermal layer,
a thick middle sterraster-armour layer, and a thin inner fibrous layer, is developed
superficially. The whole cortex is, under most parts of the surface, 700-750 y,
in some parts of the efferent areas up to 1.6 mm. thick. As such thicknesses of
the cortex have been observed only between widely open efferent cortical canals
they may be produced by the dilatation of the latter. The dermal layer is in
the afferent areas on an average about 40 « thick and excavated by systems of
subdermal canals. In the efferent areas it is on an average about 60 » thick,
solid, and occupied by numerous paratangentially disposed spindle-cells. The
sterraster-armour layer is in the afferent areas about 600 « thick and every-
where, except in the walls of the cortical canals, occupied by dense masses of
sterrasters. In parts of them the portions of the middle layer free from sterr-
asters around the cortical canals are rather extensive and considerably widened
distally (Plate 7, fig. 6). The inner fibrous layer is 35-70 » thick and occupied
by paratangential, somewhat undulating fibres, staining strongly with aniline-
blue. .

The choanosome is traversed by strands composed of large and conspicuous
elongated cells (Plate 6, figs. 1, 2). These strands are 60-90 » broad.- The cells
composing them are arranged rather irregularly, but on the whole distinctly
longitudinally. They are mostly spindle shaped, 20-30 » long and 4-7 »# thick.
Here and there (Plate 6, fig. 1, to the left below) thicker, more oval elements,
measuring 20 X 11 », are observed in the strands. The plasma of these cells is
occupied by large granules, staining strongly with haematoxylin.

Canal-system. The afferent areas of the surface are occupied by sieve-like
pore-groups 0.5-1 mm. in diameter (Plate 8, fig. 13). These pore-sieves lie
close together, being separated only by narrow poreless tracts. The afferent
pores themselves (Plate 8, fig. 15) are in the preparations, probably in conse-
quence of post mortem shrinkage, somewhat irregular in shape, 40-120 » wide,
‘and separated by dermal bands of varying breadth. The broader bands, which
form a sort of primary network, are thick, reach down to the middle layer of
the cortex, and contain asters and dermal rhabds. The narrower bands which
connect these primaries, are quite thin and contain only a few small asters or
no spicules at all (Plate 8, fig. 15). The pores of each group (pore-sieve) lead
into a system of subdermal cavities, which converge and unite to form a radial
canal. These radial afferent cortical canals, which penetrate the sterraster-
armour layer, are circular in transverse section and about 45 wide. They are
distributed somewhat regularly over the afferent areas, their centres being 0.7-1
mm. apart.

492 SIDONOPS OXYASTRA.

Below the cortex of the afferent areas numerous cavities, which appear
connected with each other by paratangential canals are met with (Plate 7, figs.
1, 2, 6). Into this system of subcortical cavities the radial afferent cortical
canals open out, and from it numerous narrow afferent canals, which extend
downwards into the choanosome, take their rise.

The intermediate tissue is poorly developed, the final ramifications of the
canals and the flagellate chambers being separated only by thin membranes
(Plate 6, fig. 3). The flagellate chambers are, so far as I could make out, spheri-
sal and measure 17—25 y in diameter.

The efferent canals join to very wide (up to 1.5 mm.) efferent canal-stems
(Plate 7, figs. 1d, 2d) which extend towards the efferent areas of the cortex,
below which they join to form a more or less continuous efferent subcortical
cavity. From this the radial efferent cortical canals take their rise. These
‘anals are 0.1-1 mm. wide, have a circular transverse section, and open out
freely on the surface. They are destitute of dermal pore-sieves (uniporal).
Their openings, the efferent pores (Plate 6, fig. 4; Plate 7, figs. le, 2e; Plate 8,
fig. 14), which occupy the efferent areas above described, are circular and, like
the canals which terminate in them, 0.1-1 mm. wide. The great difference in
size between the smallest and the largest of these pores is remarkable. The
small ones are few in number and scattered irregularly among the much more
numerous large ones. The centres of the efferent pores are, irrespective of the
width of the pores, quite uniformly 1.2 mm. apart, and the distance between
the margins of adjacent pores is consequently in inverse proportion to their size.
This and the fact that the cortex is thicker between large pores than between
small ones, seem to indicate that the great differences of width observed in the
efferent pores (cortical canals) are due to differences in degree of contraction.

Skeleton. In the interior of the choanosome numerous, rather irregularly
scattered amphioxes, some amphistrongyles, a few styles, large oxyasters, and
some sterrasters, mostly young forms, are met with. Towards the surface
rhabds, similar to those mentioned above, together with the rhabdomes of
numerous subcortical plagiotriaenes and a few small subcortical anaclades, form
radial bundles which abut vertically or somewhat obliquely on the cortex (Plate
7, figs. 1, 2, 6). In this subcortical region of the choanosome and in the inner
layer of the cortex also minute dermal rhabds occur; the (young) sterrasters
are here much less abundant than in the interior, and the large oxyasters of the
latter for the most part replaced by large oxysphaerasters. In the middle layer
of the cortex the sterrasters form a dense mass. The dermal layer contains

SIDONOPS OXYASTRA. 43

numerous small oxyasters and oxysphaerasters, numerous minute dermal rhabds,
and a few small anaclades. The small oxyasters and oxysphaerasters form a
dense coating at the surface. The minute dermal rhabds and anaclades traverse
the dermal layer more or less radially. Their proximal ends are implanted in
the distal part of the sterraster-armour layer, and their distal ends protrude
freely beyond the surface. The dermal rhabds are styles with attenuated,
proximally situated, rounded ends. In the efferent areas of the surface (Plate 6,
fig. 21) these spicules form dense masses. In the afferent areas they are not
nearly so numerous. The freely protruding ends of these spicules are in the
efferent areas very close together and nearly parallel, like grass on a good lawn.
In the afferent areas they form tuft-like groups of diverging spicules like grass on
arid ground. The anaclades are confined to the afferent areas. In one
place (Plate 6, fig. 23) I found them in great numbers. Generally they are
scarce. Where the monaxonid sponge-crusts, above mentioned, extend, these
spicules penetrate it, their cladomes lying within the attached sponge-crust
(Plate 6, figs. 19b, 20b), anchoring it to the Sidonops. These anaclades are
mostly anatriaenes, but anadiaenes, anamonaenes, and mesanaclades, chiefly
mesanatriaenes, also occur.

The large choanosomal amphioxes (Plate 6, fig. 14; Plate 8, figs. 4a, 5) are
straight or slightly curved, 1.1-1.55 mm. long and 10-82 » thick.

The rare large amphistrongyles are straight, isoactine, and 0.8-1 mm. long.
They are in the middle 18-23 thick and taper towards the two equal, rounded
ends. The degree of attenuation is variable, as the following three measure-

ments show.
Thickness in the middle Thickness at the ends
23 pL 21
20 p 17
18 uu Wp

The very rare large styles are straight and shorter and, at the rounded end,
thicker than the amphioxes and amphistrongyles. One that I measured was
850 « long and, at the rounded end, 38 » thick.

The minute dermal styles (Plate 6, figs. 21, 22) are more or less, sometimes
rather abruptly, curved, 130-230 y, usually about 200 long, and, at the thickest
point, which lies between the middle and the rounded end, 3-5.5 # thick. They
taper towards both ends; the distal end is sharp pointed; the proximal end
rounded and 1.5-3 » thick, usually a little less than half as thick as the spicule at

its thickest point.

44 SIDONOPS OXYASTRA.

The plagiotriaenes (Plate 6, figs. 6-13; Plate 8, fig. 4b, c) have a conical
sharp-pointed rhabdome, which is straight or slightly curved in its acladomal
part. The rhabdome is 1-1.65 mm. long and, at the cladome 24-40 p thick.
The clades are usually conical and pointed; rarely one (Plate 6, fig. 7), two, or
all three (Plate 6, fig. 6) are reduced in length and rounded off terminally.
The normal pointed clades of the same cladome are usually about equal in
length, more rarely distinctly unequal (Plate 6, figs. 8, 10). The basal part
of the elades is directed obliquely upward and always curved, concave to the
rhabdome, their distal part is directed outward and straight or slightly curved
in the opposite direction. The development of this upward bend of the distal
part of the clade is usually proportional to its length. The chords of the
normal (pointed) clades are 250-285 » long and enclose angles of 100-118°, on an
average 108.5°, with the axis of the rhabdome.

In a spicule-preparation I found a triaene with a clade-chord 350 # long,
enclosing an angle of 90° with the rhabdome. Perhaps this orthotriaene is a
foreign spicule.

The rhabdomes of the small dermal anaclades (Plate 6, figs. 15-18, 19b,
20b, 23) appear — I found none intact in the spicule-preparations — to be over
1mm. long. They are, at the cladome, 5-12 » thick. Their cladomes are very
variable. The most frequent forms are anatriaenes (Plate 6, figs. 17, 18, 19b,
20b) with a protuberance on the apex of the cladome. Their clades are pointed,
very rarely blunt, more or less angularly bent, concave to the rhabdome and
often rather unequal. This inequality is sometimes carried to the extent of a
complete suppression of one or two clades, whereby diaene and monaene forms
are produced. Not infrequently the apical protuberance is replaced by a long,
blunt (Plate 6, fig. 15, 16) or, more frequently, pointed epirhabd. Most of
these mesanaclades are quite regular mesanatriaenes (Plate 6, fig. 15). Some
of them are, however, rendered irregular by one of the clades extending upwards,
proclade-fashion (Plate 6, fig. 16). These mesanatriaenes were found only at
the place where the anaclades are abundant. The chords of the clades of the
more regular triaene and mesotriaene anaclades are 15-24 » long and enclose
angles of 40-65°, on an average 57°, with the axis of the rhabdome. The
clades of the diaene and monaene anaclades are longer, some of them attain-
ing a length of 30 #. The epirhabd is, when fully developed, straight, conic,
sharp pointed, and 65-75 » long.

Although the different kinds of ewasters are, to some extent, connected
by transitional forms, three categories can readily be distinguished: large

SIDONOPS OXYASTRA. 45

choanosomal few-rayed forms, without large centrum (large oxyasters); large
subcortical many-rayed forms, with large centrum (large oxysphaerasters); and
small, dermal, mostly many-rayed forms, with or without large centrum (small
oxyasters and oxysphaerasters).

The large choanosomal oxyasters (Plate 7, figs. 3-5a, 7, 8, 13-15) have a
slight central thickening 2.5-4.5 », that is two to three times the basal thickness
of the rays, in diameter, and from four to ten, most frequently seven, straight,
conical and pointed or blunt, concentric, and quite regularly distributed rays.
With the exception of its proximal end, the whole of the ray is covered with
rather large and uniformly distributed spines (Plate 7, figs. 13-15). The rays
are 11-25 » long and at the base usually 1.2-2 » thick, the total diameter of
the aster being 18-45 ». A few asters of this kind, with much thinner rays also
occur. These asters, which are less than 20 » in total diameter, have rays, at
the base, only 0.5-0.7 » thick. They are probably young forms.

The large subcortical-oxrysphaerasters (Plate 7, figs. 3c, 19, 20) have a spherical
centrum 4.2-6.5 », that is from a quarter to a third of the whole aster, in diameter,
and from sixteen to twenty-three concentric and regularly distributed rays.
The rays are straight, conical, sharp pointed, covered with rather large spines,
6-7 » long and at the base 1.1-1.4 » thick. The whole aster is 16-22 » in
diameter.

The small dermal oxyasters and oxysphaerasters (Plate 7, figs. 3-5b, 9-12,
16-18) form a continuous series. One end of this series is represented by forms
which have hardly any central thickening at all and appear as true oxyasters
(Plate 7, figs. 9 and 10 right above, 18). The other end of the series is represénted
by forms with a centrum more than a third of the whole aster in diameter. The
small dermal oxyasters and oxysphaerasters have from nine to eighteen straight,
conical, and regularly distributed rays. The rays always appear to bear nu-
merous small spines. Often however these spines are so minute that they can-
not be made out as such, a roughness of the ray then being the only indication
of their presence. The rays are (without the centrum) 2-4.5 » long and at the
base 0.7—-1.5 « thick, the total diameter of the aster being 6-13.5 p.

The sterrasters (Plate 6, fig. 21; Plate 8, figs. 1-3, 6-12) are flattened ellip-
soids, 76-85 » long, 66-73 « broad, and 50-64 « thick, the average proportion of
length to breadth to thickness being 100 : 90: 76.

Very young sterrasters, some hardly 10 # in diameter, were observed.
These appear as spheres composed of equal and regularly distributed, im-
measurably thin, straight, radial rays. In a spicule-preparation I found a

46 SIDONOPS BICOLOR.

chip of an adult sterraster, which I was able with a high power to photograph
(Plate 8, fig. 12). This photograph shows that the centre of this spicule is
oceupied by a cluster of granules (a), from which the radial lines, traversing the
body of the sterraster, arise. The central granule-cluster is about 4 # in
diameter.

The rays protruding over the surface of the solid centrum are in most of
the adult sterrasters (Plate 8, figs. 10, 11) 2-3 » thick and provided with terminal
verticils of usually five or six lateral spines. In some of the sterrasters (Plate
8, figs. 6-8) these rays are 3.5-4.5 « thick and usually provided with seven
or eight lateral and one or more obliquely arising terminal spines.

Both specimens were collected on April 13, 1888, at Duncan Island, Gala-
pagos. They were labeled F. C. 1354 and 539 Tetractinellida.

The structure of the skeleton, the cribriporal afferents, and the uniporal
efferents show that this species belongs to Sidonops. It is not at all closely
allied to any other species of Sidonops or to any species of Geodia. The species
approaching it most closely appears to be Geodia media Bowerbank from which
however it differs by the small euasters for the most part being thick- and
short-rayed strongylosphaerasters and by the presence of mesomonaenes.

Sidonops bicolor, sp. nov.

Plate 9, figs. 1-19; Plate 10, figs. 1-15; Plate 11, figs. 1-17.

I establish this species for fifteen specimens obtained off California. All
the specimens agree closely and most of them are very much lighter in colour on
one side than on the other. The specific name, bicolor, refers to this conspicuous
character.

Shape and size. These sponges are irregularly tuberous, and generally
considerably elongated (Plate 11, figs. 15, 16) or flattened (Plate 11, fig. 17).
The largest elongated one, which was obtained at Station 2958, is 101 mm. long
and 40 mm. thick. The largest flattened one, collected at Station 2981, is 62
mm. long, 59 mm. broad, and 28 mm. thick. The others are 39-73 mm. in maxi-
mum diameter. Most of the efferent pores are situated on the less extensive
concave parts of the surface, while the afferents are chiefly on the more extensive
convex parts. The areas bearing chiefly the efferent pores are either quite
smooth or slightly raised around these pores, some of which are situated on the
summits of low elevations. The areas bearing chiefly the afferent pores are
more uneven and appear to have been entirely covered with a spicule-fur.
Although rubbed off in many places, remnants of this spicule-fur can easily be

SIDONOPS BICOLOR. 47

found on the more sheltered parts. Incrusting symbionts, desmacidonid
sponges, two species of Bryozoa, etc., grow on the afferent portions of the surface
of most of the specimens, while the efferent areas are free from symbionts.

The colour of the surface varies in these spirit specimens from whitish to
reddish or purple-brown; some parts of it are, as mentioned above, usually
much lighter in colour then others, the under side appearing to be less pigmented
than the upper side. Occasionally, particularly in the specimens from Station
4420, I have noticed that the margins of the efferent pores are somewhat lighter
in colour than the adjacent parts of the surface. The interior is dirty brownish
or greenish white.

The superficial parts form a cortex (Plate 9, figs. 15-17) which contains
a sterraster-armour 0.9-1.8 mm. thick. The sterrasters do not always extend
down to the choanosome, a thin fibrous layer often intervening between them
and the latter. This layer, which is composed of paratangential fibres similar
to those connecting the sterrasters, is more clearly made out in one of the speci-
mens from Station 4420 than in the others. In the darker parts of the cortex
pigment cells are observed. These contain large spherical granules, brown in
colour, which stain deeply in azure. The number of these granules in each cell
is not great. On or just below the outer surface the pigment cells are very
numerous and often form a continuous layer which has the appearance of an
epithelium. This I observed chiefly in a specimen from Station 4420. These
cells are here massive, or somewhat elongated, irregular in outline, about 10
broad and 12-25 » long. Pigment cells also occur in the lower parts of the
cortex, within the sterraster layer, but here they are long and slender, and
arranged radially around the sterrasters. This shape and position of the deep-
lying pigment cells are apparently due to the position of the connective-tissue
fibres which radiate from the sterrasters and between which they lie.

In the choanosome of a specimen from Station 3168 I found numerous oval
bodies 20-35 » long and 10-20 » broad which consist of a nearly hyaline sub-
stance uniformly staining with haematoxylin and azure. In these bodies
neither an enveloping membrane nor a nucleus could be detected. Most of
them are densely crowded in band-like zones, some isolated and scattered.
Similar bodies, scattered singly throughout the choanosome, have also been
observed in a specimen from Station 4420.

Canal-system. In many of the specimens I have been able to make out the
afferent pores. These are, as stated above, chiefly distributed over the convex

parts of the surface and arranged in more or less circular groups (Plate 10, fig. 15)

48 SIDONOPS BICOLOR.

0.5-1.5 mm. in diameter. The centres of these pore-groups are 1.5-2.5 mm.
apart. When the groups are large and their centres close together, as is the case
on parts of the surface of one of the specimens from Station 4531, the pore-
eroups come in contact with each other and form a fairly continuous sieve.
When however, as is more frequently the case, the pore-groups are smaller and
farther apart, they appear divided by a network of belts free from pores. The
pores themselves are oval, measure 100-300 y in diameter, and perforate the
thin dermal membranes covering the distal widened parts of the afferent canals
(Plate 9, fig. 16), which traverse the cortex in a radial direction. These canals
are cylindrical in the centre, and up to 400 « wide. They widen above in a funnel-
shaped manner, and are contracted below by a stout chonal sphincter, which
lies at the level of the limit between cortex and choanosome. They lead into
subeortical cavities of no great size which lie just below the cortex and from which
the afferent choanosomal canals take their origin. The flagellate chambers
(Plate 9, fig. 18) are spherical and 20-32 » in diameter. The efferent canals are
provided with sphincter-membranes at frequent intervals and join to form tubes,
often as much as 1—-1.6 mm. in diameter, which lead up to the efferent areas of
the cortex (Plate 9, fig. 17). Some appear to end at the limit between cortex
and choanosome, while others bend round and continue their course paratangen-
tially for some distance just below this level, thus forming efferent subcortical
cavities. From the ends of the former and the roofs of the latter the efferent
cortical canals arise. These are constricted at their origin by chonal sphincters
lying at the level of the limit between cortex and choanosome. Beyond the
sphincter the canal widens to a cylindrical tube 250 #-1 mm. in diameter, which
traverses the cortex radially (Plate 9, fig. 15) and opens out freely on the surface
(Plate 10, fig. 14). The efferent pores, in which these canals terminate, either
have nearly the same width as the canals themselves or they are slightly
smaller. The centres of these efferent pores are 1-2.5 mm. apart, their dis-
tance being on the whole proportional to their size, small ones lying much
closer together than large ones. These efferents, though usually restricted to
concave parts of the surface which are generally free from afferent pores, are
also found on other parts of the surface, irregularly distributed between the
groups of afferents. On large parts of the surface no pores of any kind can be
made out.

Skeleton. Spicule-bundles which widen out distally traverse the choano-
some (Plate 9, fig. 17) radially and abut vertically or somewhat obliquely on
the surface. These bundles consist chiefly of amphioxes, of which two kinds,

SIDONOPS BICOLOR. 49

stout and slender ones, can be distinguished. The former occur chiefly in the
axial parts of the bundles, while the latter predominate in their superficial parts.
In the distal portions of the spicule-bundles plagiotriaenes also occur. The
cladomes of most of these plagiotriaenes lie at or just above the limit between
choanosome and cortex, the clades being often quite enveloped by sterrasters;
their rhabdomes extend radially inward. The radial spicule-bundles abutting
on the afferent areas do not terminate at the cortex, but penetrate it (Plate 9,
fig. 16) and protrude beyond it, thus forming the fur. The fur consists chiefly
of amphioxes but plagiotriaenes also take part in its formation. In the spicule-
fur of a specimen from Station 2958 I found a good many plagiotriaenes, with
thabdomes implanted in the cortex, and free cladomes lying a considerable
distance above the surface of the sponge. Where these spicules arise from it, the
surface is often raised conulus-fashion. In a specimen from Station 4551 I have
found a few styles and in the spicule-preparations of specimens from Stations
2958 and 3168 two small protriaenes. The latter are probably foreign spicules.
Very small and slender rhabds are often found imbedded in the superficial part
of the cortex. I think it highly probable that these belong to the symbiotic
monaxonid sponges which incrust parts of the surface.

The microscleres are strongylosphaerasters, smaller oxysphaerasters with
numerous rays, larger oxyasters with fewer rays, and sterrasters. The strongylo-
sphaerasters form a dense layer on the outer surface (Plate 10, fig. 15) and are
absent in the interior. The oxysphaerasters occur chiefly in the walls of the
cortical canals and are also met with in the region of the subcortical cavities.
The oxyasters are restricted to the choanosome, in the walls of the canals of
which they are very numerous. The sterrasters occupy the whole or nearly the
whole of the thickness of the cortex (Plate 9, figs. 15-16) in dense masses. In
most of the specimens the choanosome is free from sterrasters. In some how-
ever, particularly in the specimens from Stations 3168 and 4420, considerable
numbers of sterrasters, chiefly young ones, were found in it.

The stout amphiozes (Plate 9, figs. 9-11) are curved, isoactine, or slightly
anisoactine, attenuated towards the rather blunt ends, gradually in the central
parts and rather abruptly in the distal parts. They are 2.3-5.6 mm. long and
35-105 » thick. An inverse proportion between length and thickness is indi-
cated. Those of the specimens from Station 4551 attain a greater maximum

thickness than those from the other stations.

50 SIDONOPS BICOLOR.

DIMENSIONS OF STOUT AMPHIOXES.

Stations 2958 2981 | 3168 | 4420 4531 4551
Hengthinm= ee eet ce Oster nae 3-4.2 |2.3-5.3 | 3.14.2 3.1-3.7 |3.8-5.6)lees_5n6
SRHIGKM GSS Wile scene ee ai esas es ee 50-82 35-93 | 60-92 | 53-79 | 60-85 | 65-105

The slender amphioxes (Plate 9, figs. 7, 8) are curved in a simple (Fig. 7) «
or wavy (Fig. 8) manner, cylindrical in the central parts, and attenuated to fine
points at the ends. They are 3.5-9 mm. long and 15-40 + thick. The longest
measured were from specimens from Stations 4531 and’4551, but as the long ones
are usually broken in the spicule-preparations it is probable that slender amphi-
oxes, considerably longer than those observed and measured, occur also in the

specimens from the other stations.

DIMENSIONS OF SLENDER AMPHIOXES.

4420 4531 4551

Stations | 2958 | 2981 | 3168 |
| iz i

lengthens #484 eo odor anne 5.2-5.9|5.2-6.7| | 4.7-5.8| 3.5-9 | 69
| :

Kiiclenesa fiche Rea Pass we ache! 25 | 15-25 | 23-33 | 18-40 | 24-40 | 22-38

The exceedingly rare styles, which I observed only in the specimens from
Station 4551, are a little over 4 mm. long and at the somewhat thickened rounded
end 100-120 » thick.

The plagiotriaenes (Plate 9, figs. 1-6, 12-14) generally have a straight
conical rhabdome, rather abruptly attenuated at the acladomal end and
pointed (Plate 9, figs. 13, 14). The rhabdome, when thus normally de-
veloped, is 2.1-4 mm. long and at the cladomal end 62-110 # thick. Just
below the cladomal end it is markedly thickened and here attains a transverse
diameter of 73-120 ». This thickest part of the rhabdome is 5-20, most fre-
quently about 11 % thicker than the cladomal end, which consequently appears
constricted in a neck-shaped manner. In some plagiotriaenes the rhabdome
is reduced in length, more cylindrical in shape, and simply rounded off at the
acladomal end (Plate 9, fig. 12). Occasionally this reduction goes so far that
the rhabdome measures only 290 » in length. Such very short rhabdomes are
cylindrical and not constricted at the cladomal end. They have been observed
only in a specimen from Station 4420 and here also they are rare. The clades
are conical, quite blunt, and fairly straight (Plate 9, figs. 1, 3-6, 12) or

SIDONOPS BICOLOR. 51

slightly curved, either simply, concave towards the rhabdome (Plate 9, fig. 2),
or in the shape of an §, in such a manner that the proximal part is concave, the
distal part convex towards the rhabdome (Plate 9, fig. 13). The clades are
280-700 » long; their chords enclose angles of 103-122° with the rhabdome.

DIMENSIONS OF THE NORMAL PLAGIOTRIAENES.

Stations 2958 2981 3168 | 4420 4531 | 4551
length mm. 2 6=3),5|| DfES) 5)|| O78) | GVeRe eis |G} Tes} 6) I gy a
at the clado-| 79-93 | 62-110 | 68-8so | 68-83 | 73-100 | 80-100
mal end
Rhabdome == = - -
thickness | of the thickest
pee, eI) ae eae bie 5 ees ‘
Helomecheuclae 73-98 78-120 78-90 80-100 | 86-107 | 98-120
domal end 4
(CIAGLE, Soy Na To ene eee 320-620 | 350-590 | 280-530 | 400-550 | 400-700 | 360-650
Angle between clade-chords and rhabd- |

= 2 2-199 3-112 | Q— 199 | Bari
i, 100-118 | 108-122 | 103-112 | 108-116 | 104 122 | 114-120

The large choanosomal oxyasters (Plate 10, figs. 6-13a; Plate 11, figs. 6b,
8b, 9) have from one to twelve rays. Forms with one ray are exceedingly rare
and were observed only in a specimen from Station 4420. Forms with two rays
were not found. Three-rayed forms are met with in small numbers in the
specimens from Stations 3168 and 4551, four-rayed ones also in the specimens
from Station 4531. By far the most frequent forms are those with from five
to nine rays which occur in large numbers in all the specimens. The many-
rayed oxyasters pass into the oxysphaerasters. The rays are 1—-2.8 « thick at
the base, conical and straight. In the many-rayed oxyasters they are always,
in the few-rayed ones usually, sharp pointed. In the few-rayed oxyasters the
rays are irregularly distributed and apparently not always quite concentric,
many of these spicules appearing somewhat metastrose. In the oxyasters with
five or more rays, the rays are quite regularly distributed and concentric, but
these asters are also occasionally rendered somewhat irregular by one of
their rays being bifid. The distal parts of the rays are always spiny. In most
cases the spines are confined to the distal third of the rays, in some they cover
as much as the distal two thirds. The spines are either numerous and small,
or sparse and large. The sparse, large spines are slender and rise vertically
from the ray. A central thickening 2.8-6 » in diameter is nearly always devel-
oped. The whole oxyaster measures 20-34 in diameter. The length and

52 SIDONOPS BICOLOR.

thickness of the rays and, apart from the one-rayed forms, also the total diameter
of the aster are roughly in inverse proportion to the ray-number. In the rare
one-rayed form the centrum is 5 « in diameter, and the ray 18 long and 5 pu
thick at the base. The oxyasters with from three to five rays measure 28-34 4,
the oxyasters with-six or more rays 19-29 » in total diameter. The oxyasters
of the specimens from Station 4531 are slightly smaller than those of the others.

DIMENSIONS AND NUMBER OF RAYS OF OXYASTERS.

Stations 2958 2981 3168 4420 4531 4551
Total nae RBA arrester Koen pe. 22-34 23-34 20-32 19-28 22-33
Diameter of centrum pt. ....-...---... 3. 34.2 36 3.5-4 28-5 34.5 3-5
3asal thickness of rays #...-..-....- 1.3-2.3 | 1.5-2.8 2-3 1.5-5 12.5 | 1.62.5
NII DeLI On TVS tere ett ere oe 78 6-10 3-9 . 1 5-10 4-12 3-12

The orysphaerasters (Plate 11, fig. 7b) have a spherical centrum 4.5-10 4
in diameter, from which from twelve to twenty-five rays arise radially. The
rays are concentric, distributed regularly, 0.7-2.5 » thick at the base, conical,
straight, and pointed at the end. The distal part of each ray, usually about the
distal half of it, is covered with spines. Not infrequently a group of somewhat
larger spines, arranged in a more or less verticillate manner is situated a short
distance below the end of the ray. In the centrifugal spicule-preparations of the
specimen from Station 2958 I found several oxysphaerasters with rays entirely
destitute of spines, otherwise similar to the ordinary ones. These may possibly
be foreign to the sponge. In total diameter the oxysphaerasters measure 10-23
uw. The largest ones pass into the oxyasters described above. The oxysphaer-
asters of the specimens from Stations 4531 and 4551 have on the whole more
rays than those from other stations, and specimens with more than twenty rays
have been observed only in the specimens from these stations.

DIMENSIONS AND NUMBER OF RAYS OF OXYSPHAERASTERS.

PS) EC tere ce SOO IE Dey peeise GAA Tees 2958 2981 3168 4420 4531 4551
‘Pobal\ diameter: ji) eke ee ee 14.5-21.5} 17-21 19-23 | 14-18.5 | 14-16.5 | 10-19
Dinats of centrum yt. .. ae ee = Sie 4.5-10 67.5 6-7 5.5-7 6-6.5 47.5
Basal dicks of ae hte oma ar heya: 12.5 ¥ 1.5-2 1-2.5 |1.5-2.5|] 1.5-2 |0.7-1.8
rnnes of aos ; hes 3 Aa : We, 12-16 7 14-18 14-17 14-17 19-23 19-25

SIDONOPS BICOLOR. 53

The strongylosphaerasters (Plate 10, figs. 6b, 11-13b; Plate 11, figs. 1-5,
6-8a, 10) have a spherical centrum 4-14 » in diameter, from which generally
from nine to thirty rays arise radially. Exceptionally there is only one ray.
The rays are concentric, regularly distributed, straight, and at the base 0.7-
3.5 # thick. They are cylindrical or cylindroconical, truncate, and 1.5-6 long.
Their terminal face and the distal parts of their sides are covered with numerous
small spines; the proximal parts of the rays and the central thickening are usually
quite smooth. The only exception to this is the one-rayed strongylosphaeraster
found in a specimen from Station 4531, in which the whole of the ray and also
the central thickening are covered with spines. The centrum of this spicule is
9 «41n diameter and the single ray 2.5 » thick at the base and 5 » long. Two-
to eight-rayed strongylosphaerasters were not observed, and the nine- to eleven-
rayed forms were found only occasionally in specimens from Stations 2981 and
4420. The strongylosphaerasters of the specimens from Stations 4531 and 4551
have on the whole more rays than those of the others. The total diameter of
the strongylosphaerasters is 9-22 yn.

DIMENSIONS AND NUMBER OF RAYS OF STRONGYLOSPHAERASTERS.

7 |

TSU OIE: coed ee Se ae eee eee 2958 2981 | 3168 4420 4531 4551
Motalidiameterj-)-...-....-...--. | 11.5-21| 9-22 |11:5-16| 9-22 11-15 | 9-16
Diameter of centrum p............... 6.5-12 4-10 5.5-9.5| 6-14 7-10 4.5-9

length « 2.5-4.5| 2.5-6 2-3.5 2-4.5 1.7-5 1.5-4
Rays — ——|-— = =

basal thickness 1.5-2.8 | 1.5-3.5 | 1.5-2.5 | 1.5-3.5 | 1.2-2.5] 0.7-3
PEPEDCTIORMTAVS irae os oisc- sec esses 12-21 9-25 14-22 10-19 1.17-27| 18-39

The sterrasters (Plate 10, figs. 1-5; Plate 11, figs. 11-14) are usually flat-
tened ellipsoids, 130-170 » long, 100-133 » broad, and 77-97 » thick. A few
are somewhat rounded, triangular (Plate 10, fig. 1), not oval, in outline. In the
ellipsoidal sterrasters the ratio between the length and breadth is on an average
100 :76. The specimens from Stations 2958 and 4531 have on the whole some-
what broader (ratio 100 :78 and 100 : 82 respectively), those from Station 4551
somewhat narrower (ratio 100 :70) sterrasters. The specimens from Station
2981 have slightly larger sterrasters than those from the others. The umbilical
pit is usually about 12-15 » deep and situated in the centre of one of the broad
faces of the flattened sterraster. The distal free parts of the rays composing
the sterraster are 2.5-4 » thick and about 2 » apart. Those surrounding the

54 SIDONOPS BICOLOR.

umbilicus usually have an elongated (Plate 11, fig. 12), those remote from the
umbilicus, a circular or polygonal (Plate 11, figs. 13, 14) transverse section.
Each ray bears a terminal verticil of from two to eight stout, conical, lateral
spines arising vertically from the ray.

The young sterrasters observed in great numbers in the choanosome of
the specimens from Stations 3168 and 4420 were surrounded by stratified
capsules readily stainable with azure and apparently composed of flat endothelial
cells. The smallest of these young sterrasters was a sphere, 20 « in diameter,

composed of exceedingly fine radial rays.

DIMENSIONS OF STERRASTERS.

SS Bs SAGER OS CU DU OROR EO OROD eb EoS 2958 2981 3168 4420 4531 4551
erate terse ee te rere chante Mees 130-145 | 145-170 | 140-165 | 135-155 | 135-155 | 140-145
Breaths pics tee, arprenc Aierciera rete chore 100-120 | 105-133 | 105-120 | 105-118 | 115-120 | 100-120
Thickness TORN aes Se, SACRA Se cay RAS 80-92 i 90-97 85-90 85-87 88-90 77-85

LOCALITY AND NATURE OF ENVIRONMENT.

| | n
e | Bottom 55
ss Locality ‘ Date Depth tem pera- Bottom oe
Le ture Pad
Zn z
sare | Off southern California, 34° 04’ 47 m. WD
2958 : 9 Feb., 1889 Gray s
© | N., 120° 19" 36” W. Wseenc (26) |co4.gery| C7*Y sand :
P Off southern California, 33° 18’ | = 82 m. Coarse gray sand
2981 | - : 13 Feb., 1889 = gray 2
| N., 119° 24’ W. (45 f.) and broken shells
e1eg | Off central California, 38° 01’ | 24 March 64m.
3168 , , ——a e
25” N., 123° 26" 55” W. 1890 (34f.) Rocks andi]
Off southern California, E. of
9, | Point San Nicolas Island, S. 12 April 58-60 .
4420 , pri’, mi. =
77° W. 10.5 km. (5.7 miles),| 1904 (32-33 f.) Pine ereyezed é
drift S. 60° W.
4 | ones Bay, Cal., Point 28 May, Fine gray sand,
4531 | Pinos Light House. N. 64° E., 1904 — -- pebbles, and rock 3
3.8 km. (2.1 miles).
Monterey Bay, Cal., Point
__. | Pinos Light House, S. 9° E., 7 June 102 m. Coarse sand
ABest Bs eee : — : 1
8.4 km. (4.5 miles), drift S. 37° 1904 (56 f.) shells, and rock
FE.

GEODIA VARIOSPICULOSA. 55

The agreement between the specimens described above as Sidonops bicolor
is so great that there cannot be any doubt as to their identity; in fact not
even varieties or forms can be established.

Since the spiculation is geodine in character, the afferent cortical canals
eribriporal, and the efferents uniporal, this sponge must be placed in Sidonops.
For the reasons given in the description of Geodia agassizii, | have compared it
not only with the known species of Sidonops but also with those of Geodia.
The species of these genera which seems to be most closely allied to it is the
one described in this report as Sidonops angulata. This differs from S. bicolor
by the possession of anatriaenes and angularly bent amphioxes, by the smaller
size of the sterrasters, and by the oxyasters and oxysphaerasters always having
perfectly smooth rays. These differences are certainly sufficient for specific
distinction.

GEODIA Lamarck.

Among the megascleres are regular triaenes. The tetraxon megascleres
are confined to the superficial part and arranged radially. The dermal micro-
scleres are asters. The afferents and efferents are both cribriporal.

There are sixty-two specimens of Geodia in the collection made by the
“Albatross.” These belong to thirteen species, ten of which are new. Two
species, one previously known and one now described for the first time, are
further divided into seven varieties, five of which are new.

Geodia variospiculosa THIELE.

Zoologica, 1898, 24, p. 10, taf. 6, figs. 6-7. Lendenfeld, Tierreich, 1903, 19, p. 107.

intermedia, var. nov.

Plate 17, figs. 23-26, 34-40, 49; Plate 18, figs. 8, 10, 13-20, 22, 27; Plate 19, figs. 9-11, 19, 20, 22, 24, 31.

micraster, var. nov.
Plate 17, figs. 27-33, 41-48, 50; Plate 18, figs. 1-7, 9, 11, 12, 21, 23-26; Plate 19, figs. 1-8, 12-18, 21,
23, 25-30, 32.

Two specimens collected by the ‘Albatross’? off Honshu Island, Japan,
differ from the typical G. variospiculosa Thiele, from the var. clavigera Thiele,
and from each other, sufficiently to rank as new varieties. In both these new
varieties the choanosomal oxyasters are considerably smaller than in the type,
and in one of them they are not so large as in the other. The variety in which
they are smallest I name micraster, the one in which they are not so small

intermedia.

56 GEODIA VARIOSPICULOSA.

Shape and size. The specimen of var. micraster (Plate 17, fig. 41) is tuberous
and measures 42 mm. in length, 34 mm. in breadth, and 24 mm. in height. The
surface is somewhat undulating. Here and there slight, abrupt, step-hke
changes of level of about 0.3 mm. are observed in it. In one place there is a
round hole, a little over 1 mm. wide. This appears to be the entrance to a
tubular cavity occupied by an annelid, and not an osculum. Apart from this
no apertures are visible. Large parts of the surface are occupied by shallow
pits, the centres of which are less than 1 mm. apart. A dense fur of minute
spicules, about 100-200 » high (Plate 18, fig. 26c), visible only with the micro-
scope in sections, covers the whole of the surface. Besides this, remnants of a
sparse fur of large spicules, 2-3 mm. high, are observed in unexposed places.

The specimen of var. intermedia (Plate 17, figs. 39, 40) is thick oval, slightly
constricted near one end, 22 mm. long and 15 mm. broad. It was attached at
one side. The surface is continuous, without undulations or step-like changes
of level, free from apertures visible to the eye, and for the greater part covered
with shallow pits, more conspicuous on one side than on the other. Where
these pits are more conspicuous their centres are 1-1.2 mm., where they are less
conspicuous 0.6-0.9 mm., apart. A low dense spicule-fur is also present in this
variety, but there are hardly any traces of a sparse, high spicule-fur.

The colour (in spirit) of var. micraster is nearly white, that of var.
intermedia light coffee-brown.

The superficial part of the body forms a cortex composed of three layers:
an outer dermal layer free from sterrasters, a central sterraster-armour layer,
and an inner fibrous layer containing but few sterrasters or none at all. In the
pits the dermal layer is considerably thickened, on other parts of the surface it
is very thin. The sterraster-armour layer is chiefly composed of sterrasters.
The connective-tissue fibres radiating from the sterrasters and connecting them
with each other are very conspicuous. Around the proximal parts of the radial
canals which penetrate the sterraster-armour rather extensive zones of chonal,
fibrous tissue, free from sterrasters, occur (Plate 18, fig. 21). The sterraster-
armour layer and the dermal membrane are together 0.5-1 mm. thick. Sections
show that the above mentioned step-like changes of level in the surface of var.
micraster are caused by abrupt changes of thickness of the sterraster-armour
layer, the lower limiting surface of this layer passing smoothly and unchanged
beneath the steps. The fibrous layer of the cortex (Plate 18, figs. 21, 26) is for
the most part over 1 mm. thick and in many places excavated by cavities.
The walls separating these are chiefly composed of fibres arranged obliquely

GEODIA VARIOSPICULOSA. 57

or radially. Strands of such fibres extend from this layer far down into the
choanosome.

Canal-system. The pits on the surface are covered by pore-sieves. These

are so numerous and so close together that they join to form extensive con-
tinuous pore-areas. The pores are generally broad-oval, 20-75 » long and 15-
55 « broad. They seem to be larger in var. intermedia than in var. micraster.
The strands of tissue separating them vary very much in width. Some are so
narrow as to appear as slender threads, while others are as broad as or even
broader than the pores. On examining pore-sieves, removed by a paratangential
section in transmitted light, one clearly sees that these sieves are composed of
stout, primary, non transparent bars, the interstices between which are occupied
by secondary nets of thin, transparent strands (Plate 18, fig. 9). The stout bars
of the primary network usually exhibit a somewhat radial arrangement round
the centre of the pore-sieve (pit). Below these sieves rather extensive
cavities occur into which the pores lead. These cavities join under the centre
of each pore-sieve, that is, in the centre of each pit, to form a radial canal which
penetrates the cortex and either opens out below into a subcortical cavity
(Plate 18, fig. 21) or is continued as a narrow, usually tortuous canal leading
down into the choanosome (Plate 18, fig. 26). The proximal third or half of
each radial cortical canal is surrounded by a stout chonal sphincter, which
does not extend proximally beyond the lower limit of the sterraster-armour
‘layer. In the sections examined the radial cortical canals are constricted and
their proximal portions, which pass through the chonal sphincter, are often
quite closed. In the interior some large canals, in var. intermedia up to 0.8, in
var. micraster up to 2 mm. wide, are observed. In the latter a cavity, about
1 mm. wide, surrounded by sterrasters, was observed 2 mm. below the surface.

Skeleton. Rather loose strands of amphioxes and a few tylostyles and
styles traverse the inner part of the choanosome. These internal spicule-strands
are not arranged in a regularly radial manner. Many are very oblique, and
once I saw one extending paratangentially about 6 mm. below the surface.
Distally the megascleres form bundles which penetrate the inner layer of the
cortex and terminate at, or a little above, the lower limit of the sterraster-armour.
These distal spicule-bundles (Plate 18, fig. 8b) are vertical or oblique to the
surface and contain, besides the large amphioxes and occasional monactines
found in the interior, numerous rhabdomes of orthoplagiotriaenes and some
rhabdomes of dichotriaenes, large anatriaenes, and mesoclades, chiefly mesopro-
triaenes. The cladomes of most of the orthoplagiotriaenes and dichotriaenes

58 GEODIA VARIOSPICULOSA.

lie at, or just above, the lower limit of the sterraster-armour layer; the cladomes
of the anatriaenes and mesoproclades lie at different levels. The sparse high
spicule-fur is composed of the distal, freely protruding parts of mesoproclades,
chiefly mesoprotriaenes, and large anatriaenes. On small parts of the surface
of the specimen of var. micraster freely protruding orthoplagiotriaenes occur.
I do not believe, however, that these spicules normally take part in the formation
of the fur, but consider that the sponge must, at a previous time, have received
some injury in the places where these spicules are found.

In the inner layer of the cortex numerous small styles and minute anaclades
(Plate 18, figs. 8c, 26e) are observed. These are situated radially or obliquely,
rarely paratangentially. Some of them form groups, others are scattered singly.
Some lie between the bundles of large spicules (Plate 18, fig. 26e), others form
clusters around them (Plate 18, fig. 8c). In the thin walls of tissue separating
the subcortical cavities strand-like rows of these small spicules occur. They are
also met with in small numbers in the sterraster-armour layer. The low dense
fur, referred to above, is entirely composed of these spicules. In the spicule-
fur of var. intermedia and also in that covering the parts of the surface with thick
cortex in var. micraster the styles predominate greatly, only a few minute ana-
clades being here scattered between the dense masses of small styles. In the
low spicule-fur covering the parts of the surface of var. micraster which lie at a
lower level and below which the sterraster-armour is thin, the minute anaclades
are very abundant, more numerous than the small styles (Plate 18, fig. 24a).*
The spicules forming this low dense fur protrude for the greater part of their
length beyond the surface of the sponge; the pointed ends of the small styles
and the cladomes of the minute anaclades being situated distally and free, the
rounded ends of the styles and the acladomal ends of the minute anaclades
being situated proximally and implanted in the sponge. The manner in which
these minute dermal styles and anaclades are distributed shows that they are
formed in the distal layer of the choanosome or the proximal layer of the cortex,
that they travel up from this, their place of birth, to the surface, first, up to the
sterraster-armour layer, slowly, then, through the sterraster-armour layer,
rapidly, and that their movement in this distal direction is retarded to a great
extent or quite discontinued on their reaching their final position in the low
spicule-fur, where they remain for a considerable time.

Four kinds of microscleres occur; large oxyasters, smaller oxysphaerasters,
small strongylosphaerasters, and sterrasters. The large oxyasters are confined
to the choanosome. They are not uniformly distributed; in the distal zone of

GEODIA VARIOSPICULOSA. 59

- the choanosome these asters are very numerous; in parts of the interior they
are scarce. The oxysphaerasters are very numerous in the walls of the cortical
canals, zones exceedingly rich in them indicating the position of these canals
(Plate 18, figs. 21b, 23a, 26b). In the walls of the dermal canals these oxy-
sphaerasters extend right up to within a short distance of the surface, and they
are also met with in the walls, chiefly the roofs, of the subcortical cavities. A
few are scattered between the sterrasters: The small strongylosphaerasters
form a dense layer on the outer surface which increases in thickness in the pits
where the dermal membrane itself is thickened, and are also scattered through-
out the cortex. The sterrasters occur in the sterraster-armour, and are also
scattered in the choanosome. In all parts of the sterraster-armour layer, with
the exception of the thin portions of it in var. micraster, the sterrasters are rather
densely packed, in the thin parts of the cortex of the var. micraster they are
farther apart. The sterrasters in the choanosome are mostly young forms.

The large choanosomal amphioxes (Plate 17, fig. 42) are usually curved,
often in an irregular wavy manner, gradually attenuated to the rather sharp-
pointed ends, and isoactine or — as the one represented in the figure — slightly
anisoactine. In var. micraster they are 2.3-3.9 mm. long and 25-42 yp, usually
32-37 » thick; in var. intermedia considerably stouter, 2.5-3.1 mm. long and
42-50 yw, usually 42-44 » thick.

The large choanosomal tylostyles and styles (Plate 17, figs. 33, 43) are usually
‘curved. One of var. micraster which was intact measured 1.35 mm. in length.
These spicules gradually increase in thickness towards the rounded or tyle end.
Just below this they are in var. micraster 40-50 4, in var. intermedia 25-35
thick. Only a few of these spicules are true styles, in most the rounded end is
thickened more or less, sometimes so much so that the tyle is twice the
diameter of the shaft. But however great this thickening may be, it is never
sharply defined and passes gradually into the shaft, so that these spicules
appear more or less club shaped. The thickened end (tyle) measures in var.
micraster 46-62 » and in var. intermedia 30-70 / in diameter.

The small dermal styles (Plate 19, figs. 4, 5) are fairly straight or slightly
curved, nearly cylindrical in the central part, and gradually attenuated towards
both ends. Distally these spicules nearly always terminate in a sharp point.
Very rarely the distal end is rounded and blunt. The thickness of the proximal
end, which is always rounded off, is from 25-67 % of the maximum thickness.
The small styles with a particularly thin proximal, rounded end are amphiox-
like. Many of these spicules are slightly thickened locally at a point nearer

60 GEODIA VARIOSPICULOSA.

the rounded than the pointed end. I presume that this thickening is situated
at the point where the spicule penetrates the surface and that it is formed after
the spicule has taken up its definite position in the low spicule-fur. In var.
micraster these small dermal styles are 210-320 long, 3-7 » thick in the centre,
and 1-3.5 « thick at the proximal, rounded end. In var. intermedia they are
somewhat shorter, 200-310 » long, 3.5-7 « thick in the centre, and 1-2.5 p thick
at the proximal, rounded end.

A concrescent form of these spicules, consisting of three or four straight
rays, some pointed, others rounded at the end, is very rarely met with.

The large orthoplagiotriaenes (Plate 17, figs. 34-37, 38a, 44-47, 49) have
generally a straight, or slightly curved, conical rhabdome, pointed at the
acladomal end, more or less thickened at a distance of about 150 » below the
cladomal end, and considerably constricted above this thickest point, Just below
the cladome. In the neck-like subeladomal constriction the thickness of the
rhabdome is from 65-85 % of the thickness of its thickest part. At this point,
weakened as it is by the constriction, the rhabdome readily breaks, and parts
of these spicules (cladomes and rhabdomes) broken at that point are frequently
found in the spicule-preparations. The chords of the clades form angles of
99°-111° with the axis of the rhabdome. These spicules are consequently
intermediate between orthotriaenes and plagiotriaenes, and are named ortho-
plagiotriaenes accordingly. The clades are conical, not sharply pointed, fre-
quently quite blunt. They always arise in an ascending direction. Their
proximal part is concave to the rhabdome, their distal part straight (Plate 17,
figs. 34-37, 47) or slightly undulating (Plate 17, figs. 44, 45). A slight, abrupt,
angular bend is frequently observed at the point where the proximal part,
concave to the rhabdome, passes into the distal straight or undulating part.
This and a slight thickening of the axial thread of the clade often observed at
this point seem to indicate that these orthoplagiotriaenes are dichotriaene-
derivates. The three clades of the same cladome usually arise at similar angles
from the rhabdome; forms like the one represented, Plate 17, fig. 38a, in which
the rhabdome angles of the three clades differ to a greater extent, being rare.
In length the clades of the same cladome may be equal or unequal. The forms
with unequal clades are as numerous as the ones with equal clades, if not
more so. Most of the cladomes composed of unequal clades are sagittal, two
clades being fairly equal, while the third is very much shorter, only one half or
a third as long, as the other two. The angles between the clades are independ-
ent of the clade-length and nearly always about 120°. Rather frequently a

GEODIA VARIOSPICULOSA. 61

neck-like constriction, similar to the one at the cladomal end of the rhabdome,
is observed at the proximal end of the clades. In var. micraster the rhabdomes
of the orthoplagiotriaenes are 2.6-3 mm. long, and in the neck-like constriction
at the cladome 30-50 » thick, their maximum thickness below this constriction
being 43-64 «. The chords of their clades are 240-760 » long and enclose
angles of 99-105° with the axis of the rhabdome. In var. intermedia the
rhabdomes of these spicules are 2.4-3 mm. long, and in the neck-like con-
striction at the cladome 35-65 » thick, their maximum thickness below this
constriction being 50-75 ». The chords of the clades are 220-550 » long and
enclose angles of 100-111° with the axis of the rhabdome.

Besides these normal orthoplagiotriaenes, irregular forms are met with in
small numbers. In some of these one or two clades are bifurcate: these spicules
are transitional to the dichotriaenes. In others one or two of the clades are
cylindrical and not attenuated distally, or angularly recurved at the end. An
irregular triaene with one cylindrical and one terminally angularly recurved
clade is shown in Plate 17, fig. 46. In examining the rounded end of these ab-
normal cylindrical clades with the highest powers, I found the distal part of
the axial thread irregularly thickened at frequent intervals. At the end itself
the axial thread appeared split up into a bunch of very slender, thread-like,
divergent branches, which seemed to extend right up to the rounded, terminal
face of the clade. It seems that this terminal face is clothed with numerous
exceedingly small spines, and that one of the terminal branches of the axial
thread leads up to each of these spines. But as these structures are, in con-
sequence of their exceedingly small size, on the verge of visibility, even with
the ultraviolet light (A = 280 ») employed, I could not make them out with
any degree of certainty.

The dichotriaenes (Plate 17, figs. 88b, 48, 50), apart from the bifurcation of
the clades and the smaller size of the clade-rhabdome angles, are similar to the
orthoplagiotriaenes described above. In var. micraster their rhabdomes are
1.7-2.6 mm. long and in the neck-like constriction at the cladome 30-40 » thick;
their maximum thickness is 45-58 ». Their clade-rhabdome angles are little
over 90°. The clade-stems are 160-340 », the clade-branches 140-400 # long.
In the preparations of var. infermedia no intact dichotriaene-rhabdomes were
found, so that their length is unknown. Their thickness is 45-48 « at the
neck-like constriction at the cladome, and 55-60 » at the thickest point.
The clade-rhabdome angles are a little over 90°. The clade-stems are 150-160 p,
the clade-branches 170-280 y long.

62 GEODIA VARIOSPICULOSA.

The mesoproclades observed were nearly all mesoprotriaenes. I found only
a single mesopromonaene.

In the mesoprotriaenes (Plate 17, fig. 32) the epirhabd is straight, conical,
and usually shorter than the clades. The latter are conic, pointed, and curved,
concave to the epirhabd. Their curvature increases distally. This distal
inerease of curvature is most marked in the mesoprotriaenes with strongly
diverging clades (large clade-epirhabd angles). In the preparations of var.
micraster no intact mesoprotriaene-rhabdomes were found, so that I can not state
their length. The thickness of the rhabdome at the cladome is 7-20 4; the
epirhabd is 25-90 » long; the clade-chords are 40-140 # long and enclose angles
of 30-63° with the axis of the epirhabd. In var. intermedia the rhabdomes of
these spicules are 2.9-3.2 mm. long and at the cladome 18-23 » thick. The
epirhabd is 60-75 «long; the clade-chords are 100-142 » long and enclose angles
of 37—44° with the axis of the epirhabd.

Besides these normal mesoprotriaenes a few with one or two stunted
cylindrical clades, rounded off at the end, have been observed. One of these is
represented in Plate 17, fig. 32.

In the preparations of var. micraster | found one mesopromonaene with a
broken rhabdome, 30 » thick at the cladome. Its epirhabd is 180 y long, its
clade is recurved in a hook-like manner at the end and slightly concave to the
epirhabd. The clade-chord is 175 » long and encloses an angle of 35° with the
axis of the epirhabd.

The large anatriaenes (Plate 17, figs. 23-31; Plate 18, fig. 8d), when fully
and normally developed, have a long rhabdome, thickened towards the cladome,
and three fairly equal, conical and uniformly curved, pointed clades (Plate 17,
figs. 25, 26, 29, 30). Sometimes the clades are bent down abruptly at their
ends (Plate 17, fig. 27). The apex of the cladome is simply rounded or, more
rarely, crowned by a very slight protuberance. In var. micraster the rhabdome
is 3.6-5.2 mm. long and at the cladome 20-46 » thick. The clade-chords are
65-130 » long and enclose angles of 37°-70° with the axis of the rhabdome.
In var. intermedia I found only one intact rhabdome, 4.2 mm. in length. In
this variety the rhabdomes are 20-40 » thick at the cladome and the clade-
chords 50-135 « long. The clade-rhabdome angle is considerably smaller than
in the other variety, measuring only 35-48°.

desides these normal large anatriaenes four other, abnormal or derivate
forms are observed: 1, regular anatriaenes with pointed clades but of much
smaller dimensions; 2, irregular anatriaenes of normal dimensions and clade-

GEODIA VARIOSPICULOSA. 63

position, in which one, two, or all three clades are stunted, short, and rounded
at the end; 3, anatriaenes of normal dimensions with two simple and one bifur-
cate clade; and 4,-anatriaene-derivates of normal dimensions, in which one of
the clades is directed upward, proclade-fashion. (1) is very, and (2) fairly
abundant in both varieties; (3) and (4) are very rare and have been found
only in var. intermedia. In the small anatriaenes with pointed clades (Plate
17, figs. 23, 28) the rhabdome is only 12-20 » thick at the cladome, the clades
being 30-50 » long in var. intermedia, and 30-65 » long in var. micraster. Of
course these anatriaenes may be young forms of the normal ones. Their
abundance on and close to the surface, however, renders this assumption some-
what doubtful. The anatriaenes with stunted clades (Plate 17, fig. 31) exhibit
very different degrees of clade-reduction. In most of them only one or two
clades are shortened and rounded; in some, however, one or two clades are
reduced to low, rounded protuberances and the others (other) shortened to half
or less than half of the normal length. These extreme forms have been met
with chiefly in var. intermedia. In the few anatriaenes with one bifurcate clade
the two other (simple) clades were more or less stunted. In the anatriaene-
derivates with one clade directed upwards the clades are pointed, but much
shorter than in the normal anatriaenes.

The minute dermal anaclades (Plate 18, fig. 24a; Plate 19, figs. 3, 6-10, 14)
are mostly anatriaenes with well-developed clades. A few of them have, how-
ever, by a more or less complete clade-reduction become anadiaenes, anamo-
naenes (Plate 19, fig. 6), or even tylostyles. These latter are, however, very
rare. The rhabdome is more or less curved, simply or in an S-shaped man-
ner, and thickest at a point from a fifth to a third of its length above its acla-
domal end. From this thickest point it is gradually attenuated towards both
ends. The acladomal end is rounded. The thickness of the two ends of the rhabd-
ome is from 25-60 % of its maximum thickness. In many of these spicules a
slight local thickening of the rhabdome, situated nearer the acladomal than the
cladomal end, similar and probably analogous to the local thickening of the
small dermal styles, has been observed. The apex of the cladome is simply
rounded off or crowned by a protuberance (Plate 19, fig. 9). In var. micraster
the minute anaclades without apical protuberance greatly predominate, in var.
intermedia a much greater proportion of them possess such a protuberance.
The clades are conic, sharp pointed, strongly recurved in their basal part, and
nearly straight in their terminal part (Plate 19, figs. 3, 14). In var. micraster
the rhabdomes of the anaclades are 275-410 « long, 1.5-4 y# thick at the

64 GEODIA VARIOSPICULOSA.

eladomal end, 4-6 » thick at the thickest point, and 1.5-2 # thick at the
rounded acladomal end. The chords of the clades are 5-12 » long and enclose
angles of 38°-54° with the axis of the rhabdome. In var. intermedia the
rhabdomes of these spicules are 205-560 » long, 1-3.5 thick at the cladomal
end, 3-7.5 « thick at the thickest point, and 1-4.5 # thick at the rounded,
acladomal end. The clade-chords are 3-13 long and enclose angles of 40°
52° with the axis of the rhabdome.

The large choanosomal oxyasters (Plate 18, fig. 1, 2a, 3, 4, 5d, 6, 7b, 10a, b,
12. 14b, 15-20, 22a, b, 25a, b, 27a, b; Plate 19, figs. 25-30) have no central
thickening and are composed of fairly concentric but often not quite regularly
distributed rays. The rays are straight, conical, usually very blunt, truncate,
rarely pointed, and everywhere, except at the proximal (central) end, covered
with spines. The size of these spines is variable. Sometimes they are so
small as to be hardly discernible, sometimes they are large, 1 » or more long.
When large enough to be distinctly seen, they show an increase in size from the
base to the tip of the ray. These spines rise vertically from the ray and appear
to be bent back at the end towards the centre of the aster in a claw-shaped
manner. There are usually from one to eight rays. In the form where only
one ray is developed, short, rounded, knob-like rudiments of two or three other,
reduced rays are observed (Plate 18, figs. 5d, 6; Plate 19, fig. 28). Such ray-
rudiments also occur in most of the diactine and in some of the triactine forms.
The terminal rounded faces of these ray-rudiments are densely covered with
large spines (Plate 19, fig. 28). The monactine oxyasters appear as blunt
tylostyles with irregularly lobate tyles. In the diactine forms the fully de-
veloped rays seem never to extend in a straight line, the angle enclosed by
them being 60°-140°, always much less than 180°. Thus these spicules appear
as more or less widely opened compasses. The three rays of the triactine
oxyasters may be situated in a plane and regularly arranged so as to enclose
angles of 120° (Plate 18, fig. 10a); or they may not be situated in a plane and
form the edges of a low triangular pyramid which may be regular (Plate 18,
fig. 2a) or irregular (Plate 18, fig. 16); or finally they may be so arranged
that two lie in a straight line to which the third is vertical or oblique (Plate
18, fig. 4). Among the asters with from four to eight rays both regular forms
with fairly equal angular intervals between the rays (Plate 18, figs. 1, 15, 19),
and irregular forms with unequal angular ray-intervals (Plate 18, figs. 17, 18,
20) are met with. In these tetr- to oct-actine asters the rays are, in the same
aster, usually of equal size. The tri- to hex-actine asters are far more numerous
than those with one or two, or seven or more rays.

GEODIA VARIOSPICULOSA. 65

The number of rays is, as shown by the following table, roughly speaking,
in inverse proportion to their size.

NUMBER AND SIZE OF RAYS OF OXYASTERS.

Variety Number of rays Zone diameter | Length of rays ich ues CerayS
micraster 1-4 40-132 pt 25-72 ye 1.5-8 p
intermedia 1-4 27-180 » =f 14-90 p 1-8
micraster 5-6 21-105 11-58 ps 1.5-7 pu
intermedia 5-6 20-140 | 10-78 4 1-7 as r
micraster FS VASIae Ne | 916 SLAY)
intermedia 7-8 25-37 | 13-21 4 1-3 u

Oxyasters with from nine to eleven rays are very rare. The largest of them
observed was 23 in total diameter and had rays 12 » long and 2 » thick at the
base.

The orysphaerasters (Plate 18, fig. 23; Plate 19, figs. 12a, 13a) are composed
of a spherical centrum and numerous regularly distributed, concentric, radial
rays. The rays are in the same aster of equal size, conical, sharp pointed, and
sparsely spined. Often a verticil of larger spines, situated some distance below
the end of the ray is observed. The oxysphaerasters of var. micraster usually
have from eighteen to twenty-two rays. Their total diameter is 14-19 y, the
diameter of the centrum 5-6 ». The rays are 4-7 « long and 1-2 y thick at the
base. The oxysphaerasters of var. intermedia usually have from fourteen to
twenty-two rays. Their total diameter is 17-22 y, the diameter of the cen-
trum 5-6». The rays are 5-8 « long and 1.5-2 » thick at the base. A cor-
relation between ray-number and spicule-size is not apparent.

The small dermal strongylosphaerasters (Plate 19, figs. 12b, 13b, 19-24)
consist of a spherical or irregularly tuberous centrum and numerous short rays.
The rays are usually stout, cylindroconical, or cylindrical, and truncate, rarely
more slender, conical, and blunt pointed. Not infrequently the rays of the same
aster are unequally distributed and unequal in size, one of the rays being some-
times fully twice as large as any of the others. The distal parts of the rays,
chiefly their terminal faces, bear numerous small spines. In var. micraster the
strongylosphaerasters usually have from ten to eighteen rays. Their total
diameter is 5-8 1, the diameter of the centrum 2-4 #. The rays are 0.5-2.5 44

66 GEODIA VARIOSPICULOSA.

long and 0.6-2 : thick at the base. In var. intermedia these asters usually have
from twelve to nineteen rays. Their total diameter is 5-7 p, the diameter of
the centrum 2-4 yp. Their rays are 0.6-2 « long and 0.5-1.6 »: thick at the base.
A correlation between ray-number and spicule-size is not apparent. Occasion-
ally irregular tuberous structures with spiny surface, which I consider as derivates
of these spicules with reduced rays, have been observed. One of these, which
I found in var. micraster had the shape of a stout, short, slightly curved sausage
and was 9 «long and 4 « broad.

The sterrasters (Plate 19, figs. 1, 2, 11, 15-18, 31, 32) are flattened ellipsoids.
The proportion of length to breadth to thickness is on an average 100 :82 : 65.
In the normal sterrasters the distal parts of most of the rays have a circular or
rather regularly polygonal transverse section, are 2-3 . thick, and bear a terminal
verticil of usually from four to six stout lateral spines (Plate 19, figs. 1,2). The
transverse sections of the distal parts of the rays which surround the umbilicus
are elongated in a direction radial to the centre of the latter, 2-3 « broad, and
about 5 » long. These umbilical rays bear from six to eight or more spines,
one of which is often (Plate 19, fig. 1) considerably larger than the others.
This larger spine is directed towards the centre of the umbilicus. In the centre
of young sterrasters, 22-26 » in diameter, of both varieties, a spherical cluster,
4-5 «in diameter, of numerous radially arranged, elongated, somewhat rod-
shaped granules, are observed. The sterrasters of var. micraster are 120-133
long, 100-116 » broad, and 82-90 » thick. Those of var. intermedia are 109—
125 » long, 90-100 broad, and 70-75 4 thick.

Besides these normal sterrasters, a few abnormal ones, similar in regard to
shape and size, but very different in regard to the structure of the surface, have
been observed. Two kinds of such sterroids can be distinguished. In one of
these the rays are much thicker than in the normal sterrasters. In the other
the rays are of the same thickness as in the normal sterrasters but destitute
of spines. A sterroid of the first kind is represented in Plate 19, figs. 17, 18.
In this sterraster the ends of the rays have more or less polygonal terminal
faces, 10-20 # in diameter, densely covered with spines. A sterroid of the
second kind is shown in Plate 19, fig. 15, 16.

The specimen of var. intermedia was caught with the tangles at Station
3746 on May 19, 1900, off Honshu Island, Japan, Suno Saki N. 87°, E. 15.8 km.
(8.5 miles); depth 90 m. (49 f.); it grew on a bottom of gray sand and pebbles.
The specimen of var. micraster was trawled at Station 3758 on May 22, 1900,
off Honshu Island, Japan, Suno Saki 8. 55°, E. 3.9 km. (2.1 miles); depth
95-135 m. (73; 52 f.); it grew on a bottom of black clay and rock.

.GEODIA VARIOSPICULOSA. 67

As the description given above shows, these sponges are so similar to those
described by Thiele as Geodia variospiculosa (Studien tiber pazifische spongien.
Zoologica, 1898, 24, p. 10, taf. 6, fig. 6) and variospiculosa var. clavigera (Thiele,
loc. cit., p. 11, taf. 6, fig. 7), which also came from Japan, that I do not hesitate
to assign them to this species. Still, they differ from Thiele’s description and
also from each other in so many respects, that the question arises whether the
peculiarities wherein they differ are germinal in nature and systematically
important or merely due to differences of germ-separation or mixture before
and during fertilization, age, or individual adaptation to different conditions,
and systematically unimportant. Thiele describes the small dermal rhabds
of G. variospiculosa and of G. v. var. clavigera as amphioxes, while the corre-
sponding spicules of the “Albatross” sponges are styles. Since, however, these
styles are attenuated towards both ends and consequently similar to amphioxes,
Thiele might easily have designated the small dermal rhabds of Geodia vario-
spiculosa as amphioxes even if they had exactly the same shape as the spicules
here described as styles. The same applies to the large subcortical ortho-
plagiotriaenes, which Thiele terms orthotriaenes. According to one of Thiele’s
figures (loc. cit., taf. 6, fig. 6b) the ‘‘orthotriaenes” of his Geodia variospiculosa
are orthoplagiotriaenes in my sense. Apart from these apparent rather than
real differences, there are, however, differences in the dimensions of the spicules,
many of which are very considerable.

The specimen described by Thiele as var. clavigera is the smallest of the four.
The typical Geodia variospiculosa is larger, var. intermedia still larger, and var.
micraster by far the largest. If we were to assume that these differences in size
are due to differences of age, it would be only natural to suppose that correspond-
ing spicule-dimensions should be smallest in var. clavigera, larger in the typical
Geodia variospiculosa, still larger in var. intermedia, and largest in var. micraster.
All dimensional differences which accord with this must therefore be set aside
in studying the relative systematic position of these sponges. After eliminating
these differences possibly due to differences of age, there remain the following:
the large amphioxes and the rhabdomes of the orthoplagiotriaenes and the
dichotriaenes of the smaller var. intermedia are thicker than those of the larger
var. micraster. The dichotriaene-rhabdomes of the still smaller typical Geodia
variospiculosa are thicker than those of both the larger varieties intermedia and
micraster. The choanosomal tylostyles of var. intermedia have relatively larger
tyles than those of var. micraster. Long dermal tylostyles are present in var.
clavigera but absent in the three others. The mesoprotriaene-clades are rela-

6S GEODIA VARIOSPICULOSA.

tively and, to a smaller degree, also absolutely longer in the smaller var. inter-
media than in the larger var. micraster. The clades of the larger anatriaenes are
longest and the clade-rhabdome angles smallest in the small typical Geodia
variospiculosa; the former are smaller and the latter larger in the larger var.
intermedia, and the former still smaller and the latter still larger in the still
larger var. micraster. On comparing the figures 23-26 and 28-31 on Plate 17,
with each other and with Thiele’s figure (loc. cit. Plate 6, fig. 6e) the differences
in the appearance of the cladomes of the large anatriaenes caused by these
differences in the clade-length and clade-rhabdome angle will be noticed. The
minute dermal anaclades of the smaller var. intermedia are larger than those of
the larger var. micrasler. The size of the choanosomal oxyasters of the typical
Geodia variospiculosa and the two varieties intermedia and micraster is, like the
length of the clades of the large anatriaenes, in inverse proportion to the size of
the specimen in which they occur. Besides these differences there are others,
in the structure of the surface, the size of the pores, the relative frequency
of monactine asters and anatriaene-cladomes with stunted, rudimentary clades,
etc.

Some of these differences, but hardly all of them, may be due to differences
in the forces acting on the different individuals. In particular I should say that
the differing peculiarities in the shape and size of the cladomes of the large
anatriaenes and in the size of the oxyasters, should be considered as germinal
and therefore systematically important. The reason why these differences of
the anatriaenes and oxyasters should be considered as due to germinal peculi-
arities are: in the tetraxon sponges studied in this respect’ the anatriaenes
of young (small) specimens have not only shorter clades but also larger clade-
rhabdome angles than those of older (larger) specimens, the clade-length increas-
ing and the clade-rhabdome angle decreasing with the age of the sponge. In
the sponges here under discussion, inversely, the clade-length decreases and the
clade-rhabdome angle increases with the size of the sponge. The size of the
oxyasters is in these sponges in inverse, the size of the sterrasters in true propor-
tion to the size of the specimens. The former lie in the interior of the sponge
and must therefore be less influenced by external forces than the latter which
lie near the surface. Differences of the internal oxyasters, not also seen in the
external sterrasters, cannot therefore, in my opinion, be ascribed to differences
of the forces acting on the growing sponge. This view is further supported by

‘Por instanee in Cinachyra vertex, see R. vy. Lendenfeld, Tetraxonia. Ergeb, Deutsche Stidpo-
lar-Expedition, 1901-1903, 1907, 9, p. 318.

GEODIA VARIOSPICULOSA. 69

the fact that in var. intermedia and var. micraster the size of the dermal strongy-
losphaerasters is also in true proportion to the size of the sponge.

Formerly (Tetraxonia. Tierreich, 1903, 19, p. 107) I united var. clavigera
with the typical Geodia variospiculosa, but now, having been able to examine
sponges belonging to this species I think it better to keep these two sys-
tematically apart. In view of the above discussion on the germinal nature
and systematic importance of the differences between the two “Albatross”
specimens here described and between them and Thiele’s sponges referred to
above, I establish varieties for them. Thus Geodia variospiculosa Thiele is
divided into four varieties: the typical Geodia variospiculosa, for which I propose
the name var. typica, var. clavigera Thiele, var. intermedia, and var. micraster.

DIMENSIONS OF THE SPICULES OF GEODIA VARIOSPICULOSA THIELE.

Var. Var. Var. Var.
typica clavigera intermedia micraster
length mm. over 2 11 2.5-3.1 2.3-3.9
Large choanosomal amphioxes | am Sane = i
; 2 26%! Dar 95-42
thickness p 20 2b vail A
length mm. 1.35
diameter of
Large choanosomal tylostyles | tyle (rounded 30-70 46-62
(styles) 7 end)
thickness of 25-35 40-50
shaft
; length mm. 1.6
diameter of . 1S* |
Large dermal tylostyles tyle absent absent absent
thickness of 11*
shaft
lene 200 200-310 | 210-320
? maximum 4 3.5-7 37
Minute dermal styles thickness 1
thickness of |
rounded end | 2-5 1-3.5

! The dimensions marked * are taken from Thiele’s figures.

70

GEODIA VARIOSPICULOSA.

DIMENSIONS OF THE SPICULES OF GEODIA VARIOSPICULOSA THIELE (continued).

Var. Var. Var. Var.
typica clavigera intermedia micraster
‘length mm. — | 2 1.25 2.43 2.6-3
| thickness at | 35-65 30-50
| rhabdome | cladome 4
ae {reece Fae ||
Orthoplagio- | maximum 70 50-75 43-64
triaenes thickness
clades length «| 400-460 250-300 220-550 240-760
|angle between clade-chords 103* 100-111 99-105
and rhabdome °
length mm. 2 1.25 1.72.6
j— =
thickness at 75* 45-48 3
rhabdome cladome o)
| maximum a 55-60 45-58
Dichotriaenes | | thickness 4
main clade length 1 200* 150-160 160-340
end clade length 300* 170-280 140-400
angle between main clade- 7 a little a little
chords and rhabdome ° 90 over 90 over 90
length mm. 2.5 2.9-3.2
rhabdome F
thickness at
cladome p 12* 18-23 7-20
OE epirhabd length p 95 30-60 60-75 25-90
aenes
clades length 220 60-70 100-142 40-140
angle between clade-chords
| and epirhabd ° 40% Sie 30-63
length mm. 4.2 3.6-5.2
rhabdome :
thickness at
20* 12-40 12-4
Large anatri- cladome jt .
ene clades length » | 160-180 30-40 30-135 30-130-
angle between clade-chords ae Toe
and rhabdome ° 28-42 35-48 37-70
| length 1 286* 205-560 275-410
thickness at *
rhabdome cladome : 1-3.5 1.54
Minute dermal maximum
anaclades | thickness 1 ce 3-(-6 +6
clades length 5 3-13 5-12
angle between clade-chords 50* 40-52 38-54

and rhabdome °

‘The dimensions marked * are taken from Thiele’s figures.

GEODIA VARIOSPICULOSA. Gall

DIMENSIONS OF THE SPICULES OF GEODIA VARIOSPICULOSA THIELE (continued).

Var. Var. Var. Var.
typica clavigera intermedia micraster
number of rays 3—-6* 1-8 1-11
total diameter
i 156-164* 27-180 40-132
dimensions of length p | 84*-135 ele See
oxyasters with
1-4 rays
» rays | basal
thick- 8* 1-8 1.5-8
ness jt
total diameter
r 40—104* 20-140 21-105
Choanosomal : : bs
oxyasters | dimensions of length »| 52-56* 10-78 11-58
oxyasters with ie
5-6 r. :
a rays | basal
thiek- e7/ 1.5-7
ness
a i t
a diame ter as a
dimensions of fener 13-21 9-16
oxyasters with
7-8 rays
4 rays | basal
thick- 1-3 1-3
ness
number of rays 14-22 18-22
total diameter 16*-30 17-22 14-19
eet pase ras: diameter of centrum 5-6 5-6
ters i -
| length 5-8 » 47
rays
thickness 1.5-2 1-2
number of rays 12-19 | 10-18
total diameter 6-8 5-7 5-8
Dermal stron- = = a :
gylosphaer- | diameter of centrum 2-4 2-4
aster = 7 = —
length ps 0.6-2 0.5-2.5
rays =
thickness 0.5-1.6 0.62
length 2 80-115 100 109-125 | 120-133
Sterrasters breadth 65-95 80 90-100 100-116
thickness 70-75 82-90

ae :
1 The dimensions marked * are taken from Thiele’s figures.

GEODIA JAPONICA.

~I
lo

Geodia japonica (SouLas).
Plate 37, figs. 15-30; Plate 38, figs. 1-29; Plate 39, figs. 1-41.

Thiele, Zoologica, 1898, 24, p. 7, taf. 2, fig. 1; taf. 6, fig. 3. Lendenfeld, Tierreich, 1903, 19, p. 111.
Cydonium japonicum Souas, Rept. voy. “Challenger,” 1888, 25, p. 256.

There is in the collection a fine, dry sponge from Japan, which appears
to be identical with the species described by Sollas as Cydoniwm japonicum and
by Thiele as Geodia japonica. A part of the type of the former in the British
Museum I have, through the kindness of Mr. Kirkpatrick, been able to
reexamine. , ‘

The sponge (Plate 38, fig. 8) has the shape of a low and broad, thick-walled
cup. It is 19 em. high; its maximum and minimum transverse diameters are
24 and 22cm. Near its margin the wall of the cup is about 3 em. thick; the
@margin itself is bent inward; it is interrupted in two places by broad indentures.
The cavity of the cup is 83 cm. deep. The base of the sponge measures 11 X13
em. and exhibits a cup-shaped depression like the upper one, but much smaller,
only 44 em. deep. The inner surfaces of both the upper and the lower cup are
rather irregular and undulating but destitute of higher protuberances. The
outer surface of the sponge is covered by large and conspicuous, terminally
rounded, lobose protuberances, 13-3 em. broad and about as long, which hang
down stalactite-fashion. Most of them are attached with a considerable portion
of one side to the body of the sponge. In external appearance and in size it
corresponds with the specimens examined by Sollas and Thiele.

In some of the narrowest, most sheltered fissures between adjacent stalactite
lobes a rather dense spicule-fur about 1.5 mm. high is observed. Apart from
this the surface is now entirely destitute of a spicule-fur. The whole of the
surface, also that of the margin of the cup, is dotted with small holes, the en-
trances to the radial cortical canals. Some of these holes are partly covered
by remnants of pore-sieves. In the specimen examined by Sollas there were no
holes (canal-entrances) on the margin of the cup. On the outer, lobose surface
the width of and the distances between these holes are quite constant, their
diameter here being about 300 p and the distances between their centres about
700 w. On the margin of the upper cup the holes are much smaller and farther
apart. In this upper cup the holes are more variable in size and much less
regularly distributed than on the outer surface, some being as much as 500
wide. In extensive tracts of the lower cup, these holes are 400-450 wide and
the distances between the margins of adjacent ones smaller than their diameter.

The colour (in the dry state) is white. |

a YT

GEODIA JAPONICA. 73

The sterraster-armour layer of the cortex is a little over 0.5 mm. thick.
Sollas gives the thickness of the cortex as 0.8 mm., Thiele the thickness of the
sterraster-armour layer as 0.3 mm. According to my measurement it is in
Sollas’s type about 0.65 mm. thick.

Skeleton. In the interior of the choanosome numerous large, stout amphi-
oxes are found, also less numerous large, slender amphioxes, a few large styles,
numerous large oxyasters, numerous small strongylosphaerasters, and very
few sterrasters. In the superficial (subcortical) part of the choanosome the
same three kinds of large choanosomal rhabds, numerous orthotriaenes, a few
plagiotriaenes, many large anatriaenes, various promesoclades, numerous
minute dermal amphioxes, a few minute dermal styles and minute dermal ana-
clades, some large oxyasters, and a good many large oxysphaerasters and small
strongylosphaerasters occur. In this region the large rhabds and the rhabdomes
of the large telo- and mesoclades form radial bundles. The minute dermal
amphioxes, styles, and anaclades are situated radially or obliquely, and scattered ;
the asters chiefly occupy the canal-walls. The sterraster-armour of the cortex
is composed of dense masses of sterrasters. Where the surface has not been
rubbed off or otherwise injured numerous minute amphioxes and some minute
dermal styles and minute dermal anaclades, implanted in the distal part of the
sterraster-armour, and an outer covering of small strongylosphaerasters are
observed. Large spicules situated radially and broken off distally also occur
occasionally in this region. These are probably rhabdomes of mesoclades and
perhaps also of large anatriaenes similar to the ones found intact in the subcor-
tical layer. The spicule-fur is composed of anatriaenes and mesoproclades.
The large slender amphioxes, the large styles, the mesoproclades, the minute
dermal styles, and the minute dermal anaclades are not mentioned either by
Sollas or by Thiele. The latter also failed to find any large oxysphaerasters but,
on the other hand, observed sphaeres, which were not found in the specimens
examined by Sollas and by myself. In the type of Sollas reexamined by me,
some large styles and a good many minute dermal rhabds with one end very
blunt and more or less style in character were met with. Mesoproclades and
minute dermal anaclades appear, however, to be absent.

The large stout amphioxes (Plate 37, figs. 18-21, 22a) are generally simple,
straight or slightly curved, and rather abruptly pointed. They are 2.4-3.2 mm.
long and 30-51 » thick. According to Sollas they measure 2.7 mm. by 32 yp,
according to Thiele 2 mm. and over by 35-40 p. In the type of Sollas reexamined
by me these amphioxes were 2.43.3, mostly 2.8-3.2 mm. long and 32-50 » thick.

74 GEODIA JAPONICA.

It is known that sometimes the amphioxes of tetraxonid sponges are bifur-
cate at one end. Such forms are, however, very rare and usually considered
pathologic. In the specimen of Geodia japonica examined by me large stout
amphioxes, bifurcate at one end, are remarkably frequent. I found no less
than six in one spicule-preparation. The two branches of these bifureate amphi-
oxes are equal or unequal and 10-80 » long. The angle they enclose is always
small; the largest observed was 12°. When very short the two branches are
nearly parallel to each other and appear as terminal spines.

The rare large styles are 2.1-2.8 mm. long; their maximum thickness is
40-43 ; the rounded end measures 10-31 y, about one third to two thirds of
the maximum thickness in transverse diameter. Such spicules are not men-
tioned either by Sollas or Thiele. I found them, however, in Sollas’s type.

The large slender amphiores (Plate 37, fig. 22d) are strongly and very
irregularly curved, 1-2.2 mm. long and, in the middle, 12-22 y thick. Their
ends are exceedingly slender, thread-like. Though not mentioned by Sollas
or by Thiele, I found such spicules in Sollas’s type.

The minute dermal amphiores (Plate 39, figs. 1-9) are usually somewhat
anisoactine and rather abruptly pointed. They are straight (Plate 39, fig. 9)
or more or less, sometimes very considerably curved, usually in an irregular
manner (Plate 39, figs. 1-8). They are 195-280 y long and, in the middle,
3.5-7 p» thick. In Sollas’s specimen they were small and fusiform. Thiele
gives their length as 300 y. Neither Sollas nor Thiele mentions the remarkable
irregular curvature of somany of them. In the type of Sollas reexamined by me
these spicules are 190-270 y long and 3-6 y» thick. Most of them are more or
less anisoactine, and some strongly and irregularly curved.

The rare minute dermal styles are not so strongly curved as the amphioxes,
somewhat shorter, and in the middle 5.5-6.5 y thick. The rounded end is 2.4—
3.5 #4, about half the maximum thickness of the spicule in transverse diameter.
These spicules can be considered as derivates of the minute dermal amphioxes
in which the anisoactinity has been carried to the extent of the rounding off of
the proximal end. I found such spicules in Sollas’s type though they are not
mentioned by him or by Thiele.

The orthotriaenes and plagiotriaenes (Plate 37, figs. 15-17, 22b, ce, 23-28;
Plate 38, figs. 1-7) have a straight conic rhabdome, 2.3-3.2 mm. long and, at the
cladome, 64-85 y thick. A little below it is usually somewhat thickened and
here measures 65-89 “ in transverse diameter. This thickest part of the rhabd-
ome is, on an average, about 6 % thicker than its cladomal end. The acla-

GEODIA JAPONICA. 75

domal end is usually blunt pointed, rarely rounded and slightly thickened.
The clades are, when normally developed, stout, conic, usually blunt pointed,
and 180-350 long. Their basal part is slightly curved, concave to the rhabd-
ome, their distal part straight. Their chords enclose angles of 90-102°, on an
average 96.4°, with the axis of the rhabdome. The orthotriaene forms (with
clade-angles 90-100°) greatly predominate, the plagiotriaene forms (with clade-
angles over 100°) forming only about 15% of these spicules. The clades
of the same cladome are, in the normally developed triaenes, usually rather
unequal in length. In the normal cladome (Plate 37, fig. 17), the three
clades measured 250, 300, and 340 4. Sometimes one, two, or all three clades
are considerably reduced in length, cylindrical and rounded terminally. Ac-
cording to Sollas these spicules have a rhabdome 2.4-2.6 mm. by 78 » and
clades 230-380 » long; according to Thiele a rhabdome 2.5 mm. by 50-60 p,
clades 200-300 y long, and clade-angles of 92°. In Sollas’s type I found these
spicules had a rhabdome 2.3-2.9 mm. by 50-70 y, clades 200-300 » long, and
clade-angles of 93-100°, on an average 95.5°.

The mesoproclades (Plate 38, figs. 9-17) have a rhabdome 2.8—4.3 mm. long
and 11-21 y» thick at the cladomal end. At the acladomal end the rhabdome
thins out to a slender, irregularly curved thread. The cladomes are variable,
irregular forms with partly reduced clades predominating over the regular
mesoprotriaenes. The epirhabd is usually straight, conical, sharp pointed, and
40-105 long. Sometimes (Plate 38, fig. 16) it is shortened, rounded at the end,
and curved, rarely reduced to a mere knob on the apex of the cladome. The
clades are, when properly developed, conical, pointed and 65-125 plong. Their
basal part is curved, concave to the epirhabd, their distal part straight (Plate 38,
figs. 9, 11, 17) or curved in the opposite direction (outwards) (Plate 38, figs. 12,
15). An abrupt angular bend often intervenes between these two parts. In the
majority of these spicules one or two clades are more or less reduced in length,
cylindrical, and terminally rounded (Plate 38, figs. 15, 17). Sometimes this
reduction has gone so far that one, two, or all three clades appear as mere knobs
(Plate 38, figs. 11, 13, 16) or are altogether absent (Plate 38, fig. 12). The
chords of the clades enclose angles of 22-48°, on an average 34°, with the axis
of the epirhabd. In young mesoproclades (Plate 38, fig. 14) not only are the
epirhabd and the clades shorter but also the clade-epirhabd angles much greater.
Mesoproclades or proclades are not mentioned as occurring in this sponge either
by Sollas or by Thiele, neither could I find any in the type of the former.

The large anatriaenes (Plate 37, figs. 29, 30; Plate 38, figs. 18-29) have a

76 GEODIA JAPONICA.

more or less curved rhabdome, 3.3-4.3 mm. long, and at the cladome, 8-16 4
thick. At its acladomal end it is usually attenuated to a fine, often consider-
ably and irregularly curved thread (Plate 38, figs. 29, 30). Exceptionally the
rhabdome is somewhat shortened and thickened to a tyle at the acladomal
end. The clades arise at an angle of about 80° from the rhabdome. In very
young anatriaenes, such as the one with clades only 20 long represented in
Plate 38, fig. 18, the clades are uniformly curved, concave to the rhabdome
throughout their entire length. This curvature not being great, however, the
chords of the clades of such young anatriaenes enclose angles of over 50° with
the axis of the rhabdome. During the further development the direction of
erowth undergoes a change, the silica being thenceforth apposed to the growing
elade in such manner that its tip becomes a straight, slender, and sharp-pointed
cone, strongly inclined to the rhabdome and enclosing with it an angle of only
about 20°. The further growth of the clade may go on in the same direction:
then clades with straight distal parts are formed (Plate 38, figs. 19-22); or there
may be a continuous change in a direction opposite to that of the curvature
of the proximal part: then clades with distal ends curved outward are formed
(Plate 38, figs. 23-29). This curvature gives, when well pronounced (Plate 38,
figs. 28, 29), a sigmoid appearance to the clades. The anatriaenes with such
sigmoid clades are very characteristic of this sponge. The chords of the clades
of the full-grown anatriaenes are 80-130 long and enclose angles of 23-41°,
on an average 32.2°, with the axis of the rhabdome. The anatriaenes have,
according to Sollas, a rhabdome 18 # thick and clades 100 4 long. According
to Thiele the anatriaene-clades are 70 long. The remarkable outward curva-
ture of the ends of the clades of many of the anatriaenes is indicated in Thiele’s
figure but is not mentioned in the text either by him or by Sollas. In the type
of Sollas examined by me a few anamonaenes of similar dimensions besides the
regular anatriaenes were found. The latter have a rhabdome 2.4-5 mm. by
11-23 p, clades 70-110 » long and clade-angles of 32-45° on an average 36.7°.

The minute dermal anaclades (Plate 39, figs. 13c, 14-17, 38, 39) have a more
or less curved rhabdome, 235-310 4 long and rounded at the acladomal end.
At the cladome the rhabdome is 1-2 y, in its thickest part, which usually lies
below the middle, 2.8-5 », and at the rounded, acladomal end 1.3-3 yp thick.
The proportion of the thickness of the cladomal end to the thickest part to the
acladomal end is, on an average, 10:28:17. I have observed triaene, diaene,
and monaene forms. The clades are always curved, concave to the rhabdome;
their length is very variable. The chords of the clades are 3-10 » long and

GEODIA JAPONICA. qi

enclose angles of 30-54°, on an average 39°, with the axis of the rhabdome.
Such spicules are not mentioned either by Sollas or by Thiele neither could I
on examination find any in the type of Sollas.

The large oxyasters (Plate 39, figs. 13a, 18-26, 27a) usually have no central
thickening, and from three to seven equal, concentric, regularly distributed
rays. The rays are straight, conic, blunt, and everywhere, except quite at the
base, thickly covered with spines. The spines appear to increase in size towards
the end of the ray; those large enough to be clearly made out, are somewhat
recurved, claw shaped. The size of the whole aster and of the rays is in
inverse proportion to the number of the latter.

NUMBER OF RAYS AND DIMENSIONS OF OXYASTERS.

Number of rays Total diameter of asters Length of rays Thickness of rays at base
3-4 30-46 ps 18-23 .5 p 22.8 2
5-6 27-39 15-21 4 1.2-2.7 p
7f 25-31 p 13-16 4 1.3-1.8 p

Occasionally small oxyasters with a distinct centrum and more than seven
rays are observed. These may be considered as transitional between the true
large oxyasters described above and the large oxysphaerasters described below.
According to Sollas the large oxyasters are 32 in diameter. He describes their
rays as smooth. Thiele says that the rays of the oxyasters are few in number
and 6-14 » long. In the type of Sollas I found that the large oxyasters have
from three to seven straight, spined rays. The rays are usually conical and
pointed; very rarely some of them are reduced in length and terminally rounded.
The rays are, at the base, 1.8-2.8 » thick. The total diameter of the aster is
21-36 pu.

The large oxysphaerasters (Plate 39, figs. 27c, to the right, 33, 40, 41) have
a spherical centrum, 5-7.5 y, usually not quite a third of the whole aster, in
diameter, from which from fifteen to twenty-one concentric, regularly arranged
radial rays arise. Usually the rays are equal, rarely one or more reduced in
length and rounded at the end. The normal rays are conical, sharp pointed,
(without the centrum) 5.5-8 « long, and, at the base, 1:9-2.5 thick. They
bear a small number of spines, which are usually restricted to a belt lying
some distance below the end. Sometimes the spines are rather large and then
one perceives that they are slender and directed obliquely outward. More

78 GEODIA JAPONICA.

often they are so small that they can hardly be made out and in some of these
asters the rays seem to be quite smooth. The total diameter of the oxysphaer-
asters is 17-22 ». As stated above transitional forms connect these oxy-
sphaerasters with the oxyasters. Sollas also mentions these asters. He gives
their diameter as 20 «. Thiele did not find any such asters in the specimens
examined by him and expresses the opinion that the ones described by Sollas
were young sterrasters. In the type of Sollas I found these asters fairly abun-
dant; they have a centrum about 5 y in diameter, from sixteen to twenty
conical, sharp-pointed rays, at the base 1.4-2 » thick, and measure 15-21 p in
total diameter.

Most of the small strongylosphaerasters (Plate 39, figs. 10-12, 13b, 27b, 36,
37) are regular, but irregular forms also occur in small numbers. The regular
forms have a spherical centrum, 1.2-5 y, usually about one to two thirds of the
whole aster, in diameter, from which from twelve to twenty-one equal, con-
centric, and regularly arranged radial rays arise. The rays are (without the
centrum) 0.6-2 » long, at the base 0.5-1.2 y thick, and cylindrical. The end is
truncate or rounded. The basal part of the ray is smooth, the end bears spines,
which often form a verticil just below the tip. The total diameter of the regu-
lar small strongylosphaerasters is 4-6, usually 5-5.5 4. In the few-rayed forms
with 12-13 rays the rays are, as a rule, relatively longer and more slender than
in the many-rayed forms with 15-22 rays.

The irregular small strongylosphaerasters have a centrum 1.5-3.6 in diame-
ter, from which from six to nineteen rays arise. These are irregularly arranged,
unequal in length, not always concentric, and usually entirely covered with
spines. The rays are (without the centrum) 1.6-3 » long and, at the base,
0.6-1.3 4 thick. The total diameter of the irregular strongylosphaerasters is
5.8-7.3 pu.

The diameters of the small strongylosphaerasters given by Sollas and Thiele
are 5 w and 4 respectively. In the type of Sollas reexamined by me these
asters have a centrum 1.5-2.2 » in diameter, from sixteen to twenty rays
0.6-0.8 » thick, and measure 4-5.5 y in total diameter.

The sterrasters (Plate 39, figs. 28-32, 34, 35) are flattened ellipsoids 80-89 p
long, 65-78 y» broad, and 55-61 » thick. The average proportion of length to
breadth to thickness is 100: 83:58. The freely protruding rays which surround
the umbilicus have transverse sections elongated in a direction radial to the
centre of the umbilicus, mostly measuring 2.5 x4 y, and bear five or six lateral
spines. The spines directed towards the umbilicus are larger than the others.

GEODIA ATAXASTRA. 79

The other protruding rays, away from the umbilicus, have more or less circular
transverse sections, about 2.5 mm. in diameter, and generally bear four or five
spines. The measurements of the sterrasters given by Sollas and Thiele are
77.5 or 90 by 77.5 by 58 p, and 75 by 65 y respectively. In the type of Sollas
examined by me the sterrasters measure 84-92 by 70-80 by 61 y.

This sponge was labeled Acc. No. 31982, Cydonium japonicum, Japan, and,
as seen from the above description, although very similar to the sponges described
by Sollas as Cydonium japonicum and by Thiele as Geodia japonica in many
respects, appears to differ from them very considerably in others. The most
important of these apparent differences are the presence of large styles, large,
slender amphioxes, minute dermal styles, mesoproclades, and minute dermal
anaclades in the ‘“‘Albatross” specimen and the absence of any reference to
them in the descriptions of Sollas and Thiele. If these differences were real I
should not consider these sponges the same species. Since, however, I have
found large slender amphioxes and styles and also minute dermal styles in the
type of Sollas, the differences due to these spicules not being mentioned in the
descriptions of Sollas and Thiele in reality do not exist. It is different, how-
ever, with the mesoproclades and the minute dermal anaclades, which I failed
to find in Sollas’s type. Since, however, these spicules protrude beyond the sur-
face and are, in much worn specimens, broken off and lost, and since, judging
from the descriptions given by Sollas and Thiele and from the type examined
by me the specimens at their disposal had been much worn, I do not think
their presence in the ‘‘Albatross” specimen and their absence in Cydonium
japonicum Sollas and Geodia japonica Thiele sufficient for systematic distinction.
For this reason and on account of the great similarity of these sponges in every
other respect, I consider them all as belonging to the same species.

Geodia ataxastra, sp. nov.
angustana, var. nov.
Plate 43, figs. 9-25, 28-38; Plate 44, figs. 1-12, 14-49.

latana, var. nov.
Plate 43, figs. 26, 27; Plate 44, fig. 13.

I establish this species for eight specimens in spirit from Perico Island,
Panama. It is characterized by the possession of ataxasters, and to this the
specific name refers. Seven of the specimens are similar and have narrow
anatriaene-cladomes: for these I establish the var. angustana. One is some-
what different and has broad anatriaene-cladomes: for this I establish the new

var. latana.

SO GEODIA ATAXASTRA.

The seven specimens of var. angustana (Plate 44, fig. 25) are attached to a
stone and partly joined. The smallest are quite regularly spherical, the larger
ones more irregular, elongated, divided into lobes, or tuberous. The largest,
which is tuberous, measures 44 X41 X29 mm. Two, which are elongated, have
a length of 46 and 50mm. _ The others are spherical or tuberous and 17-37 mm.
in maximum diameter. The branch-like lobes of the lobate specimens are 8-11
mm. broad. Except in the sheltered places adjacent to the base of attachment,
where remnants of a spicule-fur can be detected, the surface is bare. In the
smaller specimens it is almost continuous, in the larger undulating. At one
place on the surface of the largest specimen (Plate 44, fig. 25, right above) there
is a row of five low warts. Apart from these warts and the most prominent |
convexities of the larger specimens the whole of the surface is occupied by pore-
sieves. The dermal membrane is, probably in consequence of post mortem
shrinkage, more or less depressed over the radial cortical canals and their distal
branches, so that the surface appears more or less pitted. In a restricted area,
6-10 mm. in maximum diameter, which is in the larger specimens situated in a
concavity, the dermal pores are rather large, everywhere else they are quite
small. I consider the small pores, which occupy by far the greater part of the
surface, as the afferents, the large ones, confined to the restricted areas mentioned,
as the efferents.

The single specimen of var. latana is fragmentary. It has the shape of a
dise and measures 23 X19X9 mm. Its contour is pear shaped, and it was at-
tached to astone. The natural surface is destitute of a spicule-fur, quite smooth,
and occupied throughout by small, apparently afferent pores.

The colour of the specimens of var. angustana is, in spirit, nearly white
throughout; one has a slight lilac-gray tinge. The specimen of var. latana is
gray with a lilac tinge throughout.

The superficial part of the body forms a cortez, which consists of a thin, in
many places hardly perceptible, outer dermal layer; a middle sterraster-armour
layer (Plate 43, fig. 25a; Plate 44, fig. 26a), for the most part 400-700 y thick;
and a thin, inconspicuous inner fibrous layer. In one place, where a foreign
body appears formerly to have been attached to the surface (Plate 43, fig. 25a,
to the left), the sterraster-armour is only 150-250 p thick.

Canal-system. More or less stellate groups of afferent pores (Plate 43,
fig. 26) occupy the largest part of the surface. These sieve-like pore-groups are
250-500 4 in diameter and quite close together, their centres being only 300-
600 » apart. The pores themselves are oval or circular, 10-50 » wide, and

GEODIA ATAXASTRA. 81

separated by rather broad strands of dermal tissue. The pores of each group
lead into a system of lacunose, subdermal canals, which converge and unite
below the centre of the pore-group to form an afferent radial canal (Plate 44,
fig. 26b). The subdermal afferent canals are wide, separated only by relatively
thin walls, vertical to the surface of the sponge, which radiate from a common
centre. These walls attach the pore-sieves to the sterraster-armour and give to
them, when viewed en face, the stellate appearance referred to above (Plate 43,
fig. 26). The radial cortical afferent canals are 120-230 » wide and circular in
transverse (paratangential) section. Their centres are 300-600 apart.

Below the sterraster-armour layer numerous small cavities, measuring on
an average about 100 » in radial diameter, are met with (Plate 44, fig. 26).
The radial cortical canals open out into these cavities, and from them the choano-
somal afferents take their rise. The flagellate chambers (Plate 44, fig. 24) are
spherical and measure 15-20 in diameter. In the interior of the choanosome
large canals, some 1 mm. and more wide, are observed (Plate 43, fig. 25).

The efferent cortical canals, which are confined to the efferent areas referred
to above, also have a circular transverse section. They are 200-300 wide and
their centres mostly 600 #-1 mm. apart. In places, particularly towards the
margins of the areas occupied by them, they are more distant. The outer
openings of these canals are covered by nets of dermal strands, only 10-20 p
broad. The meshes of these nets are oval or polygonal, 70-200 » broad, and up
to 400 » long (Plate 43, fig. 28). Thus, when the efferent area is viewed en face
one sees only a few strands or a loose net work of strands spread out over the
entrances of the efferent cortical canals.

Skeleton. Strands of rhabds, extending obliquely or paratangentially,
traverse the deeper parts of the choanosome. The rhabds composing them are
chiefly amphioxes, but amphistrongyles, styles, and angularly bent or irregularly
branched amphiox-derivates also occur in them in small numbers. In var.
latana, the amphistrongyles and styles are relatively much more numerous
than in var. angustana. In the distal part of the choanosome similar rhabds
and the rhabdomes of orthoplagiotriaenes, anatriaenes, and mesoproclades,
chiefly mesoprotriaenes, form radial bundles. The rhabdomes of some of the
orthoplagiotriaenes are reduced in length and rounded at the end. Ortho-
plagiotriaenes with such rhabdomes have been chiefly observed below the thin
part of the cortex of a specimen of var. angustana, mentioned above. The
cladomes of nearly all the orthoplagiotriaenes and of a large number of the
anatriaenes lie in the level of the lower limit of the sterraster-armour. A few

82 GEODIA ATAXASTRA.

anatriaene- and mesoproclade-cladomes were found further down in the choano-
some. Most of the mesoproclades seen protruded 800 #-1 mm. beyond the
surface. The small remnants of spicule-fur observed consisted nearly entirely
of the distal parts of such spicules (Plate 44, fig. 27b).

In the interior no minute rhabds are observed; in the distal part of the
choanosome and in the inner and middle (sterraster-armour) layers of the cortex,
on the other hand, such spicules, for the most part situated radially, occur in
considerable numbers. Where the dermal layer is better preserved, particularly
over the entrances to the radial cortical canals, large numbers of minute rhabds
are observed. They are here situated radially, their distal ends protruding a
little beyond the surface. These minute dermal rhabds are arranged in groups,
those of the same group being either parallel or diverging distally tuft-fashion.
Most of these spicules are amphiox, a few amphistrongyle or style. Minute
*amphistrongyles and styles are relatively much more numerous in var. latana
than in var. angustana. In var. angustana a few minute dermal anaclades are
intermingled with these minute dermal rhabds.

In the interior of the choanosome a few sterrasters and large, few-rayed
oxyasters occur. The latter are situated in the canal-walls. In the distal part
of the choanosome and the inner layer of the cortex oxysphaerasters, acanth-
tylasters, small strongylosphaerasters, and ataxasters are met with. The
middle layer of the cortex is occupied by dense masses of sterrasters. It also
contains the parts of the rhabdomes of the fur-spicules which traverse it, the
minute rhabds mentioned above, and small strongylosphaerasters and ataxasters.
In the dermal membrane a thin but dense layer, composed of small strongylo-
sphaerasters and ataxasters is observed. The strongylosphaerasters are much
more numerous than the ataxasters.

The large choanosomal amphiozes (Plate 48, figs. 23d, 27d) are usually slightly
curved, fairly isoactine, and gradually attenuated to sharp points, much more
rarely blunt at one or at both ends. The last-named forms are relatively much
more frequent in var. latana than in var. angustana. The measurements of
three amphioxes, two isoactine and one anisoactine, given in the subjoined table,
indicate the degree of attenuation towards the ends.

GEODIA ATAXASTRA. 83
AMPHIOXES
of Thickness
100 4 200 earthe 200 » 100
below one end mee below the other end
Var. angustana 17 « 20 4 30 1 20 1 17 pt
Var. latana 12 pu 16 p 27 p 16 12 uu
Var. latana 12 p 17 32 4 25 p 17

In var. angustana the large choanosomal amphioxes are 1.2-2.8, mostly
1.6-1.8 mm. long, and 17-43, mostly 22-38 » thick. In var. latana they are
considerably smaller, measuring only 0.6-1.9 mm., mostly 1.2-1.6 mm., by 12-
37, mostly 20-384 pu.

I have subjected the length of the large amphioxes (and also the length
of the rhabdomes of the orthoplagiotriaenes) of this species to a biometric in-
vestigation for the purpose of attaining an insight into the relative frequency
of the different lengths of these spicules. In all, the length of about a hundred
and fifty amphioxes of the two varieties, taken at random, were measured.
The amphiox-lengths within ranges of 0.1 mm. (from 0.45-0.55 mm., from 0.55
to 0.65 mm., and so on) were counted, these numbers reduced to percentage,
and the percentage frequency-numbers thus obtained plotted on the ordinates
erected at the points of the horizontal axis indicating the amphiox-lengths (of
0.5 for 0.45-0.55 mm., 0.6 for 0.55-0.65 mm., and so on). By connecting the
points thus plotted the two curves for the two varieties (Fig. A) were obtained.

Per cents of all the
large rhabds measured,

bhywptkjw Fo rnr
DEEN one se

Fig. A. Percentage frequency of lengths of the large choanosomal rhabds (chiefly amphioxes) measured
——-— Geodia ataxastra var. latana.

"mm. in length

Geodia ataxastra var. angustana,

84 GEODIA ATAXASTRA.

Although somewhat irregular, both curves have one very well-pronounced
main culmination, broad in the var. latana curve, and narrow in the var. angus-
tana curve. These main culminations show that one amphiox-length is much
more frequent than any other. The most frequent lengths represented by these
culminations are 1.4 to 1.5 mm. in var. latana and 1.8 mm. in var. angustana.
These may be considered as the normal amphiox-lengths. 'That the “normal”
amphiox-length is greater in var. angustana than in var. latana accords with the
fact that the specimen of var. angustana selected for this examination is much
larger than the specimen of var. latana.

The var. latana curve has one, the var. angustana curve two, minor culmi-
nating points besides the main one. The left one (at 1.3 mm.) of the latter
appears to correspond to the single one (at 0.9 mm.) of the former. Since the
majority of amphioxes shorter than those of ‘‘normal” length (1.4-1.5 mm. in
var. latana and 1.8 mm. in var. angustana) are probably young, still growing
spicules, the minor culmination which precedes the main one in both curves
indicates that the rate of longitudinal growth of these spicules is not uniform,
a stage of cessation or retardation of growth at 0.9 mm. in var. latana and at
1.5 mm. in var. angustana intervening between the earlier and the later periods
of rapid growth.

The rare large amphistrongyles have similar dimensions to the large
amphioxes.

The rare large styles are 0.5-1.6 mm. long and 11-30 » thick. Their thick-
ness is by no means proportional to their length.

The rare angularly bent and irregularly branched amphiox-derivates have
similar dimensions as the regular amphioxes. In the angularly bent amphiox-
derivates the bend is always near one end. The angle may be over or under
90°. Most of the branched forms are amphiclade in character and consist of a
shaft with a short branch-ray near each end. In some only one branch-ray,
situated near one end of the shaft, is observed. Generally the branch-rays are
simple, straight, conical, and pointed, rarely irregularly curved or divided into
secondary branchlets. The simple branch-rays are 20-100 y long and rise
vertically or obliquely from the shaft. The oblique ones are inclined outward,
proclade-fashion.

The minute rhabds of the distal part of the choanosome, the cortex, and the
dermal groups (Plate 44, figs. 31a, 32, 33a, 40a) are, in var. angustana, mostly
fairly straight, rather abruptly pointed amphioxes. In var. latana minute
dermal rhabds rounded at one end or at both occur in fair numbers besides the.

GEODIA ATAXASTRA. 85

ordinary amphiox forms with both ends sharp pointed. In var. angustana
these spicules are 120-200, mostly 160-200 y long and 3-5 yp thick; in var.
latana somewhat larger, 145-215 » long and 4-7 » thick. The shorter ones are
on the whole thicker than the longer ones. In var. angustana the average
thickness of the minute rhabds less than 160 in length is 4.6 4, that of the
ones more than 160 # in length 3.9 yu.

The orthoplagiotriaenes (Plate 43, figs. 9-22, 24a, 27a) have a straight,
usually conical and sharp-pointed, rarely shortened, cylindrical, and terminally
rounded rhabdome. The ordinary pointed rhabdomes are, in var. angustana,
1.6-2.3, mostly 1.9-2.2 mm. long and, at the cladome, 29-70, mostly 40-60
thick, in var. latana they are considerably smaller, 1.3-1.7, mostly 1.5-1.6 mm.
long and 30-45, mostly 30-40 » thick. The cylindrical, terminally rounded
rhabdomes are as thick as the conical ones at the cladome but only 330 y-1.4
mm. long.

I have studied the frequency of the different lengths of the ordinary pointed
orthoplagiotriaene-rhabdomes in a similar manner to the amphiox-lengths.
The curves representing the frequency of the different lengths measured are
here reproduced (Fig. B).

Per cents of all

thabdomes
measured.

mm, in length.

Fig. B. Percentage frequency of lengths of rhabdomes of the orthoplagiotriaenes measured.
Geodia ataxastra var. angustana.
4 ——-—-— Geodia ataxastra var. latana.

They are similar to the frequency-curves of the amphiox-lengths, and like
them exhibit a distinct main culmination, broad in var. lafana and narrow in

86 GEODIA ATAXASTRA.

var. angustana, which shows that the ‘‘normal” lengths of the orthoplagio-
triaene-rhabdomes are 1.5-1.6 mm. in var. latana and 2 mm. in var. angustana.
Also in these curves a minor culmination, preceding the main one, is observed,
and I think there can be little doubt that this indicates, as in the case of the
amphiox-lengths, a temporary cessation or a retardation of longitudinal growth
of the orthoplagiotriaene-rhabdomes, in this case when they are 1.3 mm. long in
var. latana and when they are 1.6 mm. long in var. angustana. In these
curves, as in the ones pertaining to the amphioxes, the part beyond the main
culmination descends very rapidly at first and less rapidly later on. This
seems to show that most of the amphioxes and orthoplagiotriaene-rhabdomes

becoming longer than the normal grow beyond this very rapidly.

°, average of
angles of clades
of that length.

- clade-length,
Fig. C. Correlation between the lengths and angles of the clades of the orthoplagiotriaenes of Geodia
ataxastra var. angustana.

The cladomes of most of the orthoplagiotriaenes are regular, the three
clades of the same cladome being nearly equal in length, position, and curvature.
The clades of these regular orthoplagiotriaene-cladomes are conical, pointed,
and curved, concave to the rhabdome, rather strongly in their proximal and
central parts but only slightly or not at all in their distal part. In var. angu-
stana they are 170-290, mostly 230-270 y long, in var. latana shorter, only 130-
220 » long. The angle enclosed between the clade-chords and the rhabdome-
axis is In var. angustana 85-116°, on an average 100.3°, in var. latana 95-1119,
on an average 101°. Considering, as I do, teloclades with a clade-angle between
80° and 100° as orthoclades and teloclades with a clade-angle of 100-120° as
plagioclades, I name these spicules orthoplagiotriaenes.

For the purpose of ascertaining the correlation between the clade-length on

GEODIA ATAXASTRA. 87

the one hand and the clade-angle and the rhabdome-thickness (at the cladome)
on the other I measured ‘these dimensions in eighty-one cladomes taken at
random. I arranged these measurements in the order of the clade-length, and
divided up the series at intervals of 10 », considering all clade-lengths from 145 to
155 as about 150 y, all from 155 to 165 as about 160 “and so on. All the meas-
urements of clade-angles and rhabdome-thickness pertaining to clades of similar
length (150, 160, and so on) were combined and their averages taken. These
averages were then plotted on the ordinates erected in the points of the hori-
zontal axis representing the clade-length (150, 160 4, and so on). The curves
connecting the points thus plotted in are here reproduced.

Averages of f,
thickness of rhabd-
ome of orthopla-
giotriaenes with
clades of that
length.

ft clade-length.

Fig. D. Correlation between the clade-length and rhabdome-thickness of the orthoplagiotriaenes of
Geodia ataxastra var. angustana.

These curves are very irregular, but they show nevertheless that the clade-
angle on the whole decreases with increasing clade-length, while the rhabdome-
thickness increases with it. That is to say, that, roughly speaking, the width of
the clade-angle is in inverse, the thickness of the cladomal end of the rhabdome
in true proportion to the clade-length. According to the curves the clade-angle
and the thickness of the cladomal ends of the rhabdomes do not seem to be
correlated with each other any more closely than with the clade-length. On the
whole it may therefore be said, that, although there undoubtedly exists a corre-

SS GEODIA ATAXASTRA.

lation between the three cladome dimensions, they are subject to considerable,
apparently independent variations.

Besides the orthoplagiotriaenes with regular cladomes, above described,
a few occur in which the clades are irregularly and unequally curved (Plate 43,
fiz. 10). Very rarely monaene and amphiclade derivates of orthoplagiotriaenes
are met with. A monaene form observed had a bifurcate clade. An amphi-
clade possessed, besides the ordinary three terminal clades, a fourth clade, which
arose some distance down the rhabdome.

The mesoproclades of var. angustana (Plate 43, fig. 24b; Plate 44, figs. 1—
12, 14) have a straight or slightly curved rhabdome 1.6-3.4, mostly 2.1-2.8 mm.
long. At the cladome the rhabdome is 9-20 », most frequently 10-15 y thick.
Farther down it thickens, and at its thickest point, which is situated near the
middle, it is usually half again as thick as at the cladomal end. At the acladomal
end the rhabdome is usually attenuated to a fine point. Generally the cladome
is regularly triaene. The clades are usually conical, about 70 long, slightly
curved, concave to the epirhabd basally and straight distally. Their chords
enclose angles of 40-50° with the epirhabd. The latter is straight, conical,
pointed, and 40-50 » long. Besides these regular forms various irregular ones
are observed. In some of these a fourth clade is added to the three normal
ones, so that the spicule appears as a mesoprotetraene. The clades of these
tetraene mesoproclades (Plate 44, fig. 1) are somewhat unequal in length and
position. The longest clade of the cladome is 50-75 long. The chords of the
clades enclose angles of 36-62° with the axis of the epirhabd. The latter is
conical, pointed, and 35-58 » long. Other derivates are produced by one or
more clades being abruptly (angularly) bent outward in the middle (Plate 44,
fig. 14). These spicules are also very rare. More frequently irregularities due
to one or two of the clades being reduced in length, cylindrical, and terminally
rounded, are observed (Plate 44, figs. 11, 14). Rarely one clade is absent
altogether, so that the spicule appears as a mesoprodiaene (Plate 44, fig. 12).
Taking all the mesoproclades together, but leaving out of account the excep-
tionally large, quite abnormal angles of some of the tetraene clades, which,
as mentioned above, measured up to 62°, we find that the longest clade of
the cladome is 44-80 and the epirhabd 40-60 » long, the angle between the
clade-chords and the epirhabd axis being 33-48°, on an average 41.3°.

In var. latana the mesoproclades (Plate 44, fig. 13) have a rhabdome 2-2.4
mm. long and, at the cladome, 7-13 » thick. As in the mesoproclades of var.
angustana the rhabdome thickens towards the middle. Regular mesoprotriaenes,

GEODIA ATAXASTRA. 89

like those predominant in var. angustana, are rare in var. latana, one, two, or
even all three clades (Plates 44, fig. 13) being reduced in length and terminally
rounded in most of the mesoproclades observed in this variety. Most of the
pointed, fully developed clades are angularly bent outward in the middle like
some of the clades of the mesoproclades of var. angustana (Plate 44, fig. 14).
The dimensions of the cladomes of the mesoproclades of var. latana are: length
of longest clade of the cladome 25-58 4, mostly 32-50 4; length of epirhabd 28-
50 », mostly 34-47 w; angle between clade-chord and epirhabd axis 25-58°, on
an average 43.6°.

The large anatriaenes (Plate 44, figs. 15-23, 40d, 41) have in var. angustana
a rhabdome 2.1-3.1 mm. long and, at the cladome, 3-10 » thick. In var. latana
the anatriaene-rhabdome is somewhat shorter and thicker, measuring 2.-2.8
mm. in length and 7-12 y in thickness. The cladomes are generally regular,
irregular anatriaene-cladomes with one clade shorter than the others (Plate 44,
figs. 17, 41) being rare. Exceptionally all three clades are greatly reduced
in length, so that the whole cladome appears as a three-lobed tyle (Plate 44, fig.
23). In the anatriaenes of var. angustana the cladome is usually destitute of an
apical protuberance. The clades of these anatriaenes are regularly conical,
taper uniformly to a sharp point, are strongly curved, concave to the rhabdome
basally, straight or just perceptibly bent outward distally, and 17-68 4, mostly
30-65 long. Their chords enclose angles of 20-42°, on an average 34°, with the
axis of the rhabdome. The cladomes of the large anatriaenes of var. latana
are different in appearance. Most of them have a distinct apical protuberance
and their clades are very thick and not very strongly curved, concave to the
rhabdome at the base, and abruptly attenuated to very slender, straight or
slightly outwardly curved, sharp-pointed tips. They are 30-55 y long, their
chords enclose angles of 41-55°, on an average 47°, with the axis of the
rhabdome.

The minute dermal anaclades (Plate 44, figs. 28d, 42, 46-49), which have
been observed only in var. angustana, have a slightly curved rhabdome 190-340 ps
long. At the cladomal end it is 0.5-2 thick, and it increases in thickness
towards the middle to 1.4-3.5 y, the central part of the rhabdome being usually
not quite twice as thick as its cladomal end. The acladomal end of the rhabd-
ome is rounded and usually slightly thicker, rarely thinner, than the cladomal
end. Most of these minute anaclades are regularly triaene. The basal parts
of their clades are generally quite strongly curved, concave to the rhabdome,
while their central and distal parts are straight. The chords of the clades are

90 GEODIA ATAXASTRA.

2-6 » long and enclose angles of 33-57°, on an average 47.6°, with the axis of the
rhabdome.

Geodia ataxastra is very rich in asterforms, Eight different kinds of asters
can be distinguished. 1. Large oxyasters with few rays. 2. Oxysphaerasters
with numerous slender, spiny rays. 3. Oxysphaerasters with numerous stout,
smooth rays. 4. Small strongylosphaerasters. 5. Ataxasters. 6. Acanthtyl-
asters. 7. Irregular sterraster-derivates. 8. Regular sterrasters. Three and
seven are very rare and have been observed only in centrifugal spicule-prepara-
tions of var. latana. Six also, which occurs in both varieties, is by no means
frequent. The five other kinds, particularly four, are abundant, five being
more numerous in var. lafana than in var. angustana. One and two, and also
four, five, and six are connected by some transitional forms; seven may also
belong to the series of forms represented by the three latter, and at the same time
exhibits great affinities to eight.

The large oxyasters (Plate 43, figs. 35, 36, 37a, 38; Plate 44, figs. 28¢, 29c,
30c, 33-35c, 39, 40c) occur in both varieties. Those of var. angustana are
usually destitute of a centrum, only the two-rayed forms possessing a slight
central thickening. They have from two to ten concentric and regularly dis-
tributed, straight or slightly curved, usually simple rays. In some oxyasters,
however, one or more of the rays are bifureate, the terminal branches being
nearly equally long and strongly divergent. Very rarely trifurcate rays have
been observed. The simple rays are conical, gradually or abruptly pointed,
and always spiny. When abruptly pointed and distally covered with very
numerous spines they have a somewhat strongyle appearance. As a rule nearly
the whole ray is covered with spines, only a narrow belt at the base being smooth.
The spines increase in size towards the end, just below which they occasionally
form a conspicuous verticil. The spines of the proximal half of the ray are
often so minute as merely to render the appearance of this part of it rough.
When large enough to be clearly made out, the spines are seen to arise vertically
from the ray and to bend down at the end in a claw-shaped manner. In regard
to their spines the branched rays agree with the simple ones. The size of the
aster is in inverse proportion to the number of its rays. The two-rayed (diactine)
asters are 40-50 y, the three- to seven-rayed 21-40 yp, and the eight- to ten-rayed
15-28 win diameter. The rays are half or a little more than half of the diameter
of the whole aster in length, and, at the base, 0.6—2.6 ys, usually 1-2 pw thick. The
basal thickness is by no means in proportion to the length of the ray, shorter
rays being often much thicker than longer ones. As, however, the relatively

GEODIA ATAXASTRA. 91

thick-rayed oxyasters (Plate 43, figs. 36, 37a) are connected with the relatively
thin-rayed ones (Plate 43, figs. 35, 38) by numerous transitions, which form a
continuous series, it does not appear advisable to separate them.

The large oxyasters of var. latana are on the whole similar to those of
var. angustana, but considerably smaller. Diactine oxyasters and asters with
branched rays were not observed in this variety. The increase in the size of the
spines towards the end of the ray is not so marked in the oxyasters of this
variety as in those of var. angustana. The large oxyasters of var. latana have
from three to eleven rays, the three- to seven-rayed are 17—28 yp, the eight- to
eleven-rayed 15-21.5 » in total diameter. The basal thickness of the rays is
0.7-1.8 ys.

The oxysphaerasters with slender spiny rays (Plate 43, figs. 29-32; Plate 44,
fig. 29e, 34e) oecur in fair numbers in both varieties. Those of var. angustana
consist of a spherical centrum 2.6-5.5 p, that is about a third, of the whole
aster, in diameter, from which eighteen to twenty-eight regularly distributed
rays arise radially. The rays are straight, conical, and sharp pointed. Their
distal parts bear spines of considerable size, some of which are often arranged
in a verticil, situated some distance below the tip. The rays are, without
the centrum, 3.5-5 «long and at the base 0.7-1.3 y thick, the total diameter
of the aster being 8-13 . The spined oxysphaerasters of var. latana have
eighteen to twenty-one rays and resemble those of var. angustana very closely
but are somewhat larger and have relatively smaller centra. Their dimensions
are: centrum 2.4-3.5 yp, a fifth to a third of the whole aster, in diameter; rays,
without the centrum, 4-5.5 long and at the base 1.2 » thick; total diameter
10-14.4 p. :

In both varieties a few asters transitional between the oxyasters and oxy-
sphaerasters above described, in regard to ray-number, size, and development of
the centrum, have been met with.

The rare oxysphaerasters with stout smooth rays, which have been observed
only in var. latana, have a centrum 4.5 4, that is a little over a third of the
whole aster, in diameter, and about eighteen, regularly distributed, straight,
smooth, conical, and rather blunt rays, which are, without the centrum, 4.5 ys
long and at the base 2 « thick. The total diameter of the aster is about 13 y.

The small strongylosphaerasters (Plate 43, figs. 33, 34, 37b; Plate 44, figs.
30b, 31b, 33-35b, 40b) are exceedingly abundant in both varieties. Those of
var. angustana have a more or less regularly spherical centrum, 0.6-2.3 p, ex-
ceptionally as much as 3 y, a sixth to a half of the whole aster, in diameter,

92 GEODIA ATAXASTRA.

from which from seven to twenty, very rarely as many as thirty, fairly regularly
distributed rays arise radially. The rays are straight or slightly curved, and
appear as short cylinders, rounded and often somewhat thickened at the end.
Indications of the presence of exceedingly small spines have frequently been
observed. The rays are, without the centrum, 0.7-2.6 » long and 0.2-0.8 « thick.
The total diameter of the aster is 2.6-6, usually 3-4.5 . The small strongylo-
sphaerasters of var. latana are very similar. They have a centrum 1-3 yp, a
fifth to a third of the whole aster, in diameter, and from eleven to nineteen,
rarely as few as nine, rays. The rays are, without the centrum, 0.6-1.9 #4, rarely
as much as 2.5 « long, and 0.3-0.8 4 thick. The total diameter of the aster is
3-6.4, usually 3.44.3 ys

The ataxasters are more abundant in var. latana than in var. angustana.
Those of var. angustana consist of a spherical or irregularly tuberous centrum,
743.5 in diameter, from which from one to eight rays arise. The rays are
conical or, more frequently, cylindrical and always rounded, sometimes thick-
ened at the end. They are rough or distinctly spined, without the centrum,
0.3-2.8 » long and 0.4-1.2 » thick. This distribution is most irregular, and
they arise radially or obliquely. Those of the same aster often differ very con-
siderably in size. When only a few, two to four, rays are developed, they
usually stand close together and form a bunch arising from one point of the
surface of the centrum, the rest of the latter being often rough or spiny,
but free from rays. When more rays are present they are usually somewhat
scattered, but in this case also a large part, generally more than a half, of the
surface of the centrum is free from rays. The whole aster is 4-7 # in diameter.
The ataxasters of var. latana have from two to fourteen rays. They are very
similar to those of var. angustana, but on the whole larger and covered with
somewhat larger spines. Among the ataxasters of this variety a few with
branched (bifureate) rays have been observed. Their dimensions in this variety
are: centrum 2-4.5 yw, a third to three quarters, of the whole aster, in diam-
eter; rays, without the centrum, 0.3-2.5 » long and 0.7-1.5 y thick; total
diameter 5.3-8.3 pu.

Strongylosphaerasters with a less proncunced irregularity in the distribu-
tion of the rays connect these ataxasters with the small strongylosphaerasters
described above.

The acanthtylasters occur in both varieties, but they are far from numerous,
and particularly scarce in var. angustana. Those of var. angustana have from
ten to fifteen rays and measure 8-16 » in diameter; those of var. latana have

GEODIA ATAXASTRA. 93

from four to nine rays and measure 10-14.5 in diameter. The acanthtylasters
have no central thickening. Their rays are usually distributed fairly regularly,
cylindrical, 1-2 » thick, and rounded at the end. The end, which usually appears
distinctly thickened, is densely covered with large spines, which are for the most
part directed outward. The central and proximal parts of the rays are quite
smooth. In the asters of this kind which I have found in var. angustana the
rays were always simple, while in several of the acanthtylasters of var. latana
some of the rays were terminally divided into two or, rarely, three short branches,
each provided with a special acanthtyle.

The rare irregular sterraster-derivates, which have been observed only in var.
latana, consist of a central mass from which tufts of large, slender spines arise.
The central mass is either simple and spherical, or oval, in which case two
tufts of spines, lying nearly opposite each other, rise from it; or it is lobate, in
which case the distal convex face of each lobe is covered with spines. The
number of the lobes of the lobate form is from two to five. Sometimes the
lobes are separated by rather deep recesses. The spines are 4-8 y long,
the whole spicule measuring 21-50 » in diameter. In their dimension and
in the shape, size, and position of the spines these spicules agree with young
sterrasters, and they make, on the whole, the impression of being early stages
of abnormal spicules of this kind. Abnormal sterrasters, which might be con-
sidered as their adult forms, have, however, not been observed.

The sterrasters (Plate 44, figs. 36-38, 43-45) are abundant in both varieties.
They are flattened ellipsoids. Those of var. angustana are 65-98 p long, 58-67 #
broad, and 47-57 4 thick, the average proportion of length to breadth to thick-
ness being 100 : 89 : 74. The sterrasters of var. latana are similar but smaller
and relatively thicker. They are 55-65 y long, 50-60 « broad, and 47-50 y thick,
the average proportion of length to breadth to thickness being 100 : 90 : 79.

In the centre of the sterraster a cluster, about 4 # in diameter, of small
granules is observed. Outside the umbilical area the rays protruding beyond
the surface are about 2.3 » thick and have a circular or broad-oval transverse
section. The transverse section of the rays surrounding the umbilicus is greatly
elongated in a direction radial to the centre of the latter and measures about
24by7 yp. The terminal spine verticils are, in the rays away from the umbilicus,
composed of from six to ten, in those of the rays surrounding the umbilicus of
a much larger number of lateral spines. On some of the periumbilical rays I
counted as many as sixteen. The spines are broad, conical, and small, mostly
under 1 pin length. The terminal faces of the periumbilical rays are very

94 GEODIA ATAXASTRA.

oblique, strongly inclined inwards (Plate 44, figs. 43-45). The umbilicus is
about 10 « deep.

The eight specimens of the two varieties of this species were collected on
October 26, 1904, on the shore of Perico Island, Panama.

Although these specimens belong without doubt to the same species, they
are not quite identical. Seven of them are quite or nearly white and have the
same structure, one is gray and differs from the rest by its ataxasters being
larger, by its megascleres and sterrasters being smaller, by the clade-angles of
its large anatriaenes being considerably wider (averages 34° and 47° respec-
tively), by its mesoproclade-cladomes being different in shape, by the rays
of none of its large oxyasters being branched, by possessing a few oxysphaer-
asters with thick, smooth rays and irregular sterraster-derivates, and by being

gdestitute of minute dermal anaclades. Some of these differences, particularly
those of the asters and the ana- and mesoproclades, seem to me to be germinal
in character.

The spiculation and the cribriporal nature of the afferents and efferents
show that these sponges belong to Geodia. For the reasons given in the descrip-
tion of Geodia agassizii I have compared them not only with the described
species of Geodia but also with those of Sidonops. The species most nearly
allied to them are Geodia tuber (tuberosa) O. Schmidt, G. distincta Lindgren, G.
hilgendorfi Thiele, G. miilleri (mulleri) Fleming, that described by Dendy as
G. ramodigitata Carter and here described as G. mesotriaenella. Geodia tuber, G.
distincta, G. muilleri, G. ramodigitata (Dendy), and G. mesotriaenella are dis-
tinguished from G. alaxastra by being destitute of ataxasters and by having
much larger dermal strongylosphaerasters. Geodia hilgendorfi differs from G.
ataxastra by possessing small oxysphaerasters instead of the strongylo-
sphaerasters.

GEODIA ATAXASTRA.

95

DIMENSIONS AND CHARACTERS OF THE SPICULES OF THE TWO VARIETIES OF

GEODIA ATAXASTRA.

Spicules. var. angustana. var. latana.
Large choanosomal chiefly amphiox, amphistrongyles | chiefly amphiox, but amphistron-
rhabds and styles rare; 1.2—2.8, mostly | gyles and styles relatively much more

Minute dermal rhabds

Orthoplagiotriaenes

Mesoproclades (nearly
all are triaene)

Large anatriaenes

Minute dermal ana-
clades

Large oxyasters

Oxysphaerasters with
slender, spined rays

Oxysphaerasters with
stout, smooth rays

Small strongylosphaer-
asters

Ataxasters

Acanthtylasters

Trregular asters (sterr-
aster-derivates)

~ Sterrasters

1.6-1.8 mm. long and 17-43, mostly
22-38 yu thick.

nearly all amphiox, styles exceed-
ingly rare; 120-200, mostly 160—
200 » long, and 3-5 » thick.

rhabdome 1.6-2.3, mostly 1.9-2.2
mm. long, and at cladome 29-70,
mostly 40-60 » thick; clades 170—
290, mostly 230-270 » long; clade-
angle 85-116°, average 100.3°.

rhabdome 1.6-3.4, mostly 2.1-2.8
mm. long, and, at cladome, 9-20
thick; (longest) clades 44-80 » long;
clade-epirhabd angles 33-48°, average
41.3°; epirhabd 33-73, mostly 40-60
p long.

rhabdome 2.1-3.1 mm. long; at
cladome, 3-10 « thick; (longest) elades
17-68, mostly 30-65 » long; clade-
angles 20-42°, average 34°.
thabdome 190-340 » long; at clad-
ome, 0.5-2 » thick; clades 2-6 » long;
clade-angles 33-57°, average 47.6°.
two-rayed 40-50, three- to seven-
rayed 21-40, eight- to ten-rayed 15—

28 in diameter.

8-13 » in diameter.
not observed.
2.6-6 » in diameter.

47 win diameter.

with ten to fourteen rays, 8-16 » in
diameter, very rare.

not observed.
65-78 p» long,

58-67 p broad,
47-57 p thick.

frequent; 0.6-1.9, mostly 1.2-1.6

mm. long, and 12-37 « thick.

chiefly amphiox, but also a fair num-
ber of styles; 145-215 y» long, and
47 p thick.

trhabdome 1.3-1.7, mostly 1.5-1.6
mm. long, and, at cladome, 30-45,
mostly 30-40 » thick; clades 130-
220 » long; clade-angle 95-111°,
average 101°.

rhabdome 2-2.4 mm. long, and, at
cladome, 7-13 thick; longest clades
25-58, mostly 32-50 » long; clade-
epirhabd angle 25—58°, average 43.6°;
epirhabd 28-50, mostly 34-47 » long.

rhabdome 2-2.8 mm. long, and, at
cladome 7-12 » thick; (longest) clades
30-55 y long; clade-angles 41-55°,
average 47°.

not observed.

three- to seven-rayed 17-28, eight-
to eleven-rayed 15-21.5 » in diameter.

10-14.4 » in diameter.

13 » in diameter, rare.

3-6.4 » in diameter.

5.3-8.3 in diameter.

with four to nine rays, 10-14.5 » in
diameter.

21-50 » in diameter.

55-65 yp long,
50-60 » broad,
47-50 » thick.

96 GEODIA MESOTRIAENA.

Geodia mesotriaena, sp. nov.
pachana, var. nov.
Plate 21, fig. 1; Plate 23, figs. 3, 5, 6, 8, 9; Plate 24, figs. 3, 5, 9.
microana, var. nov.
Plate 23, figs. 1-2; Plate 24, figs. 2, 6, 7, 10-13, 16, 19, 21.

megana, var. nov.
Plate 21, figs. 2-6; Plate 22, figs. 1-10; Plate 23, figs. 4, 7, 10-25; Plate 24, figs. 1, 4, 8, 14, 15, 17,
18, 20, 22-32; Plate 25, figs. 1-11.

I establish this species for three sponges, a complete dry specimen, a com-
plete spirit specimen, and some fragments, also preserved in spirit. These
specimens were collected at Stations 2909, 2942, 2958, off the coast of southern
California. Their spicule-fur is composed of large and conspicuous mesotriaenes
and to this the specific name refers. Although these three specimens are similar
enough to be considered as the same species, each possesses peculiar characters
so that it seems advisable to separate them as varieties. The most conspicuous
differences between them are those of their anatriaenes. In the specimen from
Station 2909, var. pachana, the anatriaene-cladomes are small and stout; in
that from Station 2942, var. microana, the anatriaene-cladomes are small and
slender; and in that from Station 2958, var. megana, the anatriaene-cladomes are
large.

The specimen of var. microana is cake shaped and appears as a low and broad
inverted cone with bulging sides, 72 mm. high, with largest and smallest hori-
zontal diameters of 116 and 104 mm. respectively. The upper side, forming the
base of the cone, is quite flat, the lower side, forming its apex, rounded. Judg-
ing from the fragments of var. megana this was similar (Plate 21, fig. 2) but larger,
probably as much as 200 mm. in maximum diameter. The specimen of var.
pachana (Plate 21, fig. 1) is relatively broader than the other two and somewhat
similar to a plane-convex lens with vertical.axis and flat upper face. It is 230
mm. in horizontal diameter and 120 mm. high. The basal part, the apex of the
lower convex side wherewith the sponge was attached, has been torn off — in its
complete state, with this apex, its height may have been 140 mm.

The colour of var. microana is yellow, that of var. megana brown and that
of var. pachana dirty olive-brown on the surface and a lighter, yellowish brown
in the interior. The first two are in spirit. The last is dry.

On the upper face and on the sides depressions are met with. In the smaller
specimens (var. microana and var. megana) these are few in number, large parts
of the surface being without any trace of them (Plate 21, fig. 2). In the larger

GEODIA MESOTRIAENA. 2

specimen of var. pachana they are more numerous and larger, 15-30 mm. wide
(Plate 21, fig. 1). Between the depressions on the upper face and the sides of
the latter low elevations, 18-27 mm. broad, arranged in curved rows and forming
gyrus-shaped ridges arise. Some of the depressions are isolated and shallow, most
of them join to form furrows, 6-12 mm. deep, separating the gyri. At the
bottom of these furrows apertures, 5-8 mm. in diameter, which lead into irregular
tubes, 4-13 mm. wide, traversing the interior of the sponge and occasionally
anastomosing, are observed. These tubes must not be confounded with true
canals; they are, as will be shown, praeoscular cavities. Masses of sponge-tissue,
on an average 30 mm. thick, separate these tubes from each other. In var.
megana and var. microana tubes of this kind are also met with. Here, however,
they are wider, in var. megana 20 mm. in diameter, and less numerous (Plate 21,
fig. 2b).

In all the specimens portions of the outer surface (Plate 21, figs. 1, 2a),
chiefly the sheltered parts in the depressions, are covered with protruding spic-
ules which form a fur. This spicule-fur is 5 (var. microana) to 10 (var.
megana and var. pachana) mm. high. The walls of the praeoscular tubes are
also hirsute, but here the protruding spicules are less numerous and do not
extend nearly so far beyond the surface. I do not doubt that the fur is pro-
duced by partial ejection of radial spicules on all parts of the surface, and
that, wherever it is now wanting, it has been lost by friction, either during life
or post mortem. As stated above, the protruding spicules forming this fur are
mostly mesotriaenes.

Where the large protruding spicules have been lost, slight depressions 1.3
(var. megana) to 2 mm. (var. microana) apart are observed on the surface.
Apart from these the surface appears, to the unaided eye, quite smooth.

The superficial parts, abutting on the outer surface and the praeoscular
tubes, are differentiated to form a cortex free from flagellate chambers.

This cortex (Plate 21, fig. 2; Plate 22, figs. 1-7, 10; Plate 23, figs. 24, 25;
Plate 25, fig. 1) is composed of three layers: an outer layer, containing small
dermal amphioxes and various euasters but no sterrasters, and traversed by
systems of mostly oblique canals (Plate 23, figs. 24, 25, Plate 26, fig. 1); a
middle layer filled with dense masses of sterrasters and traversed by the distal
parts of the narrow, radial, chonal canals (Plate 22, figs. 3-5, 7a; Plate 23, figs.
24b, 25b; Plate 26, fig. le); and an inner layer, poor in microscleres, traversed
by the proximal parts of the chonal canals (Plate 22, figs. 5, to the right, 6, 7b;
Plate 23, figs. 24,25). All three layers are penetrated by the rhabdomes of the

98 GEODIA MESOTRIAENA.

protruding mesotriaenes, the inner layer also by the distal parts of the rhabd-
omes of those orthotriaenes and anatriaenes the cladomes of which lie near its
outer limit. The outer layer is 300-500 « thick. The middle layer, being chiefly
composed of sterrasters, firmly held together by connective fibres, appears as
a strong armour (sterraster-armour). Below the. outer part of the surface this
layer is in all the three varieties 1-1.5 mm. thick, in the walls of the praeoscular
tubes thinner, in var. megana hardly half as thick (Plate 21, fig. 2). The tissue
composing the inner layer extends along some of the large, radial, choanosomal
canals a considerable distance into the interior; between these canals it is
usually 300-600 thick.

As stated above the outer surface is covered with slight depressions 1.3-2
mm. apart. These are situated between the points where the radial spicule-

gbundles abut on the surface (Plate 23, fig. 24), so that it appears as if these
depressions had been produced by a subsidence of the parts of the surface (cortex)
not supported by the radial spicule-bundles. The depressions are obviously
homologous to the depressed parts of the surface of horny sponges lying between
the conuli, while the elevated parts, supported by the radial spicule-bundles,
correspond to the conuli. Not only the outer surface but also the surface form-
ing the limit between the outer and middle layers of the cortex is raised in the
radii of the spicule-bundles, so that this also appears conulated; the “conuli”’
of this limiting surface are even higher than those of the outer surface. The
surface forming the lower limit of the middle cortical layer is not thus raised in
the radii of the spicule-bundles and nearly continuous. The thickness of the
outer and middle layers of the cortex are consequently far from uniform; in the
depressions the outer layer is thicker and the middle layer (sterraster-armour)
thinner than in the radii of the spicule-bundles (under the conuli). The small
ainphioxes in the outer layer of the cortex form radial tufts. Their outer ends
protrude some distance beyond the surface (Plate 23, fig. 251; Plate 25, fig. 1b);
on the outer surface of var. megana usually 50-80 p, in the praeoscular tubes of
this variety 200-300 ». Where the spicules penetrate it, the dermal membrane
is drawn up tent-fashion (Plate 25, fig. 1). Thus a great number of small, one
might say secondary, conuli are formed rising, everywhere from the surface
between the large (primary) conuli.

The canal-system proper. In the depressions between the conuli, groups
of pores, penetrating the dermal membrane and rendering it sieve-like, are met
with. On the outer, exposed surface of var. megana these pores are more or less
circular and 40-60 » in diameter. They lead into canals (Plate 22, figs. 1b, 2b;

GEODIA MESOTRIAENA. 99

Plate 23, figs. 24c, 25c; Plate 25, fig. 1c) 120-300 » wide. All the canals
originating from the pores of the same group converge to a point below
its centre and there join to form a cavity of considerable extent, which lies in
the outer layer of the cortex below the dermal membrane (Plate 23, figs. 24, 25).
From this subdermal cavity a narrow tube, the chonal canal (Plate 22, figs. 3-7e;
Plate 23, figs. 24, 25k) extends radially downwards, penetrating the middle and
inner layers of the cortex. This chonal canal, particularly its proximal part,
which passes through the inner layer of the cortex, is surrounded by a ring of
contractile tissue, the chone (Plate 22, fig. 6; Plate 23, figs. 24d, 25d), accord-
ing to the degree of the contraction of which the width of the chonal canal
varies. In most of the sections it is about 200 » wide above, where it enters
the middle cortical layer, and narrows centripetally to about 50 at its proximal
end (Plate 22, figs. 3-6; Plate 23, fig. 25).

At the lower limit of the inner layer of the cortex, the chonal canal opens out
into a wide choanosomal canal, likewise radial, extending down towards the in-
terior. These radial choanosomal canals (Plate 23, figs. 24, 25e) are usually
300-600 yz, sometimes as much as 1 mm. wide. Transverse membranes, protrud-
ing from their walls at intervals of 150-250 yp, partially divide the canal-lumen
into a row of chambers. Some of these radial canals are short and soon split
up into numerous narrow branch-canals which lead to the superficial flagellate
chambers, others are long and open into larger canals, 3-5 mm. wide, which
extend in a paratangential or oblique direction into the deeper parts of the
choanosome (Plate 21, fig. 2). Some of the radial canals leading down from
the chones are surrounded by thick mantles free from flagellate chambers (Plate
23, figs. 24f, 25f), which appear as centripetal continuations of the tissue form-
ing the lower cortical layer. The chamber-bearing tissue (Plate 23, fig. 24g)
occupies the interstices between the canals (canal-mantles). Within this tissue
the flagellate chambers are numerous and close together. They appear to be
more or less spherical and have a diameter of 20-30 yu.

It will be seen by the above that in general structure and in the character
of its canal-system this sponge is very similar to the Mediterranean Geodia miilleri
(cydonium).' I have been able to examine this species in various stages of
growth and thus to ascertain the true nature of the tubes leading down from the
large depressions on the surface and the different parts of the eanal-system proper.
I think there can be little doubt that in G. mesotriaena, as in G. mvilleri, both the

1R. v. Lendenfeld. Die Tetractinelliden der Adria. Denk. Akad. wissensch. Wien, 1894, 61, p.
138, ff. cs

100 GEODIA MESOTRIAENA.

afferent and efferent canals proper are cribriporal, some of the pore-groups being
the beginnings of the afferent, the others the termini of the efferent system.
Some of the radial choanosomal canals, as stated above, are surrounded by thick
mantles of tissue free from flagellate chambers, others are not so surrounded.
I think that the former are efferent, the latter afferent canals. Since in G.
mesotriaena the walls of the tubes leading down from the large depressions on
the outer surface have, as in G. miilleri, a cortex, continuous and virtually iden-
tical with the cortex of the outer, exposed parts, I do not doubt that the tubes
themselves are in the former, as in the latter, produced by a plicature and local
fusion of the growing sponge in such a manner that their lumina are in reality
outside the sponge and the tubes themselves not to be considered as canals
proper. As the afferent pores seem to predominate on the outer, exposed sur-
face, and as efferent pores only seem to occur in the walls of the tubes, I con-
sider the lumina of the latter as annexes of the efferent canal-system, that is,
as praeoscular cavities.

The specimen of var. megana, the best preserved of the three, is, in its
histological structure on the whole similar to G. miilleri, but also exhibits some
peculiarities. There is no accumulation of stainable cells (nuclei) at the surface.
The dermal membrane is traversed paratangentially by slender spindle-shaped
elements, drawn out at each end to a fine thread. The central swelling meas-
ures 1.5-2 y, each terminal thread 0.3 y in thickness. These elements consist
of a somewhat granular substance, the granulation being coarser and more
distinct in the spindle-shaped thickening than elsewhere. Lower down in the
outer layer of the cortex similar fibres, not situated paratangentially but ar-
ranged irregularly, occur. The fibres joining the sterrasters stain only very
slightly with azure but deeply with iron-haematoxylin. In the proximal part
of the cortex paratangential threads are often observed. These are not homol-
ogous to the threads in the distal part of the outer cortical layer but appear
as the connecting fibres, stretched out between the most proximal of the sterr-
asters. Since these sterrasters at the proximal limit of the sterraster-layer are
usually much farther apart than those above, the fibres connecting them are
often of considerable length. The differentiated contractile tissue, forming a
ring, 300-400 y broad, round the proximal part of the chonal canal, the chone,
is brown in colour. I assume that all the chones observed are considerably
contracted. The greater, outer part of the chone consists of a tissue composed
of circular fibres (Plate 22, fig. 10¢) and scattered euasters. This tissue does not
extend right down to the chonal canal, a layer of massive or radially elongated

GEODIA MESOTRIAENA. 101

cells and a coating of densely packed euasters intervening between the layer of
circular fibres and its lumen. The transverse membranes, crossing the radial
canals, are composed of three layers, an upper and a lower superficial granular
layer, and a central transparent, and more highly refractive, apparently fibrous
layer about 10 y thick.

In the tissue free from flagellate chambers which as stated above, envelops
some of the radial canals, large elongate cells are here and there met with singly
and in groups. These cells (Plate 22, fig. 9a, b) appear to be destitute of a
membrane; their protoplasm is coarsely granular; their nucleus oval or spherical.
Most of these cells are thick spindles drawn out at each end to a point (Plate
22, fig. 9a); some of them, however, are thus drawn out at one end only and
rounded off at the other, so that they appear pear shaped (Plate 22, fig. 9b).
They attain a length of 30-50 and a breadth of about 12 4. The nucleus meas-
ures 4 in diameter. In some of them I have observed small masses of easily
stainable granules close to the nucleus. In others, areas, more transparent and
less stained than the other parts, are met with in the protoplasm. These cells
seem to be ova.

Skeleton. In the innermost parts of the choanosome, that is, those farthest
from the outer surface and the praeoscular tubes, masses of large amphioxes
and a few large styles are found. Some of these are spicules irregularly
scattered; the majority join to form loose strands extending towards the elevated
parts. Bundles of spicules arise from these central masses (strands) and extend
radially towards the surface. In the smaller and more solid specimens of var.
microana and var. megana most of these bundles are straight or only slightly
curved and abut vertically on the surface (Plate 21, fig. 2). In the larger speci-
men of var. pachana, the structure of which is more complicated, only the bundles
extending towards the summits of the elevations (gyri) are straight and in their
distal parts vertical to the surface, while the others curve strongly, attempting
as it were, to reach the flanks of the gyri vertically, which, however, they do not
succeed in, so that their distal ends abut obliquely on the surface. Like the
central spicule-masses (strands), the proximal parts of these radial bundles are
composed entirely of large rhabds (numerous amphioxes and few styles). In
the distal parts of the radial bundles also rhabdomes of telo- and meso-clades
occur. The teloclades are mostly anatriaenes and ortho- or plagio-triaenes, the
mesoclades mesoprotriaenes. Besides these normal forms triaene-derivates with
two clades or only one occasionally occur. The rhabds (amphioxes and styles)
extend to the distal part of the choanosome, but usually do not reach the cortex.

102 GEODIA MESOTRIAENA.

Most of the cladomes of the anatriaenes lie in a zone, about 1 mm. thick, just
below the lower limit of the sterraster-armour. The cladomes of most of the
adult ortho- and plagio-triaenes are situated at the lower limit of the sterraster-
armour, appearing as if they supported this layer of the cortex. Lower down
only young ortho- and plagio-triaene-cladomes are found. These are remarkably
few in number. In the middle and outer layers of the cortex no teloclade- or
mesoclade-cladomes occur. The rhabdomes of the protruding mesotriaenes
penetrate the whole cortex; their cladomes lie high above the surface. The
free, distal parts of these mesoprotriaenes compose the fur of the sponge.

Besides these large spicules numerous small spicules, small rhabds, various
euaster forms, sterrasters, and, occasionally, sterroids are met with.

The small rhabds are irregularly scattered in the interior of the choanosome
and form tuft-like groups in the outer layer of the cortex (Plate 23, fig. 251;
Plate 25, fig. 1b). Those in the proximal part of the choanosome are on the
whole similar to but smaller than those in the superficial tufts. It is therefore
to be supposed that these spicules are formed in the depth of the choanosome
and then pushed up to the surface. The fact, however, that hardly any such
spicules occur in the distal part of the choanosome and the lower and middle
layers of the cortex, which they would have to pass through on their way from the
interior of the choanosome to the superficial tufts, makes it somewhat doubtful
whether this supposition is correct. These small rhabds are anisoactine, the
thinner, more pointed end of those forming the superficial tufts being directed
outward. The tufts of these spicules in the outer cortical layer appear as coni-
cal groups, the apices of which are situated at the limit between the outer and
middle cortical layers. From these apices the spicules of each group (tuft)
radiate outwardly, penetrate the whole of the outer layer of the cortex, and ex-
tend, as stated above, some distance beyond it, so that their distal ends protrude
freely over the surface. These spicule-tufts lie quite close together, the neigh-
bouring ones coming distally nearly or quite in contact with each other.

The sterrasters, between which occasionally a few sterroids occur, form a
rather dense mass in the middle layer of the cortex (Plate 22, figs. 3-5, 7;
Plate 23, figs. 24b, 25b; Plate 25, fig. le). Here only adult sterrasters occur;
young stages of these spicules are found in the lower cortical layer and the distal
part of the choanosome. In all the three specimens examined such young
sterrasters are, however, remarkably rare, which shows that, at the time of
capture, these spicules were in none of them being produced at all rapidly.

The sponge is very rich in euasterforms. On the surface, both the outer

GEODIA MESOTRIAENA. 103

and that surrounding the praeoscular tubes, numerous small strongylosphaer-
asters, forming a thin superficial layer, are met with (Plate 22, fig. 8). Similar
strongylosphaerasters occur in the lower parts of the outer layer of the cortex
. and in the choanosome. Occasionally I have observed much larger strongylo-
sphaerasters apparently transitional to the sterroids. In all parts of the sponge
asters with small centra, or with no central thickening at all, are found. In
many of these the rays are quite slender and fairly pointed, in others thicker
and blunt. Since in all, however, the rays taper more or less towards their distal
end and innumerable transition-forms connect the blunt-rayed with the pointed-
rayed ones, I think that all these euasters, the blunt-rayed ones as well as the
pointed-rayed ones, should be considered as oxyasters. Of these oxyasters two
kinds, a larger and a smaller, can be distinguished. The larger kind is restricted
to the choanosome. The smaller kind is met with chiefly in the cortex and the
distal part of the choanosome. Most of these asters lie superficially in the
canal-walls. In the proximal part of the cortex and in the choanosome, chiefly
in its distal part, large oxysphaerasters with numerous rays occur.

The large choanosomal amphioxes are 4.3-8.2 mm. long and 50-105 y, usually
80-100 » thick. Those of var. microana are considerably thinner than those of
the other two varieties. They are straight or slightly curved and generally
sharply and rather abruptly pointed. The two ends are similar (isoactine) or
slightly dissimilar (anisoactine). Very rarely spicules of this kind, blunt at
both ends, are met with. In one or two this bluntness is such that these spicules
might be termed amphistrongyles.

DIMENSIONS OF LARGE CHOANOSOMAL AMPHIOXES OF GEODIA MESOTRIAENA.

| Var. pachana Var. megana | Var, microana
Lark! Stile 3 oe So aoe eee | 5.5-7.1 eRe 43-78
Ad ae? ieee ee | 75-105 70-105 | 50-77

The large styles, scattered in small numbers among the large choanosomal
amphioxes, are 3-4 mm. long and 70-110 y» thick. The thickest part of the
spicule is close to the rounded end. These spicules are to be considered as
amphiox-derivates in which one of the actines has become rudimentary. They
are more frequent in var. megana than in the other two varieties.

The small dermal rhabds of the superficial tufts (Plate 24, figs. 1-6a; Plate
25, fig. 1d) are amphioxes and styles. They measure 380-680 y in length and
Saye in thickness. In var. megana (Plate 24, figs. la, 4a; Plate 25, fig. 1d)

104 GEODIA MESOTRIAENA.

and in var. microana (Plate 24, figs. 2a, 6a) most of these rhabds are quite or
nearly isoactine amphioxes, usually nearly cylindrical in the middle, and abruptly
attenuated towards the blunt ends. A few are distinctly anisoactine and
rounded at the thicker end so that they appear as styles. In var. pachana—
(Plate 24, figs. 3a, 5a) these spicules are more spindle shaped and usually taper
more gradually towards the ends. It is also to be noted that strongly anisoac-
tine amphioxes and true styles are much more numerous among these spicules

in var. pachana than in the other two varieties.

DIMENSIONS OF SMALL DERMAL RHABDS OF GEODIA MESOTRIAENA.

Var. pachana Var. megana Var. microana
Bene th mimi.soicats ete cra catia aia nen: | 420-680 335-571 380-550
ss | z = — SSE
@Uhickness p......... 10-19 10-18 9-15

The small amphioxes of the interior of the choanosome are similar to those
in the tufts but somewhat smaller. In var. megana they measure 170-440 y
in length and 7-17 pin thickness. As stated above these spicules may be young
stages of the small superficial rhabds in the tufts.

The rhabdomes of the ortho- and plagio-triaenes (Plate 21, figs. 3-5; Plate
23, figs. 16, 20-23) are nearly straight or curved, more or less cylindrical in their
cladomal part, and conical in their acladomal part. The acladomal end is sharp
pointed or blunt. The observation of thin transverse splinters lying flat, with
the rhabdome-axis vertical, shows the axial rod of the rhabdome to be triangular
in transverse section. Usually this rod (the canal wherein it lies) is quite nar-
row, 1 # broad or less. In some of these spicules, however, I found, after boiling
them in nitric acid, the axial canal in the acladomal part of the rhabdome as
much as 9 in diameter and wide open at the end, the latter having the shape of
avery thin-walled tube. The rhabdome is 4.6-7.2 mm. long and at the cladomal
end 75-120 » thick. The clades are 200-670 long; their chords enclose angles
of 85-117°, on an average 98.5°, with the axis of the rhabdome. Very often
the three clades of the same cladome differ considerably in size; ortho- and
plagio-triaenes showing such an irregularity appear in fact to be more frequent
than the regular ones. At the base the clades are always curved, concave
towards the rhabdome; their distal part is generally straight (Plate 24, figs. 16,
21, 22), or slightly curved in the opposite direction (Plate 24, fig. 20), more
rarely abruptly bent towards the rhabdome (Plate 24, fig. 23). In var. megana
the clade-rhabdome angle of these spicules is on an average only 91.9°, while it

GEODIA MESOTRIAENA. 105

is in the other two on an average 99.4° and 104.2° respectively. Thus these
spicules are mostly orthotriaenes in var. megana, but partly (var. pachana
or mostly (var. microana) plagiotriaenes in the other two. By the complete
suppression of one or two of the clades ortho- and plagio-diaenes and monaenes
are produced. These are, however, rare. A plagiomonaene which I found
among the spicules of var. megana had a rhabdome 100 # thick at the cladomal
end, and a clade 530 long, the chord of which enclosed, with the rhabdome-

axis, an angle of 102°.

DIMENSIONS OF ORTHO- AND PLAGIO-TRIAENES OF GEODIA MESOTRIAENA.

Var. pachana Var. megana Var. microana
length mm. 4.6-6.6 6.1-7.2 5.7-7.2
ENG thickness at eye 35 i“ 2
cladome 1 85-115 90-120 75-100
Clades length 250-660 200-670 320-630
Angles between the pew 94-108 85-96 95-117
clade-chords and the aie a ee as a a
FTE NE SOS | 99.4 91.9 104.2

The mesoprotriaenes (Plate 21, fig. 6; Plate 23, figs. 13, 14, 18, 19) have
It is thickest in the middle.
38-70 in transverse diameter and from here it tapers towards both the cladomal

a rhabdome 6-14 mm. long. Here it measures
and the acladomal end. At the cladome it is usually half as thick or less than
at its thickest point near the middle of its length, and here measures only 15-
40 win transverse diameter. As an example of this I give the following measure-
ments of a mesoprotriaene of var. megana. The rhabdome of this spicule was
10.5 mm. long. At the thickest point, which was situated 4.8 mm. below the
¢ladome, it was 59 p, and at the cladome 27 » thick. The epirhabd is 95-330 p
long and usually simple, straight, conic, and pointed (Plate 23, figs. 14, 18, 19).
Sometimes (Plate 23, fig. 13) it bears branches, forming an imperfect secondary
cladome above the cladome proper. The clades are not nearly so constant in
shape as the epirhabd, and mesoprotriaenes with irregular clades of very frequent
occurrence. In the most regular mesoprotriaenes (Plate 23, figs. 18, 19) the
clades are conical, pointed, or blunt, and more or less curved, concave towards
the epirhabd. Although this curvature appears rather uniform in the clades
themselves, the observation of their axial rods shows clearly that it is in reality

greatest at the base and decreases distally. The irregularities of the cladomes

106 GEODIA MESOTRIAENA.

are caused by retardation of growth or irregular bending of one, two, or, rarely,
all three clades. The first leads to irregularities in the length of the clades,
the last to other irregularities of the cladome. Not only the clades. but also
the epirhabd of the mesotriaenes may become rudimentary, whereby forms like
the one, represented in Plate 23, fig. 15, are reproduced. Such irregular telo-
clades are, however, rare. The clades are 90-310 y» long; their chords form
angles of 29-54°, on an average 42.4° with the epirhabd-axis. The meso-
protriaenes of var. megana have longer clades than those of the other two
varieties. In the mesoprotriaenes of var. pachana the epirhabd is shorter than

in those of the two others.

DIMENSIONS OF MESOPROTRIAENES OF GEODIA MESOTRIAENA.

Var. pachana Var. megana Var. microana ~
length mm. 6-14 8.8-11
thickness at
5— = 17-22
Rhabdome the cladome y oe 1)
spe thieck- 70 38-65
ness jt
Epirhabd length 95-230 140-330 100-300
Clades length ps 90-230 140-310 130-210
Angles between clade- ee aes acd 29-56 35-93 35-54
; minimum
chords and epirhabd- 3 - =| ee
spies average ° 43.7 43.5 40

The anatriaenes (Plate 23, figs. 1-12) have a rhabdome 11-16 mm. long,
22-40 » thick at the cladomal end, and attenuated to a fine, more or less
twisted thread at the acladomal end. The clade-chords are 90-270 long and
enclose angles of 34-58° with the axis of the rhabdome. They are curved, con-
rave towards the rhabdome, either (Plate 23, figs. 3, 6) quite uniformly or (Plate
23, figs. 10-12) more strongly between the first and second third of their length
than elsewhere. There is always a small protuberance on the summit of the
cladome. The clades of the anatriaenes of var. megana are considerably longer
than those of the other two varieties. The rhabdomes and clades of var. mi-
croana are considerably thinner than those of the other two varieties.

Rarely irregular, mesoclade anatriaene-derivates (Plates 23, fig. 17) with
three regular anatriaene-clades and a curved epirhabd are met with. In their

GEODIA MESOTRIAENA. 107

dimensions these spicules agree with the regular anatriaenes described above.
The epirhabd of the mesanatriaene of var. megana, Plate 23, fig. 13, is 200 long.

DIMENSIONS OF ANATRIAENES OF GEODIA MESOTRIAENA.

Var. pachana Var. megana Var. microana
| length mm. 11-16 15 11
Z = — —— == a
Rhabdome - ‘
| thickness at the 27-40 22-38 3-95
| cladome pt
Clade-chords length p 90-170 160-270 70-175
= a= = a? ee ee =e
Angles between clade- aa ene 39-56 3458 3955
chords and axis of the) minimum
rhabdome —— —|— — ~ Se
average ° 48 .2 45.2 47.9

Among the normal ewasters which occur in great numbers, small strongylo-
sphaerasters, small and large oxyasters with small centrum or without centrum,
and large oxysphaerasters can be distinguished. Besides these a few large
strongylosphaerasters are found. These aster-forms are connected by numerous
transitions.

The oxyasters (Plate 24, figs. 1b, 6b, 7a, 9b, 10-14a, 15, 19, 22, 23, 24a, b,
25a, b, 26-31) are 11-54 y in total diameter. They have a relatively small
centrum, the diameter of which is from one eleventh to one seventh of the total
diameter of the spicule, or no central thickening at all. Their rays are concentric,
straight, and conical. In some of these asters the rays are in their proximal parts
nearly cylindrical and decrease in thickness with increasing rapidity towards
the distal end (Plate 24, fig. 23). In others the rays are more regularly conical
and attenuated quite uniformly from base to tip (Plate 24, fig. 25a). The end
of the ray is usually blunt pointed, not so frequently either sharp pointed or
truneate. The basal part of the ray is smooth, the distal part, rarely also
smooth (Plate 24, fig. 26), usually covered with a varying number of smaller or
larger spines (Plate 24, figs. 27-31). The u. v. photographs (Plate 24, figs. 28-
31) show that these spines arise vertically and that they are often recurved
towards the centre of the spicule, in a claw-shaped manner. These oxyasters
have from five to twenty rays. Small oxyasters 11-20 / in diameter with from
nine to twenty rays, and large oxyasters 19-54 in diameter with from five to
fifteen rays can be distinguished. The largest oxyasters, that is those over 40
in diameter, have from six to eleven rays, while the smallest oxyasters have

108 GEODIA MESOTRIAENA.

from sixteen to twenty. The oxyasters attain a larger size in var. megana
than in the other two varieties.

DIMENSIONS OF EUASTERS OF GEODIA MESOTRIAENA.

Var. pachana Var. megana Var. microana
number of rays 6-15 5-13 6-14
total diameter 19-37 22-54 24.5-42
diameter of cen- ee
2.5-8 2-8 2-7

1 | Large oxyasters trum
length of rays 8-17 11-28 8-17
basal thickness of 1.52.7 eR 1.344
Trays -t
number of rays 9-10 8-15 11-20
total diameter 17 17-20 11-17.5

: periee te eo 1.5-2.5 3.5-5 3-5

2 | Small oxyasters trum
length of rays 1 7-8 6-6 3.5-7
be i

asal thickness of 1-1. 1-3 0.7-1.5
rays jt
number of rays 20-25 15-21 19-23
total diameter 1 20-30 19-32 26-27
diameter of cen-

3 | Large oxysphaerasters trum eat Bae gee
length of rays 7.5-8 7-12 7-11
basal thickness of
aa 1-1.5 1.6-3 1.8-1.9
number of rays 11-20 6-18 8-17

total diameter 8.2-14.5 6-14 7.5-14
Naree er fenie diameter of cen-
4 | Small strongylosphaer- trum 2.5-5 2-4 2.5-6
asters
length of rays 1 Del ar 2.5-6
basal thickness of a 5)
rays jt 1-1.8 0.5-2.5 0.8-2.1

The large oxysphaerasters (Plate 24, figs. 5b, 7c, 8b) are 19-32 y in total
diameter and have a spherical centrum 3-10 in diameter from which from fif-

GEODIA MESOTRIAENA. 109

teen to twenty-five straight, conical, sharp-pointed, and uniformly distributed
rays arise radially. The diameter of the centrum is from one seventh to one
third of the diameter of the whole aster. The rays are very spiny, particularly
in their distal part, 7-12 long, and 1-3 y thick at the base.

The small strongylosphaerasters (Plate 24, figs. 5-6c, 7d, 10-14b, 16-18, 20,
21b, 32) have from six to twenty concentric rays, usually equal in size, and
quite regularly distributed. Rarely such asters with unequal rays occur. In
these there are three or more long rays, and a number of more or less shortened,
rudimentary ones. The rays are cylindrical or cylindroconical, usually rounded
off terminally, covered with small spines, without the centrum 2-7 y long, and
0.5-2.5 » thick. The centrum is 2-6 , and the whole aster 614.5 jin diameter.
The small strongylosphaerasters of var. pachana have more numerous and more
slender rays than the strongylosphaerasters of the other two varieties.

The rare large strongylosphaerasters (Plate 24, figs. 7b, 9a, 21a), transitional
between the small strongylosphaerasters and the sterroids, have numerous,
usually cylindroconical rays densely covered with large spines. Their centrum
is from one third to a half of the whole aster in diameter. Their total diameter
is 16-33 yu.

The rare sterroids (Plate 24, fig. 8a), which have been observed in var.
megana only, have a very large centrum, two thirds or more of the whole aster
in diameter, from which very numerous, short and stout, cylindrical rays arise.
These are 5-8 yp long and 4-6 » thick. Their sides are smooth, their convex
terminal faces covered with numerous spines. These asters measure 39-58 yu
in total diameter.

DIMENSIONS OF STERRASTERS OF GEODIA MESOTRIAENA.

Var. pachana Var. megana Var. microana
IL@HYEIN Pec Gog ee ORO CRIES Cnet aaa 102-125 a 100-115 92-120
SRCACUD I eretersceictsiee sieie soi inf eek sogtsvece its ales 90-102 . 80-105 78-107
LIU OSS 11s 64 6. Oe oP aera eee eee 75-82 75-82 6778
a ea eee beth | apc: | scorns | wo.797

The sterrasters (Plate 22, figs. 1-7; Plate 23, figs. 24, 25; Plate 26, figs. 1-11)
are flattened ellipsoids 92-125 » long, 78-107 y broad, and 67-82 thick. Their
average length and average thickness is nearly the same in all the three

110 GEODIA MESOTRIAENA.

varieties; their average breadth, however, is greater in var. megana and var.
pachana than in var. microana. The proportion of length to thickness is in the
sterrasters of all three varieties 100 : 70-71; the proportion of length to breadth,
on the other hand, in var. pachana and var. megana 100 : 90-91, in var. microana
100:79.

The umbilicus is a round (Plate 26, figs. 2, 3, 9-11), or more (Plate 26, fig.
6) or less (Plate 25, fig. 7) elongated, cup-shaped depression or pit, 10-18 y in
maximum diameter. On the walls of the umbilical pit low irregular elevations,
covered with numerous minute spines, often forming protruding tufts, are met
with (Plate 25, figs. 9, 10). These elevations appear as ridges, extending from
the rays which surround the umbilical pit down towards its bottom, or as isolated
patches, the transverse diameter of which is similar to that of the rays. The
lowest part, bottom, of the umbilical pit is usually quite free from spined pro-
tuberances. The whole of the sterraster-surface, with the exception of the
part occupied by the umbilicus, is covered with protruding, cylindrical rays,
circular in transverse section, and usually about 3 y thick, which terminate
with a rather flat apical face. From the margin of the latter stout and blunt,
conical spines arise. The axes of these spines are more or less vertical
to the ray-axis, so that they appear as verticils round the summits of the rays
(Plate 25, figs. 6-8, 11). Away from the umbilicus the rays are mostly crowned
by regular verticils of five or six spines (Plate 25, fig. 8). The rays surrounding
the umbilicus are provided with a greater number of spines, sometimes with as
many as eleven (Plate 25, figs. 6, 7). From the ends of many of these perium-
bilical rays, not only the lateral spines forming the verticil, but also more or fewer
upright spines arise (Plate 25, fig. 6). The spines attain a length of about 1.3 4
and are, at the base, about 1 # thick. They are conical, straight, or more or less
curved, and blunt. Those of the periumbilical rays are on the whole larger and
more strongly curved than those of the rays on other parts of the sterraster.
In a few sterrasters, one in a hundred or less, the rays are larger and distally
crowned with a greater number of spines. These abnormal sterrasters appear
as transitions between the regular sterrasters and the sterroids.

The specimen of var. pachana was trawled at Station 2909 on January 8,
1889, in 34° 22’ N., 120° 8’ 30” W., depth 375 m. (205 f.); it grew on a bottom of
green mud; the bottom temperature was 7.3° (45.2° F.); that of var. microana
was trawled at Station 2942 on February 5, 1889, in 33° 38’ 45” N., 118° 13’ 45”
W., depth 37 m. (20 f.); it grew on a bottom of gray sand and broken shells;
the specimen of var. megana was caught with the tangles at Station 2958 on

GEODIA MESOTRIAENA. 111

February 9, 1889, in 34° 4’ N., 120° 19’ 30” W., depth 47 m. (26 f.); it grew on a
bottom of gray sand; the bottom temperature was 12.7° (54.9° F.).

These three varieties differ in many details. The specimen from Station
2909, var. pachana, is meandric and rich in praeoscular cavities, the other two
are nearly solid. The rhabdomes of all the three kinds of teloclades, the large
choanosomal amphioxes, and the small dermal rhabds, are considerably thinner
in var. microana than in the other two. Among the small dermal rhabds styles
are frequent in var. pachana but rare in the other two. The average clade-
rhabdome angle of the orthoplagiotriaenes is in var. megana 91.9°, in var.
pachana, 99.4°, and in var. microana 104.2°. The mesoprotriaene-epirhabds
are shorter in var. pachana, the mesoprotriaene-clades longer in var. megana,
and the clade-epirhabd angles smaller in var. microana than in the others.
The anatriaene-clades are considerably longer and less divergent in var. megana
than in the other two, which latter differ from each other by the anatriaene-
clades being stout in var. pachana and slender in var. microana. The oxyasters
and, to a smaller extent, also the oxysphaerasters are larger in var. megana
than in the other two. The sterrasters are relatively narrower in var. microana
than in the other two.

Since these specimens are all large and apparently full grown, these differ-
ences cannot be ascribed to differences in age. Some of them might of course
be mere individual adaptations or due to differences of germ-separation or mix-
ture before or during fertilization; others, however, particularly the differences
in the clade-rhabdome angles of the orthoplagiotriaenes, the shape of the dermal
rhabds, and the relative breadth of the sterrasters, seem to be germinal in nature
and sufficient for varietal distinction. (See table on p. 112.)

On account of their cribriporal afferents and efferents and their spiculation
. these sponges belong to Geodia. The only other species with similar spicula-
tion, either of this genus or of Sidonops, which I have also compared with
Geodia mesotriaena, are G. arabica Carter, G. agassizii, G. mesotriaenella, G.
breviana, and G. ovis.

According to the description and figures given by Carter’ and Topsent ’
the choanosomal euasters of G. arabica are different from those of G. meso-
triaena; the megascleres of the former are much smaller than those of the latter,
and G. arabica has hitherto been found only in the Red Sea, while G. meso-
triaena appears to be confined to the coast of California. G. agassizii has no

1H. J. Carter. A descriptive account of four subspherous sponges. Ann. mag. nat. hist., 1869,

ser. 4, 4, p. 4, pl. 1, figs. 13, 13a.
? £. Topsent. Eponges dela Mer Rouge. Mém. Soe. zool. France, 1892, 5, p. 23.

112

GEODIA MESOTRIAENA.

DIFFERENCES BETWEEN THE THREE VARIETIES OF GEODIA MESOTRIAENA,

|

Var. pachana

Var. megana

Var. microana

Shape |

meandrie

more solid

more solid

Large choanosomal
amphioxes

Small dermal rhabds

5.5-
105 1

7.1 mm. long, 75-
i thick

5.3-8.2 mm. long, 70-
105 p» thick.

4.3-7.8 mm. long, 50-
77 « thick.

420-680 jx long, 10-19
thick; styles numerous

335-571 p» long, 10-18 4

thick; styles rare.

380-550 long, 9-15 4
thick; styles rare.

Ortho- and _plagio-

triaenes

chiefly _ plagiotriaenes;
rhabdome 4.6-6.8 mm.
long, at cladome 85-115
uw thick; elades 250-660 yt
long; clade-rhabdome
angle 94—-108°, average
99.4°.

chiefly orthotriaenes;
rhabdome 6.1—7.2 mm.
long, at cladome 90-120
mw thiek; clades 200-670
wlong; elade-rhabdome
angle 85-96°, average
91.9°.

chiefly plagiotriaenes;
rhabdome 5.7-7.2 mm.
long, at cladome 75-100
uw thick; clades 320-630
long; — elade-rhabdome
angle 95-117°, average
104.2°.

Mesoprotriaenes

rhabdome 6-14 mm.
long, at cladome 15—
40 » thick; epirhabd
95-230 » long; clades
90-230 » long; clade-
epirhabd angles 29-56°,
average 43.7°.

rhabdome 8.8-11 mm.

long, at cladome 18—
40 y» thick; epirhabd

140-830 » long; clades
140-310 » long; clade-
epirhabd angles 35-53°,
average 43.5°.

rhabdome at cladome
17-22 » thick; epirhabd
100-300 » long; clades
130-210 » long; clade-
epirhabd angles 35-54°,
average 40°.

rhabdome 11-16 mm.
long at cladome 27-40 4

rhabdome 15 mm. long,
at cladome 22-38 thick;

rhabdome 11 mm. long,
at cladome 8-25 yp thick;

AnatrAenes thick; clades 90-170 » | clades 160-270 mm. | clades 70-175 mm. long;
Pa eee ge long; elade-chord rhabd- | long; clade-chord rhabd- | clade-chord rhabdome
ome angle 39-56°, av-| ome angle 34-58°, av-| angle 39-55°, average
erage 48.2°. erage 45.2°. 47.9°.
Small strongylo- Total diameter 8.2— 4 :
sphaerasters ise: total diameter 6-14 p. total diameter 7.5-14 yu.

Oxyasters

total diameter 17-37 y

total diameter 17-54

total diameter 11—42 p.

Oxysphaerasters

total diameter 20-30

total diameter 19-32 j«

total diameter 26-27 yu.

Sterrasters

102-125 » long, 90-102 px
broad, 75-82 , thick;
average proportion of
length to breadth to
thickness 100:91:70.

100-115 long, 80-105
broad, 75-82 y thick;
average proportion of
length to breadth to
thickness 100: 90:71.

92-120 » long, 78-107
broad, 67-78 y thick;
average proportion of
length to breadth to
thickness 100: 79:71.

praeoscular cavities, differently shaped small strongylosphaerasters, and much
Greodia breviana also has much
smaller megascleres; this species is also distinguished from G. mesotriaena by
its minute, dermal anaclades and the thickness of the clades of the large ana-
triaenes.

smaller megascleres and choanosomal oxyasters.

The species most nearly allied to G. mesotriaena are G. mesotriaenella

and G. ovis. Of G. mesotriaenella there is only a small specimen in the collec-

GEODIA AGASSIZII. 113

tion, and at first I thought that it was a young G. mesotriaena. A more careful
examination showed, however, that it differs from the latter not only by the
smaller size of its spicules, which, in view of the small size of its body, would
not, by itself, be of any systematic importance, but also by the shape of its
mesotriaenes, orthotriaenes, and oxysphaerasters. The mesotriaene-epirhabds
are in G. mesotriaena as long as or longer than the clades in G. mesotriaenella, as a
rule, they are very considerably shorter. The ortho- or plagio-triaene-clades
are in G. mesotriaena nearly straight or somewhat turned upward at the end, in
G. mesotriaenella concave towards the rhabdome right up to the tip. The oxy-
sphaerasters of (. mesotriaenella have stouter and less spiny rays than those of
G. mesotriaena. Geodia mesotriaena differs from G. ovis, by the possession of
praeoscular cavities in the interior; by its spicule-fur being not nearly so highly
developed; by having much smaller dermal strongylosphaerasters, and by the
absence of the minute anatriaenes and the oxyasters with very stout, regularly
conical, sharp-pointed rays, which characterize G. ovis.

Geodia agassizii, sp. nov.
Plate 26, figs. 1-21; Plate 27, figs. 1-19; Plate 28, figs. 1-28; Plate 29, figs. 1-17; Plate 30, figs. 1-
17; Plate 31, figs. 1-10; Plate 32, figs. 1-46; Plate 33, figs. 1-14; Plate 34, figs. 1-17.

Cydonium miilleri L. M. Lampe (non Fleming), Trans. Roy. soe. Canada, 1893, 11, p. 36, pl. 4, fig. 2.

I establish this species for twenty-two specimens obtained at nine different
stations on the west coast of North America; eight at Station 2886, one at Sta-
tion 2887, two at Station 2978, one at Station 3088, one at Station 3168, two
at Station 4193, four at Station 4199, two at Station 4228, and one at Station
4551.

The reexamination of the sponge determined by L. M. Lambe (loc. cit.) as
Cydonium miilleri Fleming in the collection of the Geological Survey of Canada,
and which was kindly placed at my disposal for examination, shows that it
differs specifically from the typical Geodia (Cydonium) miilleri and is an
immature form of the sponges here described. A new species with another
specific name must therefore be established for these sponges and Cydoniwm
miilleri Lambe 1893.

Although much has been written on the variability of sponges, our knowl-
edge concerning this subject is still very vague. The results of the examination
of the differences between these sponges, given below, throw some light upon it
so that greater general interest attaches to this species than to most of the
others here described. For this reason I have selected for it the name agassizit.

114 GEODIA AGASSIZITI.

One of the two specimens from Station 4228 is, like the one described by
Lambe, in some respects immature in character; all the others, although of
various sizes, are apparently adult. In the ease where more than one adult
specimen was obtained at the same station, these are fairly identical in structure.
The adult specimens from different stations, on the other hand, differ more or
less, but although these differences are not inconsiderable, I have, for the reasons
given below, united all in one species and have refrained from further sub-
dividing this into subspecies, varieties, or forms.

Shape and size. All the specimens have a more or less continuous surface
and are massive and destitute of vestibular or praeoscular cavities. The greater
number are attached by a small base and are either quite regularly spherical
(Plate 26, figs. 16, 19, 20; Plate 34, fig. 17), oviform (Plate 26, figs. 17, 18), or
somewhat irregular (Plate 26, fig. 21). Some are more cushion shaped and
attached by an extended base. The largest specimen, which;is a stout ovi-
form one (Plate 26, fig. 17), was obtained at Station 4193. It is 180 mm. long,
105 mm. broad, and 100 mm. high. Smaller, more or less regularly spherical
specimens, 20-55 mm. in diameter, were obtained at Stations 2886, 4228, and
4551. The smallest one of these, which is at the same time the smallest of all
the twenty-two, is the immature specimen above referred to. Two elongated
ones, measuring 68 X 35 X 35 mm. and 46 X 20 x 20 mm. respectively, were
trawled at Station 2886, and another elongated one, 47 X 32 < 30 mm., at
Station 2887. The specimens from Stations 2978, 3088, 3168, and 4199 are
broad-based, more or less cushion shaped, and not so regular in outline. The
largest of these was obtained at Station 4199. It measures 50 mm. in length,
46 mm. in breadth, and 34 mm. in height. The maximum diameter of the others
is 24-46 mm. The specimens from Station 4199 show a predilection for the
concave, inner side of tubular hexactinellid skeletons. One of them quite fills
such a tube, three quarters of the circumference of which is still present. The
immature specimen described by Lambe is smaller than any of these. It
measures only 12 by 10 mm.

In all the specimens by far the greater part of the surface is free from pro-
jecting spicules and finely granular, or, as in the specimens from Station 4228,
nearly smooth. In some remnants of a spicule-fur have been observed. In
the large specimen from Station 4193 there are a few areas with projecting spic-
ules up to 16 mm. long and lying very obliquely to the surface. In the cushion-
shaped specimens from Stations 2978 and 4199 a well-developed spicule-fur, up to
9 mm. in height, is observed in sheltered places near the base of the sponge.

GEODIA AGASSIZITI. 115

Also in the immature specimen from Station 4228 spicules protruding up to 5
mm. beyond the surface occur. From these observations I infer that large
spicules are protruded and a spicule-fur is thus formed by all these sponges,
and that this has been wholly or partly lost during life or after capture.

Larger apertures (oscules) are absent, but minute holes in the sterraster-
armour are observed in large numbers. Minute holes of two different sizes can
be distinguished. The larger, which are clearly visible to the unaided eye,
and through which the efferent cortical canals pass, are restricted to certain areas
of the surface. The smaller, which are not visible to the naked eye and through
which the afferent cortical canals pass, occupy the remainder of the surface.
In one of the specimens from Station 4199 the area perforated by the large
efferent holes is roughly circular in outline, 18 mm. in diameter, and slightly
concave, thus forming a shallow depression. In the other specimens the effer-
ent areas are not depressed. In the large specimen from Station 4139 there are
two efferent areas, one a horseshoe-shaped zone 20 mm. broad and 64 mm.
in total diameter, the other an irregularly circular patch 15 mm. wide. In
the smaller specimens there are one or two, rarely three, generally more or less
circular efferent areas 8-24 mm. in diameter. The holes piercing the sterraster-
armour in these efferent areas (Plate 26, fig. 13) are circular and measure 300-
700 in diameter, their centres being 1-1.5 mm. apart.

Concerning the mode of attachment it is to be noted that the young specimen
from Station 4228 has grown quite over part of the hexactinellid skeleton-net
which forms its support, the siliceous bars of the latter partly penetrating the
sterraster-armour of the Geodia and entering its choanosome which surrounds
them as if they formed a portion of the true internal skeleton of the Geodia.

Most of the specimens are light brown in colour. The larger one of the two
from Station 4193 has a few extensive darker, rust-brown patches on the surface.
Of the eight specimens from Station 2886 some also are light brown, while
the others are dark blue. The specimens from Stations 4228 and 4551 are
lighter in colour than the others, nearly white. The true colour of these sponges,
when preserved in spirit, is probably light brown or white; the rust-brown and
dark blue pigmentations of some of the specimens may possibly have been pro-
duced after capture.

The sponge has a cortex, which is in the adult specimens about 1 mm. thick
and composed of three layers, the dermal layer outside, the sterraster-armour
layer in the middle, and a fibrous layer within. The dermal layer is in the young
specimen from Station 4228 (Plate 32, figs. 8, 11, 12) and in several of the adult

116 GEODIA AGASSIZII.

ones (Plate 27, fig. 1) merely a thin dermal membrane. In other adult speci-
mens (Plate 27, fig. 2a) it is 150-200 y, rarely as much as 230 y, thick and com-
posed of loose tissue, containing small amphioxes and strongylosphaerasters, but
no sterrasters or oxyasters. The sterraster-armour layer (Plate 27, figs. 1b, 2b;
Plate 32, figs. 8a, lla, 12a) is in the young specimen from Station 4228 350-
400 y, in the adult specimens usually about 800 « thick. The inner, fibrous
layer is free from sterrasters and quite thin. The limit between the dermal
layer and the sterraster-armour layer is very clearly defined, the limit between
the latter and the inner, fibrous layer is somewhat indistinct.

Canal-system. The areas of the large efferent holes in the sterraster-
armour are, in all sufficiently well-preserved specimens, covered by a dermal
membrane perforated by numerous small afferent pores. These lead into
systems of canals traversing the dermal layer and converging to points lying
in the level of the limit between this layer and the sterraster-armour layer.
Here the canals of each system join to form a radial tube, surrounded by a
chonal sphincter, which occupies one of the small afferent holes in the
sterraster-armour. The afferent cortical canals are in all the sections ex-
amined very narrow, or quite closed. Below the sterraster-armour layer these
canals open out into subcortical cavities (Plate 27, fig. 2c) which are higher
than broad and often attain a radial dimension (height) of 1 mm. The chonal
sphineters do not protrude into these cavities. From the majority of these
subcortical cavities narrow afferent canals lead down to the adjacent flagellate
chambers. Some of the subcortical cavities join below to form large afferent
canals (Plate 27, figs. 1d, 2d) 1-2 mm. wide, which, repeatedly ramifying, supply
the more distant flagellate chambers. The flagellate chambers are spherical
and measure 27-35 4, usually about 30 y, in diameter. The efferent canals aris-
ing from them join to form large tubes (Plate 27, fig. le; Plate 32, fig. 5a), 1
mm. or more in diameter, which extend towards the areas of the large efferent
holes in the sterraster-armour layer. The afferent canals are not separated
from the chamber-bearing choanosomal tissue by special mantles and have
smooth surfaces. The efferent canal-stems on the other hand are, particularly
in their wider distal parts (Plate 27, fig. le; Plate 32, fig. 5a), enclosed in
sheaths, about 500 y thick, free from flagellate chambers, and greatly con-
stricted at very frequent intervals by transverse sphincter-membranes, pro-
truding far into their interior. Distally these efferent canal-stems divide
into branches which lead up to the cortex. From the summits of these
branches arise radial cortical canals (Plate 26, figs. 13, 14a, 15a), surrounded

GEODIA AGASSIZII. ity

by chonal sphincters, generally found open and usually 120-500 « wide.
Many of these efferent chonal canals are destitute of dermal sieves and open
out freely on the surface (Plate 26, fig. 13, those to the right). In some of
the specimens nearly all of them are thus naked; in most, however, some
of these efferents are covered by dermal sieves, composed of nets of threads, 50—
120 » broad, with round meshes of very variable size (Plate 26, fig. 13, those to
the left, figs. 14b, 15b). Thus, at first sight, it appears as if there were, in this
sponge, two different kinds of efferents, cribriporal and uniporal ones. A closer
examination, however, reveals remnants of dermal sieves in most of the aper-
tures appearing at first sight uniporal. I think it therefore highly probable
that all the efferents are, like the afferents, provided with sieve-membranes
(cribriporal) in the living sponge, and that, where they are now missing, they
have been lost post mortem.

The skeleton consists chiefly of large choanosomal and small dermal amphi-
oxes, orthoplagiotriaenes, mesoprotriaenes, anatriaenes, large oxyasters, large
oxysphaerasters, small strongylosphaerasters, and sterrasters. To these spicules,
which occur in all the specimens, a few large amphistrongyles, slender and
cylindrical or thick and club-shaped styles, mesoclade or amphiclade ortho-
plagiotriaene-derivates, anadiaenes, sterroids, and other irregular forms may
be added. In the specimens attached to hexactinellid skeletons, particularly
in the young specimen from Station 4228, small hexactinellid spicules, hexac-
tines, and scopules are also found imbedded in the choanosome. These foreign
spicules are by no means restricted to the base of the sponge which is attached
to the dictyonine network of the hexactinellid, but are found in all parts.

The large choanosomal amphioxes are arranged radially and form loose,
conical bundles extending from the centre or base to the surface of the sponge.
Some of the outermost of these spicules protrude beyond the surface and thus
take part in the formation of the fur (Plate 27, figs. 1, 2). The rare amphi-
strongyles, which I have observed only in the young specimen from Station
4228, are scattered in small numbers between them. The rare, large styles,
both the thinner cylindrical and the thicker club-shaped ones, are arranged
radially like the large amphioxes among which they occur, their rounded end
being situated distally, their pointed end proximally. These spicules are more
numerous in the distal than in the proximal parts of the bundles and often pro-
trude their rounded end beyond the surface. I have found the thick club-shaped
styles only in the specimens from Stations 3168 and 4193, the thin cylindrical
ones in all the specimens with the exception of those from Stations 2887, 2978,

118 GEODIA AGASSIZII.

4298, and 4551. The small dermal amphioxes form radial, tuft-like groups,
which arise from the sterraster-armour, traverse the dermal layer and expand
above (Plate 27, fig. 2), their distal ends protruding more or less beyond the
surface. In some specimens, as in the one from Station 3168, a section of
which is represented in Plate 27, fig. 2, this protrusion is very slight; in
others, as in a specimen from Station 4193 and in the young specimen from
Station 4228, the small amphioxes protrude as much as 200-280 » beyond the
surface. It seems that these differences in the degree of protrusion of the small
dermal amphioxes are, partly at least, due to differences in the state of
preservation and degree of shrinkage of the tender dermal layer; in the
well-preserved and not much shrunken specimens their protrusion is slight,
in specimens not so well preserved, it is great. Some small amphioxes,
similar to those forming the tufts im the dermal layer, are occasionally
observed in the proximal layer of the cortex and in the distal part of the
choanosome. The cladomes of the orthoplagiotriaenes generally le at the
limit between the cortex and the choanosome; their clades extend para-
tangentially in this level, their rhabdomes are directed radially inwards.
Sometimes, particularly in the young specimen from Station 4228, the ortho-
plagiotriaene-cladomes are situated a little higher, within the sterraster-armour
layer (Plate 32, figs. 8, 11, 12). The orthoplagiotriaenes do not protrude
beyond the surface and take no part in the formation of the fur. The rare
mesoclade and amphiclade orthoplagiotriaene-derivates and the quite irregular
forms belonging to this category of spicules have been found only in spicule-
preparations, so that I am unable to say what position they occupy in the sponge.
I have found mesorthotriaenes in the spicule-preparations of the specimens
from Stations 2978 and 4199, amphiclade orthoplagiotriaene-derivates in such
preparations of the specimens from Stations 3168 and 4199 and the young speci-
men from Station 4228. The mesoprotriaenes, the anatriaenes, and their
various derivates are also radially arranged. The cladomes of a few of them lie
a short distance beneath the surface, most of them protrude freely beyond it.
These spicules form the principal part of the spicule-fur. In this fur the meso-
protriaenes are generally much more numerous than the anatriaenes. Ana-
diaenes and other, irregular anatriaene-derivates have been observed only in the
specimens from Stations 3168 and 4228.

The small strongylosphaerasters form a single but dense layer on the surface
of the dermal membrane (Plate 26, fig. 15) and also occur in the interior. The
large oxysphaerasters, which appear to be much more numerous in the specimens

GEODIA AGASSIZITI. 119

from Stations 2886 and 3168 than in those from the other stations, are chiefly
met with in the walls of the cortical and subcortical canals (Plate 26, fig. 14)
and in the inner layer of the cortex. The large oxyasters are scattered through-
out the choanosome, where they chiefly occupy the canal-walls. The sterr-
asters occupy the middle layer of the cortex in dense masses (Plate 27, figs. 1b,
2b; Plate 32, figs. 8a, 1la, 12a). In some specimens, particularly the adult
specimen from Station 3168 and the young specimen from Station 4228, a good
many sterrasters, chiefly young ones, also occur in the choanosome (Plate 32,
figs. 8, 11, 12). A small number of sterroids are usually associated with the
sterrasters.

The large amphioxes (Plate 28, figs. 15, 16a, 17a; Plate 32, figs. 9, 10) are
numerous in all the specimens. They are cylindrical in their central part and
rather abruptly and not very sharply pointed, sometimes blunt at the ends.
They are usually isoactine or slightly anisoactine, a few are strongly anisoactine.
In the adult specimens they are 2.3-4.8 mm. long and 60-112 y thick, their
general average maximum * dimensions being 3.9 mm. X 86.3 y. In the speci-
mens from Station 2886 and the adult specimen from Station 4228 they are
smaller, both shorter and thinner, than in those from any of the others. In
the specimens from Station 4193 they are very slender, longer than in any and
thinner than in most of the others. In the specimens from Station 2978 they
are of medium length, but very much thicker than in any of the others. In the
young specimen from Station 4228 the large amphioxes are 2.1-3.4 mm. long
and 20-66 y thick, their average maximum dimensions being 3.1 mm. x 57 pu.
In the immature specimen described by Lambe these amphioxes measure 1.8—
3.4 mm. by 33-47 p. (See table p. 120.)

The rare large amphistrongyles, which have been observed in the young
specimen from Station 4228 are somewhat shorter than the large amphioxes,
about 55 yw thick in the middle, and attenuated to about 40 yw at the rounded,
somewhat truncate ends.

The large, slender, cylindrical styles (Plate 28, fig. 17b) which have been found
in small numbers in all the specimens, with the exception of those from Stations
2887, 2978, 4228, and 4551, are 1.5-3.4 mm. long and 60-110 y thick, gently
curved, and only slightly thickened, or not thickened at all, at the rounded end.

1Tn all cases these average maxima were obtained as follows:— first the averages of the dimen-
sions of the three largest amphioxes of the (adult) specimens from each of the nine stations were taken.
From these special maximum averages (of three), which are given in the subjoined table, the averages
were again taken. These latter averages (means) which appear in the subjoined table in the column
headed “from all stations” are the “general maximum averages.”

120 GEODIA AGASSIZII.

DIMENSIONS OF LARGE CHOANOSOMAL AMPHIOXES OF GEODIA AGASSIZII.!

Biatlons ere eer esos 2ss6 | 2887 | 2978 | 3088 | 3168 | 4193 | 4199 | 4228] 4551 [All Sta. 4228
an Ee 2.4~|2.4-13.5-|3.5-|3.1-| 301+ | 2.4 |@:3= | 27 oe
g * 3.8 | 4.1 | 4.2 | 4.4 | 4.5 | 4:8 13.9 | 3.5 | 2 eee

| ules, mm.

Length —
| average of the
three longest, || 3.4 | 3.8 | 4.1 | 4. 4.2 | 4.6 | 3.8 | 3.2 | 3.8 3.9 || 3.1
mm.

of apparently
full-grown spic-
ules, 4

60— | 80—- | 90- | 65- | 80- | 72- | 64 | 35- | 68- 60— || 20-
81 100 | 112 82 93 95 85 80 100 112 66

average of the
three thickest, 4 || 75 93 | 111 76 89 87 79 74 93 || 86.3 || 57

The large, thick, club-shaped styles (Plate 28, figs. 12-14), which oecur in
small numbers in the specimens from Stations 3168 and 4193, are straight or
slightly curved, always thickened at the rounded end, and 115-145 p thick.

The minute dermal amphiozes (Plate 27, figs. 3a, 7a, 8a) occur in all speci-
mens. They are slightly curved and usually not very sharply pointed at the
ends. Sometimes one end is much more blunt than the other, but the bluntness
never seems to be great enough to make the spicule appear as a style. In the
adult specimens these spicules are 160-480 long and 5-12 thick, their average
maximum dimensions being 342 X 9.06 ». Those of the specimens from Station
2886 are considerably larger, both longer and thicker, than those of the others.
Particularly slender ones are met with in the adult specimen from Station 4228.
In the young specimen from Station 4228 these spicules are 225-300 y long
and 4-7 » thick. In the immature specimen described by Lambe they are 180-
480 by 3-8 p.

DIMENSIONS OF SMALL DERMAL AMPHIOXES OF GEODIA AGASSIZII.

Station ...... | 2886 | 2887 | 2978 | 30ss | 316s | 4193 | 4199 | 4228 | 4551 || Fromall Stations|| 4228
160-480
| 220- | 210— | 230— | 260- | 160- | 270- 185— | 310- - 225-
Length, _
eh" | “480 | 340 | 270 | 360 | 290 | 300 | 2%? | 330 | 330 [maximum 2YeF I “so9
age 342
? 5-12
Thickness, » || 7-12 | 8-10 | 7-8 |8-10} 7 | 7-10] 8 |5-85] 8 |lmaximum aver-|| 4-7
age 9.06

‘In this and the following tables the specimen from Station 4228 is a young specimen; those from the other
Stations are adults.

GEODIA AGASSIZII. 12)

The orthoplagvotriaenes (Plate 26, figs. 3-12; Plate 29, figs. 1-6, 8-12, 14-17;
Plate 34, figs. 1-12, 14, 15) occur in large numbers in all specimens. Their
rhabdomes are straight or slightly curved and usually regularly conic, occasionally
more cylindrical. The conical rhabdomes are generally pointed (Plate 26, figs. 3,
5-12; Plate 34, figs. 1-7), more rarely blunt (Plate 26, fig. 4; Plate 34, fig. 9).
The more cylindrical rhabdomes are more or less shortened and thickened to
tyles at the acladomal end (Plate 32, figs. 12b; Plate 34, fig. 8, 10-12, 14, 15).
Orthoplagiotriaenes with conical rhabdomes occur in all the specimens. The
orthoplagiotriaenes with more cylindrical, shortened and terminally thickened
rhabdomes are exceedingly rare in the adult specimens, but quite numerous in
the young specimen from Station 4228. The ordinary conical orthoplagio-
triaene-rhabdomes of the adult specimens are 1.5-4.2 mm. long and 65-150 yu
thick at the cladome, their average maximum dimensions being 3.39 mm.
115.44 ». In the ordinary orthoplagiotriaenes of the specimen from Station
4551 the rhabdome is remarkably stout, shorter and at the same time thicker than
in the ordinary orthoplagiotriaenes of any of the others. In the ordinary ortho-
plagiotriaenes of the specimens from Station 4193 on the other hand the rhabd-
ome is very slender, its length being above the mean and its thickness less than
in the orthoplagiotriaenes of any of the others. The conical rhabdomes of the
orthoplagiotriaenes of the young specimen from Station 4228 are 2-3.5 mm. long
and at the cladome 50-100 y thick, their average maximum dimensions being
3.3 mm. X 90 w. The orthoplagiotriaene-rhabdomes reduced in length and
terminally thickened (Plate 34, figs. 8, 10-12, 14, 15) are more cylindrical, and
much less attenuated towards the distal end than the ordinary ones. The
degree of this attenuation is proportional to the length, very short ones (Plate
34, fig. 15) being not attenuated at all and regularly cylindrical. The terminal
thickening is usually fairly spherical. Its diameter is 15-25 % greater than
the thickness of the part of the rhabdome lying just above it. Sometimes one
or more slight thickenings of the rhabdome are observed above the terminal
thickening. A cylindrical, terminally thickened orthoplagiotriaene-rhabdome
of an adult specimen from Station 2978 was 900 » long and 155 y thick at the
cladomal end, the thickness of the thickened acladoma! end being 170 y. In the
young specimen from Station 4228 these more cylindrical orthoplagiotriaene-
rhabdomes are 760 /-2.35 mm. long, and at the cladome 60-105 y thick. Their
thickness is, roughly speaking, in inverse proportion to their length. The two
shortest observed, one of which is represented on Plate 34, fig. 15, were 950 and
760 » long and 95 and 105 » thick respectively.

122 GEODIA AGASSIZII.

In young orthoplagiotriaenes the entire clade, in the fully developed ones
its basal part only, is directed obliquely upward. In their further course the
clades of the fully developed orthoplagiotriaenes bend downwards, so that their
distal parts lie more or less in a plane vertical to the rhabdome. The chords of
the clades of the orthoplagiotriaenes of the adult specimens enclose angles of 73-
117°, on an average 98.2°, with the axis of the rhabdome. In the adult speci-
mens from Stations 3168 and 4228 the clade-rhabdome angles do not exceed
100°, so that here all these triaenes appear as orthotriaenes. In the adult
specimens from the seven other stations a smaller or a greater number of such
triaenes with clade-chord angles exceeding 100° and appearing as plagiotriaenes
in consequence, are observed. The average clade-angle, however, exceeds 100°
only in the specimens from Stations 2886, 2887, and 4193. The angles of the

* three clades of the same cladome are usually nearly equal. It is very rarely that
they become so different as to give the cladome a position oblique to the rhabd-
ome. Such orthotriaenes with oblique cladomes are represented on Plate 26,
fig. 3, and Plate 29, fig. 4.

The size and the shape of the clades are far from constant, not only the
clades of different orthoplagiotriaenes of the same specimen but even the clades
of one and the same triaene often being very dissimilar. Their maximum aver-
age dimensions are, however, about the same in all the specimens. At the base
the clades are a little thinner than the cladomal end of the rhabdome, the ratio
between these two dimensions varying between 7 to 10 and 9 to 10. The clades
are 240-560 » long, their average maximum length being 490.89 ». The
maximum average of those of the orthoplagiotriaenes of the specimen from Sta-
tion 4551 is the smallest, of those of the orthoplagiotriaenes of the specimens
from Station 4199 the largest. The orthoplagiotriaenes of the young speci-
men from Station 4228 have clades 300-500 long, their average maximum length
being 490 ». Their chords enclose angles of 88-108°, on an average 97° with
the axis of the rhabdome.

In all specimens orthoplagiotriaenes with simple clades, gradually de-
creasing in thickness and curvature towards the usually not very sharp-pointed
end (Plate 26, figs. 3, 4, 6, Plate 29, figs. 1-5, 8,11), are met with. In the speci-
mens from six of the stations all the orthoplagiotriaenes, or at least a very great
majority of them, have regular clades of this description. In the specimens
from Stations 3168, 4193, and 4199 on the other hand the orthoplagiotriaenes
with such simple and regular clades are not so numerous as orthoplagiotriaenes
with one or more clades rendered irregular by being either abruptly bent down

GEODIA AGASSIZII. 123

near the end, or branched. Clades abruptly bent down at the end are repre-
sented on Plate 26, fig. 8, Plate 29, figs. 6, 15, Plate 34, figs. 5,8. The ramified
clades (Plate 26, figs. 8-12; Plate 29, figs. 9, 10, 12, 14, 16, 17; Plate 34, fig. 15)
are so variable, that it is difficult to find two alike. Their branches either
extend in a longitudinal plane passing through the rhabdome, or less frequently
they diverge in different directions forming, if numerous, a terminal bunch.
They hardly ever lie in the plane of the cladome and therefore differ fundament-
ally from dichotriaene-end clades.. The simplest and most frequent forms of
these branched clades are those in which one straight, conical, thorn-like branch
arises from the lower (rhabdomal) side of the distal part of the clade. This
branch is either directed vertically downward (Plate 29, fig. 12), or, more fre-
quently, obliquely downwards and outwards (Plate 26, figs. 9-12; Plate 29, figs.
9,16). Its size is in proportion to the distance of its origin from the end of the
clade; when it arises near the end of the clade it is small (Plate 29, fig. 16)
when it arises farther away from it, it is larger (Plate 29, fig. 9). In some
clades of this kind the branch is terminally divided into small secondary
branchlets (Plate 29, fig. 17). Sometimes the clades bear two simple or second-
arily ramified branches (Plate 29, fig. 10). The most complicated forms are
_ those in which the clade terminally divides into a greater number of divergent
simple, or more often, secondarily ramified branches (Plate 29, figs. 14, 16).
In the immature specimen, described by Lambe, orthoplagiotriaenes and dicho-
triaenes occur. According to Lambe (loc. cit., p. 37), the latter are much more
numerous than the former, ‘‘few examples of the simple orthotriaenes”’ being
found. I, on the contrary, found the orthoplagiotriaenes quite as numerous
as the dichotriaenes if not more so. The orthoplagiotriaenes have a rhabd-
ome 2.1-3 mm. by 70-90 p, and clades 300-450 y long; the clade-angles are
91-103°. The dichotriaenes have a rhabdome 1-2.2 mm. by 50-75 yp, main
clades 150-300, and end clades 30-130 « long; the breadth of the whole
cladome is 350-700 y, the main clades enclose angles of 109-112° with the
rhabdome.

124 GEODIA AGASSIZII.

DIMENSIONS OF THE ORTHOPLAGIOTRIAENES OF GEODIA AGASSIZIT.

| | From
oO Ate) ee Ae 2886 2887 | 2978 3088 | 3168 4193 4199 4228 4551 See 4228
| |
. hieeeun || 2 | a9- | 2a- | 2.5- | 2.4 | 2.0 | 2.8- [a5 | 238-15 i 2
y full-grown || * a ce 9
spicules, mm. | 3.2 3.1 4.2 3.7 4 3.8 3.4 2.8 3.3 4.2 3.5
| j
length} 7 li =e
average of the | |
three largest | 3.1 Bae ie! 3.6 3.9 3.6 7, 2.5 3 3.39 3.3
. dimensions, | |
Rhabd-| ue
ome = — = — — |— — — 2) ees
|
a ta rueaent 73- | 90- | 95- | so- | 90- | 65— | 7o- | 8o= | 90- | 65— || 50—
y tusrown ! 410 | 110] 145 | 120] 137] 95 | 103] 105 | 150}. 150 || 100
spicules, » |
thick-| aii 2 ues
ness | | |
— average of the ||
three largest || 103 110 | 139 120 132 92 102 98 143 (115.44 90
dimensions, 1 || |
| SSet an hal i t
xs naparen’- || 270- | 250- | 270- | 270- | 320- | 400- | 320- | 440- | 240- | 240- || 300-
y murerown |! 500 | 530 | 520 | 530 | 520 | 550 | 560 | 520 | 450 | 560 || 500
spicules, ||
length |__| | ——_|— |———]
average of the ||
three largest || 467 507 483 493 483 520 540 497 428 |490.89|} 490
Clades dimensions, 4 ||
~ eB || eal ae is = =
| apparent- ||
ot ee | 94- 95— | 92= |) "73= || 83>) 94= |) “90> |) <95— ||) seen mean ee
| ty ful=stown || oa’ | 109 | 104 | 110 | 100 | 117 | 105 | 100 || tos 9)) atziieos
angle spicules, ||
| Ae | ee k
average, ° ! 100 102 96 99 92 104 98 97 96 98.2 97
I}

The mesorthotriaenes (Plate 26, fig. 1; Plate 29, fig. 7; Plate 34, fig. 16) are
very rare and have been found only in the adult specimens from Stations 2978,
4199, and 4228, and the young specimen from Station 4228. They consist of a
style-like shaft, from which three clades arise. The shaft is conical, 1.8-3 mm.
long and 78-164 » thick at the rounded end. It usually tapers to a simple point
at the other end. In one of these spicules, however, the thin end of the shaft
was bifid, terminating in two points, lying close together. The three clades
form a verticil situated 150-280 » below the rounded end. They are 78-300 yu
long. Their basal part is directed obliquely downward towards the pointed end
of the shaft. Distally they curve round towards its rounded end, either uni-
formly, or abruptly. It is not quite easy to say which of the two parts of the
shaft on either side of the clade-verticil is to be considered as the rhabdome and
which as the epirhabd. The fact that the pointed part is very much longer than

GEODIA AGASSIZII. 125

the rounded part is in favor of the view that the former is the rhabdome and the
latter the epirhabd. Since, however, the rounded part is the thicker of the two,
since the clades have their concave side turned towards this shorter and thicker
part; and since there can be little doubt that these spicules are derivates of the
orthoplagiotriaenes, in which the concave side of the clades is invariably turned
towards the rhabdome, it seems that the short, thick, and rounded part of the
shaft should be considered as the rhabdome and the long, thin, and pointed part
as the epirhabd.

The amphiclade orthoplagiotriaene-derivates (Plate 26, fig. 2; Plate 29,
fig. 13; Plate 34, fig. 13) are also very rare. They have been found only in the
adult specimens from Stations 3168 and 4199 and in the young specimen from
Station 4228. They differ from the orthoplagiotriaenes described above only
by possessing, besides the terminal cladome proper, a short, rounded or pointed
clade about 100-150 » long, which arises at the acladomal end of the rhabdome
or some other part of it more or less remote from the true cladome.

Besides these orthoplagiotriaene-derivates a few quite irregular spicules
have been observed, which, to judge from their general character, appear to be
derivates either of the orthoplagiotriaenes or of the large choanosomal amphi-
oxes. On Plate 28 photographs of some of these spicules are reproduced. One,
fig. 8, is a triaene with a shaft 35 y thick at the cladomal end, and three straight,
conical clades, 130 » long, approximately extending in a plane which passes
through the rhabdome. One, fig. 10, is a stout, large amphiox with two straight
and pointed, clade-like branch-rays, 165 long, arising 250 » below one of the ends
and extending very obliquely downward towards the centre of the amphiox.
One, fig. 9, has the appearance of a large amphiox, one end of which is replaced
by a centrally attached, obliquely situated, style-like rhabd, 430 » long. One,
fig. 11, is a large amphiox, from which, at a distance of 260 y from one of the
ends, a straight, conical, clade-like branch, 240 long, arises vertically.

Of other irregular spicules observed I mention an amphistrongyle, about
1 mm. long and thicker at one end than the other, with a straight, conical, branch-
ray 80 long, arising obliquely 70 j below the thinner end and directed towards
the thicker end.

The mesoprotriaenes (Plate 28, figs. 1-7, 16d; Plate 32, figs. 40, 41) occur in
all specimens. Their rhabdome, which is thicker in the middle than at either
end, is straight or only slightly curved, and in the adult specimens 2-6 mm. long
and at the cladome 7-40 thick, the average maximum dimensions being 5.1
mm. X 25.11 . The rhabdomes of the mesoprotriaenes of the young specimen

126 GEODIA AGASSIZII.

from Station 4228 are, at the cladome, 9-13 » thick, the maximum average of
this dimension being 12 . The clades are conical, pointed, and always curved,
concave towards the epirhabd, in their basal part. In their distal part they are
either curved in the same direction (Plate 28, figs. 3, 5, 6), or nearly or quite
straight (Plate 28, figs. 2, 4, 7; Plate 32, figs. 40, 41), rarely abruptly bent
(Plate 28, fig. 1). The angles between the chords of the clades and the
epirhabd are in the mesoprotriaenes of the adult specimens 22-55°, on an average
38°. The mesoprotriaenes of the specimens from Station 4193 have exceptionally
large, those of the specimens from Station 2886 exceptionally small, clade-
epirhabd angles. As a rule the three clades of the same cladome are fairly equal
in size; mesoprotriaenes with unequal clades are, however, by no means rare.
Sometimes their inequality is so great that the longest clade of a cladome is

»more than twice as long as the shortest. Sometimes one clade is reduced to a
mere knob, and the spicule appears as a promesodiaene. A few such promeso-
diaenes I found in the spicule-preparations of the young specimen from Station
4228. The rhabdomes of these spicules are much stouter and their clade-angles
much smaller than those of the mesoprotriaenes and it is possible that they are
foreign to the sponge. The chords of the clades of the mesoprotriaenes of the
adult specimens are 60-250 y long, their average maximum length being 161.78
pw. The longest clades are observed in the mesoprotriaenes of the specimens
from Station 2978, the shortest in those of the specimens from Station 4198.
The chords of the clades of the mesoprotriaenes of the young specimen from
Station 4228 are 95-125 » long (maximum average 120 4) and enclose angles of
about 42° with the axis of the epirhabd.

The epirhabd is straight, conical, and pointed. In the majority of the
mesoprotriaenes it is about as long as the clades (Plate 28, figs. 1-6; Plate 32,
figs. 40, 41). In not a few, however, it is either considerably shorter or con-
siderably longer (Plate 28, fig. 7). It is in the adult specimens 25-320 p long,
its average maximum length being here 140.33 4. Of all the spicule-dimensions
the length of the mesoprotriaene-epirhabd is the most inconstant, the differ-
ences of the adult specimens from the nine stations in this respect being very
great indeed. The longest epirhabds are met with in the mesoprotriaenes of
the specimens from Station 4199, the shortest in those of the specimens from
Station 4193. In the young specimen from Station 4228 the epirhabds of the
mesoprotriaenes are 85-100 y long, their average maximum length being 95 yp.
In the immature specimen described by Lambe the rhabdome is 20 y thick, the
clades are 60-90 y long, the clade-angles are 36-47°, and the epirhabd is about
70 » long.

GEODIA AGASSIZII. 127
DIMENSIONS OF MESOPROTRIAENES OF GEODIA AGASSIZITI.
. 3 || From
SiG, Oe sscoqmeonessenbeeues 2886 | 2887 | 2978 | 3088 | 3168 | 4193 | 4199 | 4228 | 4551 |lall sta-|| 4228
tions |}
of apparent-
ly full-grown Dee eee | 416] 6 | 25 | 26
spicules, mm. :
length
average of the
three longest, 4.7 |- 5.5 ysil 5 6 4 5.1
nm,
Rhabd-
ome
of apparent-
= || ies | i |) prs |) ee |) ee SN) es |) ie 7-
peer 8 || 57 | 4g. | 30. |) go’. 20 | 32 | 22 | 28 ay |||
spicules, 1
thick-
ness
average of the
three thick-|| 20 27 34 26 28 19 30 21 21 || 25.11 || 12
est, “
a habe || 29 | 150- | 125- | 60- | 156- | 90- | 100- | 90-.| 90- || 60- || 95-
SS ee S| a 180 | 250 | 150 | 200 | 100 | 210 | 180 | 175 250 || 125
spicules, »
|
length |
average of the || Hine
three longest,|| 120 | 180 | 250 | 143 | 178 | 95 203 | 145 | 142 |/161.78|| 120
Clades #
pee IT as- | 3o- | 32- | ao- | ao- | 22- | 32- | a3- |] 22 |] ,,
Sas Stowe 6 40 50 55 40 55 48 50 55 55 zs
angle | spicules, °
average, ° 26 37 39 39 36 47 36 41 41 38 42
ea iio | 119- | zo- | 60- | 25- | 60- | 90- | 78- | 6o- | 25- || 35-
; 2 2 20 || 10
eames x 150 | 210 | 135 | 150°] 100 | 320°] 125 | 110 || 3 0
Epirhabd |
length
average of the |
three longest, |! 339 | 150 | 192 | 125 | 136 | 80 | 268 | 109 | 93 |{140.33|| 95
“

The anatriaenes (Plate 28, figs. 16c, 18-27; Plate 32, figs. 43, 45, 46) occur
in all specimens. They appear to be particularly numerous in the specimens

from Station 2887.

The rhabdomes of the anatriaenes of the adult specimens

are 4-9 mm. long and at the cladome 10-50 #« thick, their average maximum di-
mensions being 6.5 mm. X 32.56 4. The rhabdomes of the anatriaenes of the
specimens from Station 4199 are considerably thicker than those of the others.
The rhabdome is straight or slightly curved and thicker in the middle than at

128 GEODIA AGASSIZII.

either end (Plate 28, fig. 16¢). The acladomal end usually thins out to a fine
point, it is rarely blunt. The rhabdomes of the anatriaenes of the young speci-
men from Station 4228 are at the cladome 18-27 » (average maximum 25 4)
thick. Among them a few with rhabdomes shortened and terminally thickened,
like the rhabdomes of some of the orthoplagiotriaenes, have been observed.

The basal parts of the clades are curved, concave to the rhabdome, the
distal parts straight. Where the basal curved part passes into the distal straight
part, a slight, abrupt, angular bend is sometimes discernible (Plate 28, figs. 19, 20,
25). In most of the anatriaenes the distal straight part of the clade is about as
long as the proximal curved part (Plate 28, figs. 18-21, 24, 25); in some the
former is considerably longer than the latter (Plate 28, figs. 23, 26; Plate 32, figs.
43, 45,46). Anatriaenes of this kind occur in the specimens from Stations 4199
and 4228. The chords of the clades of the anatriaenes of adult specimens en-
close with the axis of the rhabdome angles of 32-65°, on an average 45.8°. In
the anatriaenes of the specimens from Station 2886 this angle is considerably
larger than in those of the others. The three clades of the same cladome are
usually about equal in length (Plate 28, figs. 19-26). Sometimes, however, their
length is unequal (Plate 28, figs. 18,27; Plate 32, fig.45). In the adult specimen
from Station 3168 and in the young specimen from Station 4228 anatriaenes
with clades of unequal length are relatively more numerous than in the speci-
mens from the other stations. The chords of the clades of the anatriaenes of
the adult specimens are 40-155 y long, their average maximum length being
118.11 ~. In the specimens from Station 2886 the clades of the anatriaenes
are very considerably shorter than any of the others. The chords of the
clades of the anatriaenes of the young specimen from Station 4228 are 60-
110 » (average maximum 95 y) long and enclose angles of 31-45° (average
38°) with the axis of the rhabdome. In the immature specimen described
by Lambe the anatriaenes have a rhabdome 3.3-4.7 mm. by 22-28 yp, clades
45-100 » long, and clade-angles of 41—52°.

In the centrifugal spicule-preparations of this specimen also a few minute
dermal anaclades were observed. These have a rhabdome about 290 # by 1-1.5
eat the cladome and 3-5 » at the thickest point below the middle; their clades
are 4-6 » long; the clade-angles are 38-62°. These spicules may be foreign.

GEODIA AGASSIZII. 8)

DIMENSIONS OF ANATRIAENES OF GEODIA AGASSIZII.

| | |
: | From ||
SUnaA Aaa eRe ee 2886 | 2887 | 2978 | 3088 | 3168 | 4193 | 4199 | 4228 | 4551 |Jall Sta-|| 4298
| tions
| |
| | |
of apparent- j | |]
ly full-grown 5.3-9 ian e wee | Ge 48.2] 4.6 || 4-9 ||
spicules, mm. 2 Eee || |
length = | |
average of the I]
three longest, 8.7 5.4 7603 tf 6 4.6 6.5
mm.
Rhabd-
ome J
: iistee = |} we || We Woe || cee |S Or ere pee) see
9 ©
Saree eal 20 35 35 32 33 40 37 32 50 27
thick-
ness
average of the
three thick-|| 29 35 33 31 32 31 40 36 26 || 32.56|| 25
est, fl
‘a Pe 4o- | 70- | 90- | zo- | 75- | 90- | 100-| 95- | s0- || 40- || 60-
ears 95 150 | 140 | 115 | 155 | 140 | 130 | 140 | 140 || 155 || 110
length ==
average of the
three longest, || 67 133 | 140 | 98 | 145 | 123 | 120 | 130 | 107 |/118.11]| 95
“a
Clades |
et Ritson Hd= | a= |) see || ae |) eee |) spe) cee || Spe ae |p ee || aye
Seeeaor 60 65 | 60 55 53 47 46 51 48 || 65 45
spicules,
angle |
average, ° 55 49 | 43 46 44 43 42 44 46 || 45.8 || 38

The anadiaenes (Plate 28, fig. 28; Plate 32, fig. 44) are rare. They have
been found only in the adult specimens from Station 3168 and the young speci-
men from Station 4228, where also anatriaenes with clades of different length
are more frequent than in the specimens from the other stations. In shape and
size they perfectly resemble those anatriaenes, and I consider them as such
anatriaenes, in which the inequality of the clades is carried to the extent of the
complete suppression of one of them.

The irregular anatriaene-derivates differ from the ordinary anatriaenes by
one of the three clades being directed upwards. These anatriaene-derivates are
rare. I have observed them only in the specimens from Station 4228.

The large choanosomal oxyasters (Plate 27, figs. 3b,6-14b; Plate 30, figs. 1b,
2b, 4, 5, 10b; Plate 32, figs. 4, 6,7) of the adult specimens have from four to
sixteen rays and a small centrum the diameter of which is usually from two to

130 GEODIA AGASSIZII.

three times as great as the basal thickness of the rays. The rays are usually
radial and quite regularly distributed only in the rare, exceptionally large,
few-rayed oxyasters of the adult specimen from Station 4228 is an irregular dis-
tribution of the rays observed. The rays are straight, at the base 0.8-3.2 p
thick and conical. They taper uniformly to the end, which is pointed, blunt,
rounded, or rarely, truncate. The distal parts of the rays are always more or
less spiny. In some oxyasters the spines extend down nearly to the base of the
rays, in others they are confined to the distal two thirds, and in a few —such as
I have found chiefly in the specimens from Station 4199 — they are more or less
restricted to verticils lying just below the tips of the rays. The oxyasters with
spines arranged in this manner appear somewhat acanthtylaster-like. The
oxyasters of the adult specimens measure 9-31 » in diameter, their average
maximum diameter being 24.22 4. Oxyasters more than 26 in diameter with
irregularly distributed rays have been met with only in the adult specimen from
Station 4228. Among the others the specimens from Stations 3168 and 4551
have the largest, those from Station 4199 the smallest oxyasters. The size of
these asters is in inverse proportion to the number of their rays. None of the
oxyasters over 20 y in diameter observed by me had more than nine rays,
all those with ten or more rays being less than 20 in diameter. The large
choanosomal oxyasters of the young specimen from Station 4228 are similar to
those described above, usually have from nine to fourteen rays 0.8-1.7 thick
at the base, and measure 13-25 y» in total diameter. In the immature speci-
men described by Lambe the oxyasters have from seven to nine rays 1.3-2.3 p
thick and a central thickening; their total diameter is 13-20 p.

TOTAL DIAMETERS OF THE LARGE OXYASTERS OF GEODIA AGASSIZII.

2886 2887 | 2978 3088 | 3168 | 4193 | 4199 | 4228 4551 || From all Stations 4228
ara a. = : mere 2
| | 9-31
at) 99 Ss) > 9K 9e,, | ‘ Sy, |esoes=osl! ‘ F
14-212 | 10-22y | 12-23 | 18-25 | 14-26 | 18-24 | 18-20 9-26 » |)maximum aver-|| 13—25p
fe | age 24.22 4

The large oxysphaerasters (Plate 26, fig. 14; Plate 27, figs. 4c, 14c; Plate
30, fig. 3) appear to be more numerous in the specimens from Stations 2886
and 3168 than in those from the other stations. The oxysphaerasters of the
adult specimens consist of a spherical central thickening (centrum), 3.5-11 y in
diameter, from which from fourteen to twenty-eight and perhaps more (it is
exceedingly difficult to count them accurately) equidistant radial rays arise.
These are usually shorter than the diameter of the centrum, regularly conical,

GEODIA AGASSIZIL. 131

1-2 » thick at the base, and sharp or blunt pointed. From the distal parts of
the rays a few quite large spines arise. These are often arranged in a some-
what verticillate manner near the tip of the ray. These oxysphaerasters are
10-21 » in total diameter, the average maximum being 18.2 . In the young
specimen from Station 4228 the large oxysphaerasters have from eighteen to
perhaps thirty rays and measure 12.5-18 u in total diameter, the diameter of
the centrum being rather less than in the oxysphaerasters of the adult specimens
and rarely exceeding 4 ». In the immature specimen described by Lambe

- the oxysphaerasters have from ten to twenty rays, 0.9-2 y thick, the centre is
2.7-7 p, the whole aster 8-21 y, in diameter.

TOTAL DIAMETER AND DIAMETER OF THE CENTRUM OF THE LARGE OXYSPHAER-
ASTERS OF GEODIA AGASSIZII.

SiO s.co.6 SO 9 ReOdDOr 2886 | 2887 | 2978 | 3088 | 3168 | 4193 | 4199 | 4228 | 4551 All Stations 4228

eee) ya |e 18— | 10- | 1s= | a4= | t0= | 15- ae Waps=
eee 5 || 3 |) oy |] a aN TG eye ayy) en | eee | ks
M | average 18.2
Diameter
of ; 3.5-11
an 7-9 |7-7.5|7-11| 7-9 | 6 | 10 | 58 |3.5-7| 68 || maximum || 35-4

average 8.4

The small strongylosphaerasters (Plate 26, fig. 15; Plate 27, figs. 3-14d;
Plate 30, figs. 1a, 2a, 6-9, 10a; Plate 32, figs. 2, 3) are abundant in all specimens.
They consist of a central sphere (centrum), from which from six to twenty
fairly equidistant rays arise radially. The length of these rays is usually smaller,
rarely as great as or greater, than the diameter of the centrum. In the small
strongylosphaerasters of the adult specimens the rays are at the base 0.6-1.6 y
thick and taper towards the truncate end, or are nearly cylindrical (Plate 30,
figs. 6-9). The rays bear small spines, which often appear massed at their ends.
The total diameter of these asters is 3.5-11 , the average maximum being 9.1 y.
The centra are 1.5-6 in diameter, the average maximum being 4.3 4. The
centra of the small strongylosphaerasters of the specimens from Stations 2978
and 4193 are smaller than the centra of those of the others. The small strongy-
losphaerasters of the young specimen from Station 4228 are similar to those
of the adult specimens. They usually have from twelve to nineteen rays
0.5-1 « thick; and a centrum 2-3 «in diameter; their total diameter is 5.5—
9 ». In the immature specimen described by Lambe the small strongylo-
sphaerasters have from ten to twenty-eight rays 0.6-1 y thick, the centre is
2-3.5, the whole aster 5-7 y, in diameter.

GEODIA AGASSIZII.

DIAMETER OF THE CENTRUM OF THE SMALL STRONGYLO-
SPHAERASTERS OF GEODIA AGASSIZIT.

DIAMETER AND

Station || 2886 | 2887 | 2978 | 3088 | 3168 | 4193 | 4199 | 4228 | 4551 See 4228
| Peis a Pe da :
| 3.5-11
jaster || 48.5) 4-9 | 4-10 | 4-11 | 5-9 |3.5-7|4.5-9| j9 | 4,8 || maximum || 5.5-9
| | average 9.1
Diameter amy Ae) pe bat] sae] aay
es | | | 1.5-6
centrum || 5 4 3:5 } 1.5- 5-6| 2-4 ae 25 1.5-4| 3-4.5|| maximum 2-3
ue |} 1 Yen | 3.2 4.5 average 4.2
| | |

The normal sterrasters (Plate 27, figs. 15-19; Plate 30, figs. 11-17; Plate
31, figs. 1, 2, 5-7; Plate 32, figs. 32, 35; Plate 33, figs. 1-8, 12, 13) are abun-
dant in all specimens. The full-grown sterrasters of the adult specimens are
flattened ellipsoids 82-118 y long, 75-100 y» broad, and 58-83 y thick, their
average maximum dimensions being 103.55 by 88.56 by 69.22 4. The largest
are those of the specimen from Station 3088. The proportion of length to
breadth to thickness is fairly constant. ‘Those of the specimen from Station
3088 are relatively somewhat longer and those of the specimens from Stations
In the

young specimen from Station 4228 the full-grown sterrasters are similar in

2886 and 4228 relatively somewhat thinner than those of the others.

shape, 76-100 # long, 70-85 y broad, and 60-70 » thick, their average maximum
dimensions being 95 by 83 by 68 #. In the immature specimen described by
Lambe the sterrasters are 90-110 by 74-92 by 67-75 yw. (See table p. 133.)

On one of the two broader sides of the normal full-grown sterrasters an
umbilicus, usually more or less circular in outline, 12-15 in transverse diameter,
and 6 4 deep, is observed (Plate 27, figs. 15-19; Plate 31, figs. 1,2, 5-7). With
the exception of a small, smooth, central patch at its bottom, the wall of the
umbilical pit appears to be roughened (Plate 31, figs. 5, 7). I am not quite
positive, however, whether there really is a roughness there, it being quite
possible that its appearance in this place may be an optical illusion, caused by a
refraction at the surface of the umbilical pit, that in fact this apparent roughness
is in reality nothing but a blurred ultraviolet light-image of the rays and spines
on the other side of the surface, which are traversed by the light before it reaches
the umbilicus. Observations with high powers in ordinary light failed to decide
this question.

From the whole of the surface of these normal, full-grown sterrasters, with
the exception of the part occupied by the umbilicus, the distal ends of the rays

GEODIA AGASSIZIL. 133

composing the sterraster protrude a short distance. The freely protruding
distal parts of these rays are usually circular or somewhat polygonal, four- to
seven-sided, in transverse section, 1.3-4 y» thick, regularly distributed, and
hardly 1 » apart. They are truncate, and from the margin of their terminal
face a verticil of from four to seven, most frequently six, spines arises. These
spines extend either transversely, vertical to the axis of the ray, or, less fre-
quently, obliquely outward and a little upward. The spines of the rays remote
from the umbilical pit are stout, straight cones, about 1.7 » long and 1.3 y thick
at the base (Plate 31, figs. 1, 2, 6, 7; Plate 33, figs. 12, 13). Those of the
spines of the rays surrounding the umbilical pit, which extend towards the
umbilicus, are often larger, as much as 2.5 » long, and not regularly conical but
irregular, their ends being broad and sometimes covered with small, secondary

spinelets.
DIMENSIONS OF STERRASTERS OF GEODIA AGASSIZII.

raHOUer ee tet oso. 2886 | 2887 | 2978 | 3088 | 3168 | 4193 | 4199 | 4228 | 4551 Netti 4228
z | 90- | 97- | 90- | 110- | 100- | 82- | 85- | 95- | 95- || 82- || 76-
ice || 100) | 104 | 112 | 118° | 110 | 97 | 97 | 100 | 110 || 118° || 100
sterrasters, /«

Length |
average of the |
three longest, || 96 | 102 | 111 | 118 | 109 | 94 | 97 | 98 | 107 ||103.55|| 97
L
3d ae estas Si= ONS SO— Negb—n| 60" |) 8s— eile rp=
aa || ss | 90 | 100 | 95 | 96 | 83 |- 88 | 90 | 90 || 100 |! ss
sterrasters, 1

Breadth
average of the
three broad-|| 82 | s8 | 94 | 94 | 94 | 82 | 85 | 88 | 90 |/88.56]] 83
est, yt |
of apparent | 5s- | 65~- | 74- | 78- | @5- | 63- | 67- | 58- | 69- |] 58- || 60
pedo (60) | 72) |) 7 | 93 | vo | 70 |. 70 | 64 | 72 I) S83 Il 70

e sterrasters,

Thick- = —s

Bes average of the
three thick-|| 59 | 71 | 75 | 81 | 7o | 66 | 69 | 61 | 72 ||/69.22]| 68
est, fs

Besides these normal forms of full-grown sterrasters, which form the great
majority, some others with fewer and usually stouter protruding rays and more
numerous or larger and differently shaped spines, which I propose to name
sterroids (Plate 31, figs. 3, 4, 8-10; Plate 32, figs. 13-28, 33, 34, 36-39; Plate
33, figs. 9-11, 14), are met with.

134 GEODIA AGASSIZII.

In one kind of sterroid (Plate 31, figs. 3, 4; Plate 32, figs. 33, 34, 36-39;
Plate 33, fig. 10) the free distal parts of the rays are considerably thicker, 4—
13 win transverse diameter, and farther apart than in the normal sterrasters.
Some of them, chiefly those surrounding the umbilicus, but also others, have
an irregularly elongated transverse section. In these sterrasters the thick rays
bear, besides a terminal verticil of from eight to fourteen lateral spines, rather
larger than those of the normal sterrasters, several others which arise obliquely
from their terminal face.

In another form of sterroid (Plate 31, figs. 8-10; Plate 33, figs. 9, 14)
the freely protruding distal parts of the rays are not very much thicker,
but very much farther apart than in the normal sterrasters, and provided with
very different spines. The part of the surface of the solid centrum lying between

» the protruding rays is in these sterroids covered by large numbers of small
projections of various shape, and appears irregularly granular. Each ray bears
from two to eight mostly lateral, but in part also terminal, spines, which are
2.5-5 w long, up to 3 yw broad, slightly curved down at the end and covered
with numerous small secondary spinelets. When viewed from above, the
spines are somewhat similar to serrated leaves.

In a third kind of sterroid (Plate 32, figs. 25-28; Plate 33, fig. 11) the
rays are thicker, up to 15 in transverse diameter, and farther apart than in
the forms above described. They are terminally rounded and covered with large
numbers of recurved, somewhat claw-like spines. In most of these sterrasters
the rays are all fairly equally developed, their free distal parts covering the whole
of the solid centrum of the spicule and protruding equally far beyond it (Plate 32,
figs. 25, 26). In some, however, there are only a few groups of protruding rays,
the greater part of the surface of the centrum being destitute of such (Plate 32,
figs. 27, 28) but covered with groups of spines similar to the spines on the pro-
truding rays.

In the specimens from Station 4228, both the young and the adult, I have
found a few oxysphaerasters, about as large as the sterrasters, three in the
former and one in the latter. These spicules have from thirteen to fifteen
straight, conical and smooth, radial and concentric, rather irregularly distributed
rays. They measure 90-100 in total diameter; the diameter of the centrum is
25-27 w; the rays are (without the centrum) 35-40 «long and 10-17 yp thick at
the base. I found these asters in situ in sections in the subcortical layer and
I do not think that they are foreign to the sponge. For the reasons given
below, I considered them as sterroid-derivates.

GEODIA AGASSIZII. 135

In the choanosome of several of the specimens of Geodia agassizii, particu-
larly the adult one from Station 3099, and the young one from Station 4228,
numerous young sterrasters in various stages of development have been observed.
In the adult specimens all these young sterrasters exhibit the well-known
form of spheres composed of very numerous, regularly distributed, slender,
radial rays. In the young specimen from Station 4228 young sterrasters of this
kind are also abundant, but here besides these ordinary forms numerous other
asters, similar in dimension, but very different in appearance occur. <A close
examination showed the latter to be the young forms of the sterroids. By
comparing a large number of these and the ordinary young sterrasters with
each other and with the fully-developed ones I was able to trace the develop-
ment both of the normal sterrasters and the sterroids.

On Plate 30, figs. 11-17, Plate 32, figs. 29-31, and Plate 33, figs. 1-8,
two series of developmental stages of the normal thin-rayed sterrasters, the first
from an adult specimen from Station 3188, the second from the young specimen
from Station 4228, are represented. One of the youngest stages observed, Plate
33, figs. 1, 2, is a sphere 17 » in diameter. This young sterraster consists of
about 460 straight and exceedingly slender, thread-like, concentric rays, which
are equal in length, regularly distributed, and jointed proximally. Adjacent rays
enclose angles of about 10°. As the spicule grows these rays increase in thick-
ness, their proximal parts thickening first and this process of thickening then
extending distally. The basal thickening parts of the rays coalesce, as they
come in contact with each other, to form a solid centrum. In the next stage
(Plate 33, figs. 3, 4) the basal parts of the rays protruding from the solid cen-
trum thus formed show a slight bulbous thickening which increases so that in
young sterrasters 50 » in diameter (Plate 33, figs. 5-7) the rays are at the
base already 2 y thick, their distal ends, however, being still quite thin. In this
stage each ray appears as a cone, widened below like a bulb, and drawn out to
a fine thread above. As development progresses the thickening of the rays
extends farther and farther towards their distal ends (Plate 30, figs. 11-13), the
whole sterraster and its solid centrum increase in size, and the fine terminal
points of the rays become shorter and shorter, until they are entirely enveloped
in the ascending thickening and thus altogether disappear. In this stage (Plate
30, fig. 14; Plate 33, figs. 7, 8) the young normal sterrasters appear as solid, at
first still fairly spherical, central masses from which cylindrical rays, which
stand close together and are about 4 y thick and simply rounded at the end,
protrude. Without changing much in appearance, these young sterrasters in-

136 GEODIA AGASSIZII.

crease in size and begin to assume the flattened ellipsoidal appearance of the
full-grown ones (Plate 32, figs. 29-31). Then the basal parts of the protruding
rays coalesce farther and spines begin to grow out from the margin of their
terminal face (Plate 30, figs. 15, 16). These spines at first appear as small
rounded knobs. Later (Plate 30, fig. 17) the rays are thickened terminally
and the spines grow in length. They are in such young sterrasters very slender
and do not attain their full thickness for some time.

The early stages of the sterroids differ from those of the normal sterrasters
described above by the rays composing them being not nearly so numerous and
usually also not so regularly arranged. In accordance with the smallness of
the number of the rays the angles between them are much larger than in the
young forms of the normal sterrasters. The thickening and concrescence of
the basal parts of the rays, which in the latter lead to the early formation of
a solid centrum, here therefore does not have this effect until a very much later
stage. Young stages of the thick- and few-rayed sterrasters 60-70 y in diam-
eter (Plate 32, figs. 13-20), which correspond to the stages of the normal sterr-
aster represented on Plate 32, figs. 29, 30, accordingly have a much smaller
centrum and much longer cylindroconical, terminally rounded protruding rays.
As stated above these rays are often irregularly distributed and the angular
distances between them are very unequal. Rays of such young sterrasters
standing particularly close together coalesce as early as all rays of the normal
sterrasters do, whereby the irregularity in the appearance of these spicules is
greatly enhanced (Plate 32, figs. 15, 16, 19, 20). The distal parts of the rays
cover themselves with numerous small spines, which later grow in size. Sub-
sequently, through the continued thickening and concrescence of the basal
parts of the rays, the centrum increases in size (Plate 32, figs. 21-26). Finally
sterroids are formed, the centra of which are as large as the centra of the normal
sterrasters, the surface of which, however, bears much fewer and usually thicker
rays covered with a much larger number of spines. While the rays of the
normal sterrasters are nearly always equal in length, the rays of the sterroids
are occasionally unequal. In such sterroids with rays unequally long the con-
crescence may reach up to or even beyond the shorter rays, which are then
totally enveloped by the mass of the centrum, their positions being indicated
in the full-grown sterroid only by the groups of spines on the parts of the sur-
face of the centrum free from protruding rays. In this way irregular few- and
thick-rayed sterroids like the one represented on Plate 32, figs. 27-28 are formed.

The sterroids differ from the normal sterrasters accordingly not only when
full grown but also, and even to a greater degree, when young.

GEODIA AGASSIZII.

137

The large oxysphaerasters of the specimens from Station 4228 above re-

ferred to, which in their dimensions equal the sterrasters, are rather similar to

some of the young stages of the sterroids. They may therefore be such spic-

ules, in which the development has, as it were, gone wrong, the thickening of

the rays, which normally leads to the formation of a large solid centrum, and

the formation of the spines on the rays having been suppressed.

LOCALITIES AND NATURE OF ENVIRONMENT.

abe és Se
52 ; Be 52
< = Locality Date Depth a ey Bottom 25
Sn a = 2
Zz 2 Za
Off Oregon, 43° 59’ N., 124° 9°
2886 56’ 30” W. 19 Oct., 1888 | 91 m. (50 f.) (48.1° F) Rocky 8
Off Oregon, 43° 58’ N., 124° s EC .
2887 57’ W. 19 Oct., 1888 | 77 m. (42 f.) (47.1° F.) Clay, pebbles 1
Off southern California, 33° aa S 13.6° F
2978 59’ 45” N., 119° 22’ 15” W. 12 Feb., 1889 | 84 m. (46 f.) (56.5° F.) Gray sand 2
Off Oregon, 44° 28’ N., 124° 4 8°
3088 25/30" W. 3 Sept., 1889 | 84 m. (46 f.) (46.3° F.) Clay, pebbles 1
Off central California, 38° 01’
, 2 ) —
3168 25” N., 123° 26 55” W. 24 Mar., 1890 | 62 m. (34 f.) Rocky coral 1
Gulf of Georgia: Halibut
Bank; Cape Roger Curtis, | , 33-40 m 10.2°- é
ee tieowentisiand. S. sg° B29 |)72 9M 13) a 53 fy |(50.3° F.)| Pe erermsand) 2
km. (10.8 miles), drift S. 1° E.
Queen Charlotte Sound: off
Fort Rupert, Vancouver Is- ' 154106 an. 77° Soft green mud
4199 | land, B. C. Centre of Round | 25 June, 1903 (68-107 f.) 45.9° F.)| and volcanic 4
Island. S. 46° W., 11.5 km. Se Mets (GG eee
(6.2 miles), drift S. 85° E.
Vicinity of Naha Bay: Behm
Canal. S. E. Alaska, Indian 75-245 m. 8.8° Gravel and 5
coe Point. N. 18° E., 1.7 km. (0.9 Ee Le (41-134 f.) |(47.S° F.)| sponge spicules] ~
miles), drift N. 2° W.
Monterey Bay, Cal.; Point Gharces Renna
4551 | Pinos Light House. S. 9° E..|7 June, 1904 | 102 m. (56 f.) — shells eae 2 1
8.4km. (4.5 miles) drift S. 37°E. a :
oa Houston Stewart Channel, 1893 =e oe Aas 1

Queen Charlotte Island.

The young stages of the sterroids have been observed only in the young
specimen from Station 4228 and in the immature specimen described by

138 GEODIA AGASSIZII.

Lambe. They are much more numerous in the former than in the latter.
Fully developed sterroids occur sparingly in all adult specimens. In other
Geodidae, where I have found them, they also occur in small numbers. These
facts lead me to consider the sterroids as spicules produced, like the milk-teeth
of mammals, in the immature state only. If this assumption is correct, the
sterroids might be similar to the ancestral form of the normal sterrasters, and
represent a link connecting the latter with the sphaerasters from which I should
be inclined to derive them.

To simplify the references to these sponges, I will, in the following discussion

of their relative systematic position, designate: —

Those from Station 2886 as A Those from Station 4193 as F

MM ie eat cy ee eT «iggy ng
EN Me Oe tyler oY Ao ae
RD opie ogee OSS e aL) CS AB acme
ry Ace A OLAS E CAE)

To avoid confounding characters immature in nature with the systematically
important peculiarities of the full-grown sponges, the young specimen from
Station 4228 and the immature specimen described by Lambe are not taken
into account in the following discussion.

In the character of their shape these sponges are very similar, their differ-
ences in this respect not exceeding the individual variations usually met with in
the species of geodine sponges. In the structure of their canal-system, their soft
parts, and the general arrangement of their skeleton they are also uniform. In
their colour and in the shape, relative frequency, and dimensions of their spicules
however, only the specimens collected at one and the same station agree. The
differences in the colour of specimens from different stations are certainly very
considerable. But since it is very likely that these differences have been pro-
duced post mortem through differences in the external influences to which the
different lots were exposed after capture, they are without systematic signifi-
‘cance. The differences in the spicules, on the other hand, are systematically
important, and it is therefore necessary to study them with care if we wish to
decide in what systematic relation these otherwise similar sponges stand to each
other.

The differences in the shape and relative frequency of the spicules of the
several lots of these sponges are as follows: in EZ and F both thick club-shaped
and thin cylindrical styles are met with in small numbers; in A, D, and G thin
ones only; and in B, C, H and I no styles at all. In E, F, and G the ortho-
plagiotriaenes with clades either abruptly bent down at the end or terminally

GEODIA AGASSIZII. 139

branched are more numerous than the orthoplagiotriaencs with simple and
regular clades. In the others all, or nearly all, orthoplagiotriaenes have simple
and regular clades, orthoplagiotriaenes with clades like those of the majority
of EL, F, and G being very rare or absent altogether. In the clades of the ana-
triaenes of G and H the straight distal part is considerably longer than in those
of the others. In E and H anatriaenes with clades unequal in length or position
have been met with. In some of the anaclades of EF one clade is absent alto-
gether, these spicules appearing as anadiaenes. Oxyasters over 26 in diameter
with irregularly arranged rays have been found only in H. The small strongylo-
sphaerasters of C and F have on an average smaller centra than those of the
others. In A and E£ the large oxysphaerasters are much more numerous than
in the others.

In dimension the spicules of the lots from the different stations differ more
or less. To obtain a base for studying these differences the sets of measure-
ments of the fifteen spicule-dimensions obtainable with the greatest accuracy
of each of the nine lots of adult specimens from the nine stations, were selected
for further study. These dimensions are:— the diameter of the oxyasters;
the length and breadth of the sterrasters; the length and thickness of the large
choanosomal amphioxes, the small dermal amphioxes, and the rhabdomes of the
orthoplagiotriaenes; the thickness of the rhabdomes of the anatriaenes and
mesoprotriaenes; the length of the clades of the orthoplagiotriaenes, anatriaenes,
and mesoprotriaenes; and the length of the epirhabds of the mesoprotriaenes.
Of each of these (15 x9 =) 135 sets of measurements the largest alone were taken
into account. Of three, the length and thickness of the small dermal amphioxes
and the diameter of the large oxyasters, the largest dimension observed in each
lot was taken by itself. Of the twelve other dimensions averages of the three
largest dimensions measured were taken. These single maxima and maximum
averages of three of the fifteen spicule-dimensions taken into account in the nine
lots are given in IV of the subjoined table. From these maximum averages
(maxima) the general averages (means) were taken. These general maximum
averages (means) are given in II of the table. The deviation of the average
maximum (maximum) of each of the fifteen dimensions of each of the nine lots
(IV) from the mean (general maximum average, II) of the same dimension was
ascertained by subtraction. These (135) deviations are given in V of the table.
Referring as they do to spicule-dimensions of very different size the numbers
giving these deviations are not commensurate and directly comparable with
each other. To obtain numbers expressing these deviations from the general

140

GEODIA AGASSIZII.

maximum averages (per cent) in a commensurate manner, the number, different
in each case, with which the general maximum average must be multiphed to
make the product 100 mm. was ascertained by dividing 100 mm. by the general

maximum average of the dimension in question (II).

With these numbers,

which are given in III, the deviation of each dimension from the means (V) was
then multiplied. The product thus obtained is the percentage (commensurate)
deviation given in VI.
To bring out more clearly the significance of the commensurate numbers
given in VI, I have represented the variations of the spicule-dimensions ex-

=
i=

IV

Spicule-dimensions

General maximum averages.

Averages of the three largest dimensions measured of the large am-
phioxes, the orthoplagiotriaenes, mesoprotriaenes, anatriaenes, and
sterrasters; and maximum dimensions measured of the small dermal
amphioxes and the oxyasters.

Number with which to multiply the general
average dimension (col. IT) to make it 100 mm,
(100 mm, divided by the number in col. IT).

| |. A B Cc D E y G H I
| Station |Station Station Station Station |Station, Station |Station |Station
| |} 2886 | 2887 | 2978 | 3088 | 3168 | 4193 | 4199 | 4228 | 4551
|
}
length mm. 3.9 || 25.65 || 3.4 | 38 | 41 4.3. | 4.2 | 46) 1) S330) seme ars
Large amphioxes
thickness /£ || 86.3 ||1158.75 75 93 ill 76 89 87 79 74 93
| length 4 34.2 292.4 480 340 270 360 290 390 290 330 330
Small dermal amphioxes —_—— =
thickness /¢ || 9.06 )|11037.5 12 10 8 10 7 10 8 8.5 8
length mm. || 3.39 29.5 || 3.1 3.1 4 3.6 3.9 3.6 3.7 2.5 3
Orthootael | rhabdome | ~
trigenee thickness # || 115.44 || 866.2 || 103 | 110 | 139 | 120 | 132 | 92 | 102 98 | 143
clades length /£ 490.89 || 203.72 | 467 507 483 493 483 520 540 497 428
———_—_——_]} — |
rhabdome thickness /£ || 25.11 3982.48. 20 27 34 26 28 19 30 21 21
Mes i- sy ees
acne | clades | length || 161.78 || 618.13 || 120 | 180 | 250 | 143 | 178 95 203 | 145 | 142
| epirhabd length / 140.33 || 712.59 || 110 150 192 125 136 80 268 109 93
—_—_ —_—_— — |
rhabdome thickness /£ 32.56 ||3071.26| 29 35 33 31 32 31 40 36 26
Anatriaenes — |
clades | length || 118.11 || 846.66 67 133 140 98 145 123 120 130 107
Large oxyasters total diam-" | 24.22 |larza.ea] 21 | 22 | 23 | “2s\'| ‘25 | 24) eons Gren
‘length #2 || 103.56 || 965.66 | 96 | 102 | 111 | 118°| 109 | 94 97 98 | 107
Sterrasters |
|| 88.56 | 82 88 94 94 94 82 85 88 90

breadth #

|}1129.24

-

GEODIA AGASSIZII.

14

1

pressed by these numbers also graphically. This graph, Figure F, was obtained
by erecting fifteen ordinates, representing the fifteen spicule-dimensions under
discussion, at equal intervals on a baseline. These ordinates are arranged in the
order of the maximum variations of the dimensions they represent, the one to
the left representing the most constant dimension. The mean general averages
(II) multiplied by the corresponding numbers in III of course all gave products
of 100 mm. The points graphically representing these numbers, plotted on
the fifteen coordinates 100 mm. from the axis, all lie in the straight horizon-
tal line m. This line graphically represents the ideal mean of all the sponges

vi

Deviations of the maximum averages (maxima) in col. IV from the
mean general maximum average in col. II

Commensurate percentage expression of the deviations in col, V
obtained by multiplying the numbers in col. III and col V.

mm,
-

fatten Btatlon Statfon Btation Btation Station Station Station Station |Station Station|Starion Station Station Station Station Istation Station
2886 | 2887 | 2978 | 3088 | 3168 4193 4199 4228 551 || 2886 2887 2978 3088 3168 4193 4199 4228 4551
05 —0.1 | +0.2 | +0.4 | +0.3 40.7 —0.1 0.7 0.1 12.82) —2.56 | +5.13 | +10.26) +7.69 | +17.96] —2.56 |—17.96| —2.56
13 +6.7 | +24.7|—10.3|} +2.7 | +0.7 —7.3 |—12.3| +6.7 —13.09| +7.77 +28.62|—11.04 + 2.67 | +0.81 | —8.46 |—14.26] + 7.77
+138 —2 —12 +18 | —52 +48 —52 —12 ze +40.35) —0.58 |—21.06) +5.27 |—15.21| + 14.04/—15.21| —3.51 | —3.51
+2.94 +0.94 | —1.06 | + 0.94 | —2.06 | +0.94| —1.06 | —0.56 | —1.06 |+32.45) +10.38/—11.69) +10.38 — 22.74) + 10.38|—11.69| —6.25 |—11.69
—0.29 | —0.29 | +0.61 | +021 | +0.51 | +0.21 +0.31 0.89 0.39 8.56 | 8.56 | +18.00) +6.2 +15.05) +6.2 | +9.15 |—26.26/—11.51
= jeeee* Ie 2 =.

—5.44 | + 23.56) +4.56 | + 16.56|—23.44] —13.44 |—17.44| + 27.56||—10.78| —4.72 + 20.41) +3.97 | +14.35 oe ea = 15-10) 423.88

+16.11) —7.89 | +2.11 | —7.89 | +29.11| +49.11 | +6.11 |—62.89 —4.86 | +3.28 | —1.61 | +0.43 | —1.61 +5.93 | +10.00 +-1.24 |—12.97

+1.89 | +8.89) +0.89| +2.89 | —6.11 +4.89 |—4.11 | — ils 20.35 +7.53 3° 35.41 $3.55 | + 11.51 24.34 +19.46 —16.39 —16.39

+18.22) + 88.22|—18.78) + 16.22|\66.78| +41.22 |—16.78|—19.78 25.83 +11.26) +54.53 11.61 +10.03/—41.28| + 25.48|—10.37/—12.23

+9.67 | +51.67\—15.33) —4.33 |—60.33| +127.67 —31.33 —47.33 —21.62) +6.89 |+36.82 10.98 —3.09 | 42.99 + 90.98/—22.33|—33.73

+2.44 | +0.44 | —1.56 | —0.56 |—1.56| +7.44 | +3.44 —6.56 '|_10.94| +7.50 | “41.35 | —4,79 | —1.72 | —4,79 422.85) + 10.57|;—20.15

+14.89) + 21.89|—20.11| + 26.89) +4.89) +1.89 | +11.89/—11.11 43.28 $12.61) +18.54|—17.03 peel +414) +1.60 | +10.07 — oat

— 2.22 | —1.22| +0.78 | +1.78 | —0.22| —4.22 | +6.78| +1.78 12.0 —9.17 | 5.04 +322) +7.35 | —0.91 |—17.42) +28.0| +7.35
“7.58 —1.56 | +7.44|+14.44) +5.44/—9.56| —6.56 |—5.56| +3.44 ] —7.30 | —1.51 | +7.19 | +13.94 +5.26 | —9.23 | —6.34 | —5.37 | +3.32
6.56 —0.56 | +5.44) +5.44|) +5.44|/—6.56| —3.56 |—0.56| +1.44 |—7.41 | 0.63 +6.15 |) +6.15 | +6.15 | —7.41 | —4.02 | —0.63 | +1.63

142 GEODIA AGASSIZIT.

of this group examined in respect to the fifteen spicule-dimensions under dis-
cussion. From the fifteen points forming this line of mean general averages,
the numbers in VI, giving the deviations from the means in a commensurate
manner were plotted on their respective ordinates,— above if positive and below
if negative — and designated with the letters A —I standing for the lots from
the nine different stations. The points with the same letter, commensurately
representing the different dimensions of the spicules of the same lot, were then
connected by lines. The nine lines (curves) thus obtained are maximum spic-
ule-dimension curves and graphically represent the peculiarities of each of the
nine lots in respect to the fifteen spicule-dimensions selected.

The nine stations where these sponges were collected are situated on the
Pacifie coast of North America, between 33° and 56° N. They are not uniformly
distributed but form five groups separated by wide intervals, and thus these
sponges may be said to come from five distinct regions. These are, from north
to south: 1. Behm Canal Station (4228, H); 2. Vancouver Island Stations
(4193, 4199, F. G.); 3. South Oregon Stations (2886, 2887, 3088, A. B. D.);
4. Middle California Stations (3168, 4551, E. I.); and 5. South California
Station (2978 C).

The spicule-curves, Figure F, pertaining to the lots from these five different
regions are differently drawn as follows: Behm Canal (H, 4228) ------ :
Charlotte Sound and Gulf of Georgia (F, G, 4193, 4199) —-——---; coast
of southern Oregon (A, B, D, 2886, 2887, 3088) ——————-;; coast of middle
California (E, I, 3168, 4551) ——-—-; and coast of southern California
(C, 2978) —— —

GEODIA AGASSIZII. 143

(20° Ww

~* 4228 (mn)

2s
= Sag y sone nea nan aan ans on = = G49 O1(G)
2 2 ORCS RES | 419316)

oe ee SE ee Oe 3088 (0)
i ofon-orrriiiiiii 988 6 (4) 298718)
ow
Oe ees eee IL
- 455108
292B(c)
20°

Fig. E. Pacifie coast of North America, showing Stations of Geodia agassizit. A. Station
2886. B. Station 2887. C. Station 2978. D. Station 3088. E. Station 3168. FP.
Station 4193. G. Station 4199. H. Station 4228. TI. Station 4551. (See page 144.)

ASSIZIT.

\ODIA AG

GE

144

‘saAmmo anoidg

‘nzsspbp mypoay *f “BLT

‘6 IL poytuseur
SyZuey ‘pqeysida ‘souovnosdosoyy

‘eT'S19 poyruseur
SyqZuoey] ‘sepeo ‘souovtordosoyy

‘99°9F8 poyruseur
Sy Sue] ‘sopeyo ‘souowmyeuy

‘P'26S poywusour
‘yjSue] ‘soxorydure [euep [peug

‘SP'Z86E poytuseur

‘ssouyory} ‘oumopqeys ‘souseujosdosayy,

'¢-LEOLT poyrusvur
fssouxorg} ‘sexorydure jeuep [peug

“FRSSLP poyruseur
{19} 9UIBIP [B10] ‘s1ajsBAXO aDIvT

‘6% poeytuseur

SyqZuey ‘omopqeys ‘sousvyomedoyywG

°Z'998 poytuseut

fssauyory} ‘euopqeys ‘sousvujomeldoyyO

‘97 TLE poytusvur
{ssouxyoryy ‘owopquys ‘souoeuyeuy

"CL'SSTL poyrudeur
‘ssouyory} ‘soxordure os1eT

"¢9'¢7 poytuseur
fyqaue] ‘sexoryduie ad1eT

*99°G9G poyluDeut
Sy yBugy] ‘s19}SB.110}9

‘ZL 20% poytuseur
fyqdug] ‘sopeyo ‘souovnjomeldoyysO

‘PSGSIT poylaseur
‘UPB ‘S1a}SBIIIIS

GEODIA AGASSIZII. 145

The graph shows that only very few of the maximum averages (maxima)
of the same dimensions of spicules from different lots are identical with each
other, and that not a single one coincides with the mean. The extent of their
maximum and average deviations from the mean are tabulated below.

HES (EVO | Total range of | Averages of the
aoe | cecaione iia aulneee
Dimension above the mean | below the mean
per cents of the respective dimensions

Breadth of sterraster 6.1 7.4 13.5 4. 46
Length of Pea inamiotrincne-ciade 10 | 12.8 22.8 4.64
Length of sterraster 13.9 f 9.2 ; 23.1 6.61
Length of large amphiox 17.9 17.9 35.8. | 8.83

Thickness of large amphiox 28.6 14.3 i 42.9 “10.60 oe
Thickness of anatriaene-rhabdome 22.8 20.1 | 42.9 9.41
Thickness of orthoplagiotriaene- 23.9 | 20.3 44.9 13.91

rhabdome
Length of orthoplagiotriaene-rhabd- 18 j 26.3 44.3 12.17
ome

Diameter of oxyaster 28 17.4 | 45.4 | 10.20
Thickness of small dermal amphiox 32.5 22.7 ri 55.2 14.18
ea of mesoprotriaene rhabd- 35.4 24.3 | 59.7 17.21
Length of small dermal amphiox 40.4 21 61.4 13.19
Length of anatriaene clade 22) | on en ees
Length of mesoprotriaene clade 54.5 41.3 | 95.8 | 22.51
Length of mesoprotriaene epirhabd 91 . 43 134 | 29.93

Having thus ascertained the differences in the spicules and stated them in a
manner suitable for discussion, the question of their systematic and zoégeographic
significance may be taken up.

There can be no doubt that external forces, acting on the growing sponge,
exert an influence on the shape and the dimensions which the spicules attain.
We know that at the stations where these sponges were obtained the bottom
temperature and the nature of the bottom differ more or less. From this it
follows that the forces which acted on the several lots during growth were,

146 GEODIA AGASSIZII.

generally speaking, different, and we may therefore expect to find certain dif-
ferences in the spicules. The question therefore arises whether all the differences
observed are entirely due to the differences of the external forces which acted
on the growing sponges, or whether they are in part germinal (hereditary)
in character.

None of the differences in the shape and relative frequency of the spicules
seems to me to exceed the limits allowable for such somatic (individual) non-
germinal (non-hereditary) variation. The average and maximum deviations
of the dimensions of the oxyasters, sterrasters, large amphioxes, and ortho-
plagiotriaenes from the mean are only 4.46-13.91 and 6.1-28.6 % respectively,
and also fairly within these limits. The average and maximum deviations of the
dimensions of the small dermal amphioxes, anatriaenes, and mesoprotriaenes,
which range from 9.41 to 29.93 and from 20.1 to 91 % respectively, are so con-
siderable that at first sight it seems necessary to consider them as germinal.
The greatest deviations are observed in the length of the clades of the anatriaenes
and the clades and epirhabds of the mesoprotriaenes, the total range of maxi-
mum deviation of the latter exceeding 130 %.

The position of the oxyasters, sterrasters, orthoplagiotriaenes, and large
amphioxes is different from that of the small amphioxes, anatriaenes, and pro-
triaenes. The former lie within the sponge and are thus to a certain extent
sheltered from the influence of the external forces; the latter lie superficially
and are protruded more or less beyond the sponge and thus more exposed to the
external forces. These external forces are different in the nine stations. It is”
therefore only to be expected, that the last named, exposed, spicules on which
the external forces act more directly, should be much less constant than the
first named, sheltered ones, on which they act more indirectly.

These facts and considerations clearly show that the peculiarities of the
internal spicules must be systematically much more important than the pecu-
liarities of the protruding ones. I therefore thought it desirable to ascertain
what systematic result an examination of the internal spicules by themselves —
leaving the external ones out of account — would lead to. To do this I selected
the most accurately determinable dimensions of the internal spicules, namely
the length of the sterrasters and orthoplagiotriaene-clades and the length and
thickness of the large choanosomal amphioxes and orthotriaene-rhabdomes.
These six dimensions of each lot I compared with the six corresponding dimen-
sions of each of the other eight lots. In each of the thirty-six possible pairs
(combinations of nine in the second class without repetition) I added up the

GEODIA AGASSIZII. 147

differences of the commensurate proportional amounts of the maximum averages
of the homologous dimensions and then divided the sums by six. In this way
the average percentage differences of the nine lots in respect to the dimensions
of these internal spicules, least subject to the influence of external forces, were
obtained. In the following table, where the thirty-six pairs are arranged in the
order of their similarity in respect to these dimensions, the numbers thus arrived

at are given.

Average percentage differ- | Average percentage differ- Average percentage differ-
ences in the length of the ences in the length of the ences in the length of the
sterrasters, the orthopla- sterrasters, the orthopla- sterrasters, the orthopla-
Pairs of giotriaene-rhabdomes, Or-| pairs of giotriaene-rhabdomes, or- Pairscof siotriaene-rhabdomes, or-
thoplagiotriaene - clades, thoplagiotriaene - clades, thoplagiotriaene - clades
lots. 5 lots. ' lots. “ 4
and large amphioxes, and and large amphioxes, and and large amphioxes, and
the thickness of the ortho- the thickness of the ortho- the thickness of the ortho-
plagiotriaene - rhabdomes plagiotriaene - rhabdomes plagiotriaene - rhabdomes
and large amphioxes. and large amphioxes. and large amphioxes.
AH 6.1 1H 11.6 DI 15.6
CE 6.5 BH 11.9 FH 16.2
DE 7.8 DF 12).2 AE 16.8
FG 8.1 BD 12.3 DH Wie
AG 8.2 CI Pte GI 17.4
AB 8.5 EG 1PARTS CG 18.4
BG 8.7 EF 12.9 HI 18.8
BI 8.7 AD 13.3 CF 19.5
DG 10.7 AF 13.6 FI 20.1
EI 10.7 CD 13.7 EH 21.1
BF 11.3 Al 14.6 AC 22.5
BE 11.6 BC 15.6 CH 26.8

The table shows that these differences are very unequal and vary between
6.1% (AH) and 26.8 % (CH). It would therefore seem that it might be possible,
by joining the pairs differing only slightly, and by keeping apart the pairs
differing greatly, to arrive at some systematic grouping. The regular gradation
in the increase of the average percentage difference from the lowest to the highest,
however, makes it difficult anywhere to draw a line of demarcation between
differences insufficiently and differences sufficiently great for systematic dis-
tinction. If we consider, as we surely may, differences in respect to these
dimensions up to 10.7 % insufficient for systematic distinction, we must unite
all the pairs differing 10.7 % or less. These pairs are AH, CE, DE, I'G, AG,
AB, BG, BI, DG, and EI. Now if these pairs are united: A joined to H, G, and
B; CtoE; DtoE, andG; FtoG; BtoG, andI; and E to I; it will be seen that
all nine lots are united, however different certain pairs may be. In truth, we
can say that biometrically a continuous series of transitional forms connect the
most aberrant members of the whole group.

One of the most remarkable features of the peculiarities of the spicules of
the nine lots is the want of correlation between them. So far as their dimensions

148 GEODIA AGASSIZII.

are concerned this want of correlation is most clearly brought out in the graph:
no two of the spicule-dimension curves are near to each other and parallel
throughout, lots similar in some respects being invariably different in others.

This want of correlation of the peculiarities and the results obtained by the
comparison of thirty-six pairs given above, force us either to consider all these
nine lots of sponges as representatives of the same systematic unit, or to estab-
lish a distinct group (species, subspecies, variety, or form) for each one.

Before entering upon the discussion of this question it will be well to ascer-
tain whether there is any correlation between the degree of the difference be-
tween the pairs and the distance of the localities where they were obtained.

As stated, these sponges come from five different regions of the Pacific coast.
To ascertain what correlation there is between the distances of their localities
and the degree of difference between them in regard to the six spicule-dimensions
here under discussion, I arranged the thirty-six pairs of specimens from the
different localities in five groups: 1, pairs of specimens from localities in the same
region; 2, pairs of specimens from localities in adjacent regions; 3, pairs of
specimens from localities in regions between which one other region lies; 4, pairs
of specimens from localities in regions between which two other regions lie; and
5, pairs of specimens from localities in regions between which three other regions
lie. Then the ranges and averages of the differences of these groups of pairs were
ascertained for each group.

To the first group (pairs from localities in the same region) the pairs AB,
AD, BD, EI, and FG belong; the differences of these pairs range from 8.1 to
13.3, their average being 10.6.

To the second group (pairs from localities in adjacent regions) the pairs
AE, AF, AG, AI, BE, BF, BG, BI, CE, CI, DE, DF, DG, DI, FH, and GH
belong; the differences of these pairs range from 6.5 to 16.8, their average being
Nef

To the third group (pairs from localities in regions with one other region
between) the pairs AC, AH, BC, BH, DC, DH, EF, EG, FI, and GI belong;
the differences of these pairs range from 6.1 to 22.5, their average being 14.9.

To the fourth group (pairs from localities in regions with two other regions
between) the pairs CF, CG, EH, and HI belong; the differences of these pairs
range from 18.4 to 21.1, their average being 19.4.

The only pair of the fifth group (pairs of localities in regions with three other
regions between) is CH; the difference of this pair is 26.8.

This shows that the variation of the specimens from different localities in

GEODIA AGASSIZII. 149

the same region is inconsiderable (never over 13.3) and that the averages of the
differences between the specimens from localities in different regions are pro-
portional to the distance between the regions.

As in the case of the morphological aspect of the differences between the
pairs we also find, when examining them from this distributional point of view,
that these differences are regularly graduated: forms intermediate in struc-
ture, growing in places intermediate geographically, connect the very different
forms C and H occupying the two extremities of the stretch of coast off which
these sponges grow.

This morphological and distributional continuity of the whole series of
forms renders it, in my opinion, advisable to consider the great differences of the
protruding spicules as due to mere individual adaptation, and to place the
whole series in one species not divided into minor systematic units.

Although I refrain from systematically separating the different forms grow-
ing in the different regions, I think that they might well be considered as
“ineipient”’ varieties or subspecies, which, adapting themselves more and more
to the different peculiarities of their surroundings, may, and very likely will, in
course of time, become systematically distinct — particularly if, through some
cause or other, they should disappear from the central stretch of coast they now
inhabit.

Of course there can be no doubt, that these sponges belong to the Geodidae,
but it is not so easy to decide the genus to which they should be assigned. As
some of their efferents are not covered by pore-sieves I was at first inclined to
place them in Sidonops. A careful investigation of the apparently uniporal
efferents made it highly probable, however, that these had, like the ones over
which sieve-membranes extend, been covered by such when the sponge was
alive and had been lost post mortem. For this reason I place these sponges in
Geodia. It is, in some cases, particularly when the specimens are not well
preserved, difficult to decide whether a geodid sponge with apparently uniporal
efferent apertures should be considered as a Sidonops or as a Geodia, and it is
probable that some of the species placed by me * in Sidonops ought to be con-
sidered as species of Geodia. In studying the affinities of Geodia agassizvi with
other previously described species, I have, for these reasons, compared it not
only with the sponges described as species of Geodia but also with those
described as species of Sidonops.

The species of Geodia which have orthoplagiotriaenes, subcortical teloclades,

1R, yon Lendenfeld. Tetraxonia, Tierreich, 1903, 19, p. 100-104.

150 GEODIA AGASSIZII.

sterrasters, and euasters more or less similar to those of Geodia agassizii are
G. spherastrella Topsent and the sponges here described as G. (Cydonium) ja-
ponica Sollas, G. mesotriaena, G. mesotriaenella, and G. breviana. Geodia japonica
is distinguished from G. agassizii by being cup shaped, having rounded, lobate
protuberances on the outer surface, and possessing, although very large in size,
considerably smaller megascleres; the large amphioxes of this species are less
than half as thick as those of G. agassizii. Geodia spherastrella is distinguished
from G. agassizii by being incrusting and having much larger dermal strongylo-
sphaerasters. Whether there are also differences in the size of the megascleres,
as is very likely, I cannot say, since Topsent ' gives no measurements. Geodia
mesotriaena is more or less meandrie and provided with praeoscular canals; and
it has choanosomal amphioxes and orthotriaene-cladomes twice as long, and also
larger euasters. (Geodia mesotriaenella has more slender and differently shaped
anatriaene-clades, no mesoprotriaenes with epirhabd exceeding the clades in
length, and somewhat different strongylosphaerasters. Geodia breviana pos-
sesses minute dermal anaclades, its large anatriaenes have much thicker clades
and its strongylosphaerasters are larger.

The only species of Sidonops which has orthoplagiotriaenes, subcortical
teloclades, sterrasters, and euasters similar to those of @. agassizvi 1s the sponge
deseribed by Dendy?’ as G. areolata Carter. This sponge differs from G. agassizti
by the absence of mesoprotriaenes, by having somewhat smaller spicules, particu-
larly thinner choanosomal amphioxes and orthoplagiotriaene-rhabdomes, and
by the reticulate structure of its surface. The differences in the size of its
spicules and the absence of mesoprotriaenes, which, if present only in the fur,
have perhaps been lost post mortem, may not be sufficient for distinguishing it
specifically from G. agassizii. The reticulate appearance of the surface of G.
areolata is caused by the presence of a superficial network, which consists,
according to Carter (Ann. mag. nat. hist., 1880, ser. 5, 6, p. 133, pl. 6, fig. 37),
of rows of small protruding dermal amphioxes, according to Dendy (loc. cit.),
of subdermal pigment-cells. In none of the specimens of G. agassizii could I
detect the slightest trace of a network of either of these kinds on the surface. I
am not prepared to express an opinion on the systematic value of the presence
or absence of this network. Since, however, it may be systematically important;
since there is a difference in the spicules, although by itself perhaps too slight for
specific distinction; and since the localities where these sponges occur, Pacific

‘BE. Topsent. Spongiaires des Acgores. Résult. Monaco, 1904, 25, p- 70.
* A. Dendy. Report on the sponges. Rept. pearl oyster fisheries, 1905, pt. 3, p. 87.

GEODIA MESOTRIAENELLA. 151

coast of North America and coast of Ceylon (Gulf of Manaar), are so far apart,
I think it advisable to keep them systematically distinct.

Geodia (Cydonium) miilleri Fleming, to which species Lambe (loc. cit.)
assigned the specimen of G. agassizii examined by him, differs from this species
in having much larger and more elongated sterrasters.

Geodia mesotriaenella, sp. nov.

Plate 34, figs. 18-26; Plate 35, figs. 28-35.

I establish this species for a specimen captured off the coast of southern
California at Station 4417. It is similar to the species described as Geodia meso-
triaena, but in regard to the size of the body and the dimensions of its spicules
is much smaller and to this the name refers.

Shape and size. The sponge (Plate 34, fig. 19) is nearly spherical; its small-
est and largest diameters are 15 and 19 mm. respectively. The surface is
quite smooth and covered by a spicule-fur still fairly intact in several places,
chiefly near the base. There are no larger oscules.

The colour, in spirit, is dirty white, rather darker above than below.

The superficial part of the body forms a cortex composed of three layers:
a dermal (outer) layer free from sterrasters, 30-60 » thick; a middle sterraster-
armour layer, 320-380 thick; and an inner, fibrous layer, free from sterrasters,
120-220 yp thick.

Canal-system. Apart from a few patches on the upper side, the largest of
which is roughly circular and 8 mm. in diameter, the whole of the surface is
occupied by groups of afferent pores (Plate 34, fig. 20), about 700 # in maximum
diameter and separated by poreless strips, usually 15-50 « broad. The pores
themselves are oval, 20-80 in diameter, and the strands of tissue separating
them narrower the larger the pores are. In the efferent areas smaller groups,
only 200-500 jin diameter, containing a smaller number of larger pores, 40-160 px
wide (Plate 34, fig. 25), are met with.

Skeleton. Radial bundles of megascleres traverse the choanosome and
penetrate the cortex; their distal parts protrude freely beyond the surface and
form the spicule-fur. The proximal parts of these bundles in the interior of the
choanosome are composed entirely of amphioxes. Distally orthotriaenes,
mesoprotriaenes, and anatriaenes are added to the amphioxes. Most of the
orthotriaene-cladomes lie in the inner layer of the cortex. Anatriaene-cladomes
are found in considerable numbers, both at this and at a lower level. The spicule-
fur is composed of the distal parts of amphioxes, mesoprotriaenes, and anatriaenes.

152 GEODIA MESOTRIAENELLA.

Intact mesoprotriaene-cladomes are very numerous in the proximal portion of
the fur, a short distance above the surface. The spicules forming the distal
(superficial) portion of the fur are mostly broken off. So far as I could see, most
of these terminally broken spicules are anatriaene-rhabdomes. In the dermal
layer small, more or less radially arranged rhabds, mostly styles attenuated
towards both ends, occur. Their proximally situated rounded ends are im-
planted in the sterraster-armour; their pointed distal ends protrude a short
distance beyond the surface. Small strongylosphaerasters form a thin but
rather dense and continuous layer on the outer surface (Plate 34, figs. 20, 25)
and also occur in the lower parts of the dermal layer. Large oxysphaerasters
are imbedded in the walls of the cortical and subcortical canals. OxyYasters are
scattered in large numbers throughout the choanosome. Everywhere, except
in the vicinity of the radial canals, which traverse it, the middle layer of the
cortex is occupied by dense masses of sterrasters. Many sterrasters, particu-
larly young forms, also occur scattered in the choanosome. It is to be noted
that the sterrasters are not nearly so numerous in the lower layer of the cortex
and the distal part of the choanosome as in the proximal part of the latter.

The large amphiozes (Plate 36, fig. 31c) are slightly curved, gradually atten-
uated towards the pointed ends, and isoactine or anisoactine. Anisoactine forms
were chiefly found among the stouter amphioxes. The amphioxes are 2-2.6
mm. long and 20-50 y thick.

The minute dermal rhabds are 196-260 » long and 4-5 p thick. The greater
number of them are styles with attenuated rounded ends. In some of these
rhabds this attenuation is so great that the proximal ‘‘rounded” end is hardly
less pointed than the distal. These spicules appear as anisoactine amphioxes.

The orthotriaenes (Plate 35, fig. 31a) have a fairly straight, conical rhabd-
ome, usually sharp pointed, very rarely rounded at the acladomal end. The
rhabdome is 2.1-2.4 mm. long and, at its thickest point, 75-120 y thick. This
thickest point is situated either at the cladome or a short distance below it, and,
in the latter case, separated from the cladome by a slight neck-shaped con-
striction. The clades are always conical and curved, concave to the rhabdome
in their entire length; the degree of curvature and the size of the clades are,
however, variable, the clades of the same cladome often differing from each other
considerably in these respects. The clades are sharp pointed or, more rarely
rounded at the ends, and generally simple, but cladomes with one or two bifur-
cate clades have also been observed occasionally. The clades are 350-600 y
long and their chords enclose angles of 90-96°, on an average 93°, with the axis
of the rhabdome.

GEODIA MESOTRIAENELLA. 153

Some intact rhabdomes of the mesoprotriaenes (Plate 36, figs. 28-30, 31b)
were 2.8-3.4 mm. long, but several of the fragments observed appeared to be parts
of rhabdomes longer than that. The thickness of the rhabdome at the cladome
is9-19 . The rhabdomes which are thick at the cladome are in their cladomal
half nearly cylindrical, hardly perceptibly thickened towards the middle; the
rhabdomes thin at the cladome, on the other hand, are very considerably thick-
ened towards the middle of their length. The epirhabd is straight, conical,
pointed, and 70-165, usually 90-120 » long. The clades are slender, conical,
and curved, concave to the epirhabd. Their chords are 100-220 » long and
enclose angles of 30-47°, on an average 36°, with the axis of the epirhabd. The
clades of the same cladome are equal or, more frequently, unequal in length
(Plate 35, figs. 28-30); all of them or, if they differ considerably in length, two,
or at least one, are longer than the epirhabd. The proportion of the length
of the epirhabd to the length of the longest clade of the same cladome being
100: 150 to 100 : 233.

The anatriaenes (Plate 35, figs. 32-35). The longest intact rhabdome
measured was 3.7 mm. long. Fragments indicate that many anatriaenes have
rhabdomes considerably longer. The thickness of the rhabdome at the cladome
is 18-30, usually 20-22 ». The clades are conical, pointed, uniformly and not
strongly curved, concave to the rhabdome. Their chords are 87-140, usually
110-130 long and enclose angles of 41—57°, on an average 48°, with the axis of
the rhabdome. Besides these normal anatriaenes with simple clades I have
observed a few, similar to them in all other respects, with one of the clades
bifureate.

The large oxyasters (Plate 34, figs. 18a, 21, 22, 24) have a small centrum the
diameter of which is usually about double the basal thickness of the rays. From
this centrum from five to eleven fairly regularly distributed rays arise radially.
The rays are straight, conical, usually quite blunt, 9-14 long, and, at the base,
1.5-2.8 » thick. The basal parts of the rays are smooth, the distal parts covered
with spines (Plate 34, figs. 21, 22). The size of these spines is variable and their
number is, roughly speaking, in inverse proportion to their size. The largest
spines measured were 0.7 » long. The spines arise vertically from the ray.
Those large enough to be properly seen are usually sharp pointed, but occasion-
ally I have also observed stout and cylindrical, terminally rounded spines.
The whole oxyaster is generally 17-26 » in diameter. In the centrifugal
spicule-preparations also large ones, up to 40 # in diameter, have been ob-
served. As, however, I never found such large asters in situ in the sections

154 GEODIA MESOTRIAENELLA.

and as they are exceedingly scarce in the spicule-preparations, I am by no
means sure that they are proper to the sponge. A correlation between the
ray-number and the size of the spicule is not clearly discernible.

The large oxysphaerasters have a spherical centrum, 6-9 # in diameter,
from which from fifteen to twenty-three regularly distributed rays arise radially.
The rays are straight, conical, sharp pointed, 6-7 « long and 2-2.5 yp thick at
the base. Each ray bears a small number, sometimes only one or two, vertically
arising spines. The whole aster is 20-21 in diameter.

The diameter of the centrum of the small strongylosphaerasters (Plate 34,
fig. 26) is 1.6-4.5 , usually from a quarter to a half of the diameter of the whole
aster. From this centrum generally from ten to seventeen regularly distributed
rays arise radially. Very rarely strongylosphaerasters with fewer (one or three),

»or with more numerous (up to twenty-five) rays have been observed. The rays
are cylindrical or cylindroconical and truncate. In the strongylosphaerasters
with small centrum (Plate 34, fig. 26a) the rays are much longer than thick and
attenuated towards the end. In the strongylosphaerasters with large centrum
(Plate 34, fig. 26b) the rays are only twice as long as thick or even shorter, and
cylindrical. The rays are covered with spines, which are larger in the one- and
three-rayed than in the many-rayed forms. The ray of the one-rayed strongylo-
sphaeraster is 7.2 « long and 2 » thick; the rays of the three-rayed forms are
4-5 wlong and 1 y thick; those of the many-rayed forms are 2-4.5 » long and
0.5-1 » thick. The total diameter of the aster is 6-11 4. The ray-number is
in so far in inverse proportion to the size of the spicule as the one- and three-
rayed forms are 8.4-11 y, the forms with more rays 6-10 in total diameter.

The sterrasters (Plate 34, fig. 23) are flattened ellipsoids, usually 87-97 p,
rarely as much as 107 long, 77-86 , rarely as much as 92 » broad, and 58-69 pt
thick. The proportion of length to breadth to thickness is, on an average,
100: 89: 71.

This sponge was caught with the tangles at Station 4417 on April 12, 1904,
near Santa Barbara Islands, 8. W. rock Santa Barbara Island, N. 8° W.,
11.7 km. (6.3 miles), drift S. 73° W.; depth 53 m. (29 f.); it grew on a bottom
of fine yellow sand and coralline rock.

The cribriporal character of the afferent and efferent apertures and the
ellipsoidal sterrasters show that this sponge belongs to Geodia. It is most
closely allied to the sponges described by Dendy (1905) as Geodia ramodigitata
Carter and G. areolata Carter, and to those described here as Geodia ataxastra,
G. breviana, G. agassizii, and G. mesotriaena. In Dendy’s Geodia ramodigitata

GEODIA BREVIANA. 155

Carter the oxyasters are much larger and destitute of spines. In Dendy’s Geodia
areolata (Carter) the orthotriaene-rhabdomes are much more slender, the ana-
triaene-clades shorter, the mesoprotriaenes absent, and the surface marked with
a reticulate tracing. In Geodia ataxastra the mesoproclades and anatriaenes
have differently shaped cladomes, the dermal strongylosphaerasters are much
smaller, and ataxasters are met with; in G. breviana minute, dermal anaclades
are present, the clades of the large anatriaenes are much stouter and the strongy-
losphaerasters larger; in G. agassizii some of the spicule-dimensions are con-
siderably greater, the anatriaene-clades stouter and more angularly bent, the
orthotriaene-clades often provided with one or more spine-like branches, some of
the mesotriaene-epirhabds much longer than the clades of the same cladome,
and smaller oxyasters abundant; and in G. mesotriaena most of the spicule-
dimensions are from two to three times as great, the anatriaene-clades stouter
and curved in a different manner, the mesotriaene-epirhabds usually longer than
the clades of the same cladome, the rays of the strongylosphaerasters more
slender, and the oxysphaerasters much more spiny and provided with relatively
smaller central thickenings.

Geodia agassizii and G. mesotriaena appear to be more closely allied to G.
mesotriaenella than the other four. It is true that many of the spicule-dimensions
of these sponges, particularly of G. mesotriaena, are much greater; in view of the
fact, however, that the specimen of G. mesotriaenella is very much smaller, and
therefore probably much younger, than the specimens of G. agassizii and G.
mesotriaena examined, this might be of no systematic importance. The differ-
ences between these species and G. mesotriaenella in regard to the shape of the
spicules may, of course, also be due to differences in the degree of development
(age), in which case G. mesotriaenella would have to be considered as a young
form of one or the other of them. Since, however, this seems very doubtful,
I deem it better, in order to avoid confusion, to describe this sponge as a dis-
tinct species.

Geodia breviana, sp. nov.
Plate 35, figs. 1-27; Plate 36, figs. 1-12.
1893. Cydonium miilleri L. M. LAMBE (non Fleming), Trans. Roy. soc. Canada, 1893, 10, p. 72, pl. 4,
fig. 1; pl. 6, fig. 1, la-i.

I establish this species for a specimen dredged at Station 2894, off southern
California. The large anatriaenes have unusually stout and relatively short
clades, and to this the specific name refers. The study of the sponge deter-
mined in 1893 by Lambe (loc. cit.) as Cydonium miilleri Fleming which is in the

156 GEODIA BREVIANA.

collection of the Geological Survey of Canada and which was kindly placed at
my disposal for examination, has shown that it differs from the type of Geodia
(Cydonium) miilleri and is identical with the specimen referred to above.

Shape and size. The specimen is fragmentary, 32 mm. long, 21 mm. broad,
and 10 mm. thick. Its surface is now fairly smooth, but the living sponge was
probably covered with a spicule-fur. There are three small groups of circular
or oval pores 300-800 « wide, most of which open freely on the surface. The
specimen described by Lambe is a thick-walled cup, 6 cm. high, with a rather
small cavity, occupied by a high spicule-fur, which consists of long anaclades
and mesoproclades.

The colour, in spirit, is dirty white on the surface, darker in the interior.

The superficial part of the body forms a cortex, composed of a very thin

uter, dermal layer, a middle sterraster-armour layer 700-900 y thick, and an
inner fibrous layer, which latter is, like the dermal layer, destitute of sterrasters.
The thinness of the dermal layer may be due to a post mortem collapse in con-
sequence of indifferent preservation.

Canal-system. A closer investigation of the wide apertures 300-800 y in
diameter, mentioned above, revealed remnants of pore-sieves in many of them.
From this I conclude that, in life, all of them were covered by pore-sieves and
that, where the sieves are now missing, they have been lost post mortem. Radial
canals of equal width and only very slightly constricted below, at the lower
limit of the sterraster-armour layer, lead down from these apertures to the
choanosome. Outside the regions occupied by the groups of these wide aper-
tures (cortical canals) no open canals and only few pores, about 50 s wide, were
observed. I take the latter for the afferents and the large apertures in the groups
for the places where, in life, the sieves with the efferent pores were spread out.

Skeleton. Radial bundles of megascleres traverse the choanosome, and
penetrate the cortex. Their distal parts appear to have protruded freely and
to have formed a spicule-fur. Their proximal parts are composed of amphioxes;
distally orthotriaenes, anatriaenes, and probably also mesoprotriaenes are added
to the amphioxes in the bundles. The cladomes of the orthoplagiotriaenes lie
partly at, partly a little above, and partly a little below, the lower limit of the
sterraster-armour layer, the most distal ones being partly enveloped in sterr-
asters. Anatriaene-cladomes are quite numerous below the sterraster-armour
layer. In the spicule-preparations a few mesoprotriaenes were found, but I saw
none im situ in the sections. Minute amphioxes and anaclades are implanted
in the distal part of the sterraster-armour. These spicules are situated radially

GEODIA BREVIANA. 157

or obliquely. They traverse the thin dermal layer and protrude freely beyond
its outer surface. The spicule-fur appears to have consisted of a high portion,
composed of the distal parts of the large amphioxes, anatriaenes, and perhaps
also mesoprotriaenes protruding several millimeters; and a low portion, forming
a sort of undergrowth, composed of the distal parts of the minute dermal am-
phioxes and anaclades, protruding only 100-200 yp.

The microscleres are thick-rayed and perhaps also thin-rayed oxyasters,
large oxysphaerasters, small strongylosphaerasters, and sterrasters. The thick-
rayed oxyasters are numerous throughout the choanosome. The thin-rayed
oxyasters are rare and were found only in the centrifugal spicule-preparations;
they may be foreign. The large oxysphaerasters occur chiefly in the walls of
the subcortical canal. The strongylosphaerasters form a single layer on the
outer surface and also occur scattered in the lower parts of the cortex, The
sterrasters occupy the sterraster-armour layer in dense masses.

The large amphioxes (Plate 35, figs. 1-4) are slightly curved and uniformly
and gradually attenuated towards their pointed ends. They measure 1.8—3.7
mm. in length and 30-88 in thickness. Their thickness is by no means always
proportional to their length. In the specimen described by Lambe the ordinary
choanosomal amphioxes are 3-5 mm. by 50-70 (Lambe, 1893, 2.77-3.81 mm.
by 80 y).

In this specimen I have also found some large styles; these are shorter than
the amphioxes and up to 90 » thick; they are not mentioned by Lambe.

The minute dermal amphioxes (Plate 36, figs. 10-12) are simply curved
(Plate 36, fig. 10) or angularly bent (Plate 36, figs. 11, 12) and gradually attenu-
ated towards the abruptly pointed ends. Examination with high powers
shows that both ends of these spicules are usually somewhat drawn out to ex-
ceedingly sharp terminal spines. The angular bend is, when present, generally
not in the middle of the spicule, but considerably nearer to one end than to the
other. It amounts to about 25°, so that the angle which the two parts of the
spicule enclose is usually about 165°. The minute dermal amphioxes are 280-
450 » long and 4.5-8.5 » thick. The shorter ones, 280-365 1 in length, are much
more numerous than the longer ones, 366-450 y in length. In the specimen
described by Lambe these amphioxes are 340-440 « by 2-5 (Lambe, 1893, 288
by 13 4, according to his figure 8.5 /4).

The ortho- and plagio-triaenes (Plate 36, figs. 15-17) have a fairly straight
thabdome, 1.8-3.2 mm. long and 90-130 y thick at the cladome. Usually the
thabdome is conical and sharp pointed, rarely rounded at the acladomal end.

158 GEODIA BREVIANA.

One of these rhabdomes possessed several irregular thickenings near the rounded
end, under which the axial thread passed smoothly without any thickening or
other modification. The clades are conical, pointed, and uniformly curved in
their entire length, concave to the rhabdome. The degree of the curvature
usually is, as in the two triaenes represented on Plate 36, figs. 15, 16, quite con-
siderable, rarely so slight as in the long (left) clade of the triaene (Fig. 17).
Occasionally triaenes with clades bent much more strongly than those represented
in Figs. 15 and 16 have been observed. The clades are nearly always simple;
very rarely one of them is bifureate. The clades of the same cladome are usually
similar, irregular cladomes being exceptional. The chords of the clades are 280—
500 «long and enclose angles of 94-108°, on an average 101.6°, with the axis of
the rhabdome. Of the seventeen triaenes the clade-angles of which were meas-
ured, four were orthotriaenes with clade-angles less than 100°, the other thirteen
plagiotriaenes, with clade-angles 100° or more. In the specimen described by
Lambe the plagio-orthotriaenes have rhabdomes 2.2-4.1 mm. (Lambe, 1893,
2.4 mm.) by 60-105 yp, clades 340-680 «4 (Lambe, 1893, 700 ») long, and clade-
angles of 97-113°.

The mesoprotriaenes (Plate 35, fig. 14) have a rhabdome about 15 yp thick at
the cladome, an epirhabd 105 u long and clades, curved concave to the epirhabd,
the chords of which are 115-130 y long and enclose angles of about 38° with
the epirhabd-axis. In the specimen described by Lambe the mesoprotriaenes
have rhabdomes 7-11 mm. (Lambe, 1893, 7.84 mm.) by 15-32 p, clades 65-250
(Lambe, 1893, 95 ») long, clade-angles of 20-44°, usually 32-44°, and an epirhabd
120-310 y long.

The large anatriaenes (Plate 35, figs. 5-7) have a nearly cylindrical rhabd-
ome, 25-40 » thick in its cladomal part. I found no intact rhabdomes of these
spicules; the fragments observed indicate that they attain a considerable length.
The clades are slightly and uniformly curved or somewhat angularly bent near
the end, concave to the rhabdome. They are remarkably thick at the base,
and usually simple and quite sharply pointed, but anatriaenes with one clade
bifureate have also been observed occasionally. The two terminal branches of
such clades are nearly parallel and lie close together. The chords of the clades
are 60-115 » long and enclose angles of 50-65°, on an average 57.3°, with the
axis of the rhabdome. In the specimen described by Lambe the large ana-
triaenes have a rhabdome 9-11 mm. (Lambe, 1893, 7.5 mm.) by 25-37 p, stout
clades 47-82 4 (Lambe, 1893, 60 2) long, and clade-angles of 52-63°.

In the specimen described by Lambe a few anadiaenes similar to the ana-

GEODIA BREVIANA. 159

triaenes but with longer clades (up to 110 « long) and smaller clade-angles (about
45°) were also observed.

The minute dermal anaclades (Plate 36, figs. 1-9) are triaene, diaene, or
monaene. The triaene forms are more numerous than the other two. Their
rhabdomes are usually quite strongly and somewhat irregularly curved (Plate
36, figs. 2, 4, 6, 8), thickest about two thirds of their length from the cladome,
and rounded at the acladomal end. They are 480-560 long and at the cladome
2.5-4.5 p thick. At their thickest point they are about twice as thick as at
the cladome and here measure 5-8.6 y in transverse diameter. The rounded
acladomal end is 3.4-7 y thick, slightly thicker than the cladomal end. The
clades are conical and usually pointed and uniformly curved, concave to the
rhabdome. In the triaene forms (Plate 36, figs. 1-5) the chords of the clades
are 7-9 long and enclose angles of 48-58° with the axis of the rhabdome; in the
diaene and monaene forms (Plate 36, figs. 6-9) the chords of the clades are 11-—
12 » long and enclose angles of 42—46° with the axis of the rhabdome. In young
forms like the left one of the two represented in Plate 36, fig. 1, the clade-angles
are larger. In the specimen described by Lambe the minute dermal anaclades
(not mentioned by Lambe, 1893) are exceedingly abundant. Their rhabdome
is 350-610 by 1-3.5 » at the cladome, and 5-7 y» at the thickest point below
the middle; the clades are 2-12 y» long, the clade-angles 42-60°. Some of the
minute anaclades of this specimen have a straight, conic, apical ray, an
epirhabd, 5-8 » long, and therefore appear as mesanaclades.

The thick-rayed oxyasters (Plate 35, figs. 18a, 19a, 22a, 24, 27) have five to
twelve rays and a small centrum, the diameter of which is from two to three
times as great as the basal thickness of the rays. The rays are usually simple,
radial, and regularly distributed. Occasionally irregularities are observed due
either to an irregular position or to a bifurcation of one or more of the rays.
The two branches of the bifurecate rays are nearly parallel and lie close together.
The rays are straight and conical, pointed, or blunt. Their proximal part is
smooth, their distal part covered with rather large spines. The spines are not
very numerous and rise vertically from the rays. Their ends appear to be
curved backward towards the centre of the aster in a claw-shaped manner. The
rays are (without the centrum) 7.5- 11 # long and 1-2.2 » thick at the base, the
total diameter of the oxyaster being 16-26.5 w. A correlation (inverse propor-
tion) between the size and the ray-number is observed in so far as the oxyasters
with the most numerous (twelve) rays do not exceed 18.5 » in total diameter.
In the specimen described by Lambe these asters have from five to nine rays 1.4—

160 GEODIA BREVIANA.

2.3 thick, and are 19-25 » (Lambe, 1893, 3-13, according to his figures
11-15 ) in total diameter.

The rare thin-rayed oxyasters which, as above stated, may be foreign, have
from nine to fourteen rays and no centrum. The rays are 3-11 y long, and very
thin, only 0.25-0.7 « thick at the base. They are not very much attenuated
towards the end and bear spines which sometimes form terminal verticillate
clusters, in which case these asters appear as acanthtylasters. The total diame-
ter of these asters is 7-23 If they are not foreign, they may be young forms
of the thick-rayed oxyasters described above.

The large oxysphaerasters (Plate 35, figs. 21¢e, 25, 26) have a spherical cen-
trum, 7-9 in diameter, from which from twenty-three to twenty-seven straight,
conical, and sharp-pointed, usually regularly distributed rays arise radially.
These rays are (without the centrum) 6-8 » long and 1-2.7 » thick at the base.
They are covered with a greater or smaller number of good-sized spines. The
rays of the oxysphaerasters with twenty-six or twenty-seven rays are basally
only 1-2.2 », those of the oxysphaerasters with twenty-three or twenty-four rays
basally 2.5-2.7 » thick. This indicates that there is an inverse proportion
between the ray-number and the ray-thickness. The whole aster is 14-21.5 p
in diameter. In the specimen described by Lambe these asters have up to’
thirty rays 1—-1.7 » thick, the centre is 3-5.5 , the whole aster 12-18 y, in diame-
ter. They are not mentioned by Lambe (1893).

The small strongylosphaerasters (Plate 35, figs. 8-13, 18-22b) have a centrum
3-5.5 # in diameter and usually from thirteen to twenty-one, very rarely only
seven or three rays. The rays are generally radial and regularly distributed,
rarely arranged irregularly. Such an irregularity is chiefly observed in the
rare few-rayed forms which are evidently derivates of the ordinary many-rayed
ones, produced by the suppression of a smaller or greater number of rays. The
rays are cylindrical or cylindroconical and truncate or terminally rounded. Their
basal parts are smooth, their distal parts covered with a‘number of good-sized
spines. The rays are 1.8-5 » long and 0.6-1.7 thick; the whole aster measures
7-12 » in diameter. A correlation between ray-number and spicule-size is not
pronounced. In the specimen described by Lambe these asters have from four-
teen to twenty-five rays, 0.5-1.2 u thick, the centre is 2-4 yw, the whole aster
6-9 w, in diameter. They are not mentioned by Lambe (1893).

The sterrasters (Plate 35, fig. 23) are flattened ellipsoids, 87-105 y long,
80-98 » broad, and 70-77 » thick. The proportion of length to breadth to thick-
ness is on an average 100 : 91:74. In the specimen described by Lambe the

GEODIA OVIS. 161

sterrasters measure 84-97 by 75-85 by 55-70 » (Lambe, 1893, 92 »). In the
centre of a young sterraster measuring 37 » I observed one large granule, about
1 » in diameter, and several quite small ones:

This sponge was dredged off southern California at Station 2894 on Janu-
ary 5, 1889, in 34° 7’ N. 120° 33’ 30” W.; depth 97 m. (53 f.); it grew on a bottom
of sand and broken shells; the bottom temperature was 13.7° (55.6° F.). The
specimen described by Lambe was obtained in the Strait of Georgia near
Comox, Vancouver Island, depth 7 m. (4 f.).

Since this sponge is provided with sterrasters and regular triaenes and
many of its cortical canals open out freely on the surface while others are
provided with pore-sieves, it might be supposed to belong to Sidonops. Since,
however, as stated above, a closer examination reveals remnants of pore-
sieves at the mouths of the apparently freely opening canals, its sidonoptic
appearance is probably deceptive. Believing that, in life, it had not only
eribriporal afferents, but also cribriporal efferents, I place it in Geodia. The
species of Geodia and Sidonops most nearly allied to Geodia breviana are the
sponges described by Dendy as Geodia ramodigitata Carter and G. areolata
Carter, and those here described as G. mesotriaenella, G. mesotriaena, and G.
agassizii. From all these it is distinguished by the stoutness of the clades of
its large anatriaenes and the possession of minute dermal anaclades. From
Geodia (Cydonium) miilleri, to which species Lambe assigned the specimen
described by him in 1893, G. breviana differs by its sterrasters being larger
and the clades of its large anatriaenes being much shorter, stouter, and less
inclined to the rhabdome.

Geodia ovis, sp. nov.

Plate 40, figs. 1-30; Plate 41, figs. 1-20; Plate 42, figs. 1-40; Plate 43, figs. 1-8.

I establish this species for a spirit specimen from the coast of southern Cali-
fornia (Station 2975). It has an exceedingly dense and high spicule-fur which
is somewhat woolly in character, and to this the specific name refers.

The specimen is a part of a larger sponge which appears to have been
horizontally extended, cake shaped, and about 4 cm. thick. The specimen itself
(Plate 40, fig. 28) is 127 mm. long, 50 mm. broad, and 27 mm. high. Its natural
surface is somewhat undulating and covered with a woolly spicule-fur up to
20 mm. high (Plate 40, figs. 5, 28).

The colour, in spirit, is light brown.

The superficial part of the body is differentiated to form a cortex (Plate 40,

162 GEODIA OVIS.

figs. 5b, 28; Plate 42, figs. la, b, 2a, 8a, b, 9a, b) composed of three layers: an
outer dermal layer, 100-200 thick; a middle sterraster-armour layer, 150-300 yz
thick; and an inconspicuous, inner fibrous layer, not sharply defined from the
choanosome.

Canal-system. The parts of the surface which have lost their spicule-fur,
and which are consequently exposed to view, are occupied by pore-sieves (Plate
40, fig. 25). The pores in these sieves are oval or, more rarely, circular, and
measure 30-45 in maximum diameter. The strands of dermal tissue separat-
ing the pores are narrow, usually only 10-20 « broad. The pores lead into cavi-
ties excavated in the dermal layer. From the latter radial canals, penetrating
the middle and inner layers of the cortex, arise. These radial canals are sur-
rounded by chonal sphincters which protrude inwards 300-400 » beyond the

» sterraster-armour layer. The cylindrical chonal structures thus formed (Plate
42, figs. Id, 8) usually are 150-200 y in transverse diameter. The chonal
canal in the axis of these chones is cylindrical and usually open and up to 50 p
wide. Some of these canals open out below with funnel-shaped extensions.

The choaonosome is traversed by canals the widest of which are 2 mm. in
diameter. Many of them are provided with membranous sphincters at very
frequent intervals. The pore-sieves described above I take to be afferent. I
did not observe any that looked like efferents. To find these it would have been
necessary to remove the spicule-fur, and this I did not want to do as it would have
injured the unique and valuable specimen. From the general appearance of the
sponge I am inclined to conclude that the efferents are, like the afferents, eribri-
poral.

The skeleton of the internal part of the choanosome consists of irregularly
disposed large amphioxes, a few styles (tylostyles), not very numerous asters,
mostly large thin-rayed oxyasters, some minute rhabds, and a few sterrasters.
The distal part of the choanosome is traversed by radial bundles of large mega-
scleres which abut vertically on the cortex. These bundles consist of numerous
large amphioxes, a few styles (tylostyles) with the rounded end situated distally,
and the rhabdomes of orthotriaenes, anatriaenes, mesoproclades, a few proclades,
and very few anamonaenes. The cladomes of most of the orthotriaenes and of
some of the anaclades are situated in the level of the lower limit of the
sterraster-armour layer. A few cladomes of these spicules and of the promeso-
clades (proclades), chiefly young forms, also occur at lower levels. Between
these spicule-bundles minute rhabds, large masses of asters, chiefly large thin-
rayed oxyasters, and a few sterrasters are met with. The asters are much more
numerous in this region than in the interior of the choanosome.

GEODIA OVIS. 163

In the inner layer of the cortex, below the sterraster-armour, large thick-
rayed oxyasters and also smaller euasters occur. The sterraster-armour layer
is occupied by sterrasters lying rather loosely and on an average four deep. It is
traversed by large megascleres the proximal parts of which take part in the
formation of the radial spicule-bundles above referred to, and the distal parts
of which protrude freely beyond the surface. The proximal parts of the minute
rhabds forming the dermal tufts are implanted in the sterraster-armour layer.
- Many of these spicules traverse nearly the entire thickness of this layer and
extend down to within a short distance of its lower limit; others quite reach
that level, and some even protrude beyond it. The dermal layer is traversed
by tuft-like groups of more or less radially disposed minute rhabds which
diverge above (Plate 42, fig. 9). The proximal parts of these spicule-groups
are, as stated above, firmly implanted in the sterraster-armour layer; their
distal parts protrude freely beyond the surface for a distance of 100-300 yp,
occasionally as much as 400 4, and form a sort of undergrowth at the base of the
spicule-fur. In the lower parts of the dermal layer large thick-rayed oxyasters,
similar to the subcortical ones, are met with. Its superficial part, that is, the
dermal membrane, is occupied by dense masses of small asters, for the greater
part strongylosphaerasters.

The spicule-fur (Plate 40, fig. 5a; Plate 42, fig. 2), which, as stated above,
in places attains a height of 20 mm., is composed chiefly of large amphioxes,
mesoproclades, and anaclades. Tylostyles, proclade mesoproclade-derivates,
and orthotriaenes also occur in it. Most of these spicules are implanted in the
sponge with their proximal end, some appear to lie in it quite freely. The large
amphioxes of the spicule-fur (Plate 42, fig. 2b) are fairly uniform in size and
for the most part arranged radially. The mesoproclades and their proclade-
derivates are more variable, particularly in regard to the shape of the cladome,
but all of the same order of magnitude and mostly arranged in a fairly regular
radial manner. The anaclades, which are nearly all anatriaenes, on the other
hand, exhibit an extraordinary diversity in size. Besides large, radially disposed
ones, numerous smaller and smallest anaclades are observed in the proximal part
of the spicule-fur, particularly on the lower surface of the sponge. The rhabdomes
of these small spicules are much and irregularly curved and although vertical
to the surface of the sponge in their basal part, do not extend radially through-
out their whole length. In places, particularly on the lower side, orthotriaenes
(Plate 42, fig. 2d) similar to the subcortical ones take part in the formation of

the proximal part of the spicule-fur.

164 GEODIA OVIS.

The large amphioxes (Plate 40, figs. 6-13, 27; Plate 42, fig. 2b) are slightly
curved, not very sharply pointed, 4-9, mostly 7-8 mm. long, and in the middle
30-100 yz, the long ones 70-100 y thick. Their terminal parts (Plate 40, fig. 27)
are more rapidly attenuated towards the ends than is the central part. The
two ends are usually slightly unequal. Two of these amphioxes measured: —

A B
100 » from one end 20 23 “in thickness
200 nh * an 31 31 yee a
300 nw a 39 35 Ties a
in the middle 76 72 a ss
300 » from the other end 32 37 joe :
200°2 = oa Pe a 29 30 Tie i
100 p ae a By * iI) rail roe Me

The rare large styles and tylostyles are of two different kinds. The one kind is
short and stout, the other very long and slender. Those of the former are
probably amphiox-derivates, those of the latter perhaps anaclade-derivates.
The short and stout ones are true styles, or, if tyle, only slightly thickened at
the rounded end, the tyle exceeding the adjacent parts of the spicule not more
than 10 % in transverse diameter. These styles (tylostyles) are 2.6-4 mm.
long; the rounded end (tyle) is 85-116 y thick. The long and slender spicules
of the latter kind are all tylostyles. Their length is considerable, but could not
be exactly ascertained because all I observed were broken off. These spicules
are about 40 y thick; their elongated, oval tyle measures 60-65 y in trans-
verse diameter.

The minute rhabds (Plate 42, figs. 3a, 4-7, 24a), which form the dermal
tufts and also occur scattered in the choanosome, are mostly amphioxes. Many
of them are distinctly anisoactine, and in some one end is rounded off. The
latter appear as styles. These minute rhabds are much more variable in length
than in thickness. They are 270-440 y, occasionally as much as 550 p, long or
even longer, and 8-13 y thick.

In the minute dermal rhabds of several geodine sponges, and particularly
frequently in those of the species under consideration, I have noticed a remark-
able anisoactinity of the axial thread (axial canal), this being very much thicker
(wider) at one end than at the other. One half of the axial thread (canal) is
in these spicules normally developed and appears as a thin, cylindrical thread.
The other half increases in thickness (width) from the centre of the spicule to
the end in a funnel-shaped manner. The end of the spicule in which this dis-
tally widened half of the axial thread (canal) terminates, appears as a thin-
walled tube. In the style-rhabds it is always the pointed end which contains

GEODIA OVIS. 165

the distally thickened (widened) half of the axial thread (canal). I am inclined
to consider the minute rhabds with such anisoactine axial threads (canals) as
young, not fully developed ones. If this assumption is correct it would follow
that in these minute rhabds the centre of growth (silicification) is not situated
in the centre but at one end. This seems to indicate that these spicules,
although usually diactine (amphiox) in shape, are in reality monactine in
character.

The orthotriaenes (and plagiotriaenes) (Plate 40, figs. 1-4, 19-23) have a
rhabdome 5-8 mm. long. At the cladome it is 74-100 y thick, farther down,
it generally thickens to 77-110 p. At this its thickest point, which lies a short
distance below the cladome, the rhabdome is usually 3-15 % thicker than at
the cladomal end. At the acladomal end it is in most cases attenuated to a
fine, terminally pointed thread. The clades are curved, concave to the rhabd-
ome, either uniformly throughout (Plate 40, fig. 21) or, more frequently, less
strongly towards the end of the clade than at the base (Plate 40, figs. 19, 20).
Sometimes this distal decrease of curvature is so great that the end of the clade
appears as a nearly straight cone or rod (Plate 40, fig. 22). Occasionally the
end of one of the clades is bent down abruptly. This, however, is observed only
in clades much reduced in length. Usually the clades are conical and blunt
pointed, less frequently cylindrical and rounded at the end. Such cylindrical
clades may be short or long. The clades of the same cladome are fairly equal
or, more rarely, very unequal. The longest clade of the cladome is 310-640 yu
long, the shortest often much shorter than 310 . The chords of the clades
enclose angles of 86-101°, on an average 94°, with the axis of the rhabdome.
Nearly all the adult spicules of this kind observed were true orthotriaenes,
plagiotriaene forms with clade-angles exceeding 100° being very rare among
them. Small, young forms have clade-angles of 109° and more, and appear as
plagiotriaenes.

I have observed two quite abnormal megascleres. Both are partly broken.
One (Plate 40, fig. 26) consists of a shaft, in the middle 80 y thick, and broken
off at one end. From the other end, which is 55 thick, one stout clade arises
at aright angle. This is only 70 long and divides at the end into a bunch of
small truncate axial threads radiating in all directions. All but one of these
branches are broken off, the one intact measures 70X15 y. The other abnormal
spicule is a hexactine with rays 40-55 thick, two of which are broken off.
Of the other four three are rounded and one pointed at the end. The longest
of these rays is 900 , the shortest only 75 long. This spicule is particularly

166 GEODIA OVIS.

interesting as hexactine abnormal megascleres are exceedingly rare in tetraxonid
sponges.

The mesoproclades and their proclade derivates (Plate 43, figs. 1-8) have a
rhabdome 6-17 mm. long and, at the cladome, 20-41 » thick. At its thickest
point, which lies a little above (nearer the cladome than) the middle of the length
of the rhabdome, the rhabdome is from two to three times as thick as at the
cladome. The cladomes of these spicules are very variable. Triaene forms
greatly predominate. In the most regular mesoprotriaenes (Plate 43, figs. 3, 4)
the clades are fairly equal, conic, blunt pointed, slightly curved, concave to the
epirhabd, and 140-170 » long. Their chords enclose angles of a little over 45°
with the axis of the epirhabd. The epirhabd is straight, conic, about 110
long, and pointed at the end. Mesoprotriaenes with cladomes rendered irregular
by the clades being of unequal length (Plate 43, figs. 6, 8) are very abundant.
In these spicules one clade may be very much longer than the other two, which
latter again may be fairly equal (Plate 43, fig. 6) or very unequal (Plate 43, fig. 8).
In these irregular mesoprotriaenes the clades are 40-260 » and the epirhabd
is 70-100 » long.

In some of the mesoproclades one or two clades are completely suppressed ;
these appear as mesoprodiaenes (Plate 43, fig. 1) and mesopromonaenes. The
dimensions of the clades and epirhabds of these spicules are similar to those
of the triaene forms. The clades of the mesoprodiaenes are not opposite each
other but occupy the same positions —in planes passing through the rhabd-
ome and enclosing an angle of 120°— as they would if the third clade were
present.

Some of the teloclades observed I am inclined to consider as mesoprotriaene-
derivates in which the epirhabd has been suppressed. These spicules have a
rhabdome as long and thick as or slightly thicker than the mesoproclades, and
three more or less ascending clades, convex to the rhabdome, the chords of
which are 100-360 long and enclose angles of 31-76° with the continuation of
the rhabdome-axis. The large-angled forms of these spicules (Plate 43, fig. 2)
appear as plagioclades, the small-angled ones (Plate 43, fig. 7) as proclades.

The anaclades (Plate 42, figs. 2c, 3b, 10, 11, 23b, 24b, 25-40) are nearly all
anatriaenes. I have observed only one or two anamonaenes among them.
The anaclades are remarkable for the great differences in their size. The small
ones observed cannot be considered as the young of the large ones, because they
are found in abundance, more or less extruded from the sponge, in the spicule-
fur, where they can hardly be expected to continue to grow.

GEODIA OVIS. Ove.

I measured forty-nine cladomes of anaclades of this sponge and arranged
these measurements in a series according to the thickness of the cladomal end
of the rhabdome. Apart from a gap between thicknesses of 7-16 y this series
is fairly continuous. Although the gap between 7-16 y could probably be filled if
the search were continued long enough, it nevertheless indicates that anaclades
of these dimensions are not nearly so numerous as larger and smaller ones,
so that small and large anaclades can, to acertain extent, be distinguished.
This distinction is, however, not nearly so clearly marked as in the other Pacific
geodine sponges which possess small as well as large anaclades.

The small anatriaenes (Plate 42, figs. 3b, 10, 11, 23b, 24b, 25-27) have a
rhabdome 670 #-2.5 mm. long and, at the cladome, 2-7 » thick. In some of
these spicules (Plate 42, fig. 3b) the rhabdome is nearly éylindrical and rounded
at the acladomal end. In others (Plate 42, fig. 25) it is distinctly spindle
shaped, thickest in the middle, and attenuated towards both the cladomal and
the acladomal ends, the former being less than half as thick as the thickest,
central part of the rhabdome, and the latter pointed. The clades are slightly
curved, concave to the rhabdome. They arise either terminally (Plate 42, figs.
3b, 10, 23b, 25-27) or a little below the end of the rhabdome. In the latter
case a more (Plate 42, fig. 11) or less (Plate 42, fig. 24b) clearly pronounced
protuberance arises from the apex of the cladome. The chords of the clades of
these spicules are 6-43 » long and enclose angles of 41-65° with the axis of the
rhabdome.

The large anatriaenes (Plate 42, figs. 28-40) have a rhabdome up to 23 mm.
long, which is, at the cladome, 17-45 » thick. The clades are usually equal in
size. They arise nearly terminally, and there is only a slight protuberance on
the apex of the cladome. The clades are conic, pointed or somewhat blunt, and
curved, concave to the rhabdome. This curvature decreases distally, the ends
of the clades usually being nearly straight. An abrupt bend is sometimes ob-
served at the point where the more strongly curved basal part passes into the
nearly straight distal part (Plate 42, figs. 31, 39). The chords of the clades are
70-205 y long and enclose angles of 36-55°, on an average 42.5°, with the axis
of the rhabdome.

Although the different kinds of ewasters are connected by transitional forms
to a much greater extent than in most of the other geodine sponges, three kinds
can be fairly well distinguished. These are: 1. Large asters without centrum
and with slender, conic, pointed rays, in which the length of the rays is more
than five times as great as their basal thickness, These asters are here described

168 GEODIA OVIS.

as large thin-rayed oxyasters. 2. Large asters, here described as large thick-
rayed oxyasters, with very stout, conic, pointed rays, in which the length of
the rays is less than five times as great as their basal thickness. 3. Smaller
asters, here described as small thick-rayed asters, with stout, truncate, blunt or
pointed rays.

The large thin-rayed oxyasters (Plate 41, figs. 3, 9b, 12, 15, 19; Plate 42,
figs. 13b, 14b, 17b, 21b, 22b) are destitute of a central thickening and have
from three to ten usually quite concentric and simple, straight, conical, sharp-
or blunt-pointed rays. The proximal third of the rays is smooth, the distal two
thirds are covered by spines which are usually large and conspicuous (Plate 41,
figs. 15, 19), more rarely so small as merely to give to its distal part a slightly
roughened appearance (Plate 41, fig. 3). The rays are 11-18 » long and, at the
base, 1-3.2 thick, the total diameter of the aster being 20-34.5 p.

* The many-rayed large thick-rayed oxyasters (Plate 41, figs. 10b, 11b, 16-
18, 20; Plate 42, figs. 12-15a, 20a, 21a) appear, in consequence of the con-
erescence of the basal parts of the exceedingly thick rays, as sphaerasters; in
those with few rays, however, no trace of a central thickening can be detected.
The rays of the many-rayed forms are fairly concentric and regularly distributed
(Plate 41, fig. 16; Plate 42, figs. 12-15a, 20a, 21a), while those of the few-rayed
forms are sometimes eccentric and, as a rule, not regularly distributed (Plate
41, figs. 10b, 11b). The rays are straight, conic, and usually very sharp pointed.
Most of them are simple but in a good many of these asters one or even two of
the rays are bifurcate (Plate 41, figs. 10b, 11b). The extreme tip and the basal
portion of the rays are usually quite smooth, their remaining part covered with
large, vertically arising, terminally recurved, claw-like spines. In some of these
asters the spines are quite numerous, in others rather scarce and restricted to a
verticillate belt situated some distance below the end. These asters have from
four to nineteen rays, 14-24 » long and, at the base, 3-6.3 4 thick. In some of
the many-rayed forms the central thickening attains a diameter of 12 yp. The
total diameter of the aster is 28-45 p.

The small thick-rayed asters (Plate 41, figs. 1, 2, 4-8, 9a, 10a, lla, 13,
14; Plate 42, figs. 16c, 21c) are without centrum or have a central thickening
up to 6 win diameter. They have from six to fifteen concentric, regularly
distributed, and usually equal, more rarely unequal (Plate 41, figs. 13-14),
truncate, blunt or pointed rays. The distal parts of the rays are densely covered
with large spines which are not, as is the case in the large thick-rayed oxyasters,
restricted to an intermediate zone but extend right up to the tip of the rays.

GEODIA OVIS. 169

The spines arising from the terminal parts of the rays are directed obliquely out-
ward and upward. The rays are (without the centrum, when present) 5-12 y
long and, at the base, 1-3.2 » thick. The total diameter of the aster is 11-24 p.

The normal sterrasters (Plate 40, figs. 14-18, 24) are flattened ellipsoids,
82-92 yw long, 70-83 » broad, and 54-61 y thick. The average proportion of
length to breadth to thickness is 100 : 87 : 66. The rays protruding beyond the
surface are 3.5-4.5 m thick and have a terminal verticil of usually five to seven
lateral spines. In the sterraster represented in Plate 40, figs. 17-18, the rays
surrounding the umbilicus are hardly larger than the others. The umbilicus
is 8-15 » broad and about 10 y deep.

Besides these normal sterrasters a few sterroid forms (Plate 40, fig. 29) of
similar dimensions, but with protruding rays 7-8 broad, and provided with a
much larger number of lateral spines, are met with. I have also observed a
larger sterraster with exceedingly thin protruding rays in one of the spicule-
preparations, but this I believe to be a foreign spicule.

This sponge was trawled off southern California, at Station 2975 on Febru-
ary 12, 1889, in 34° 1’ 30” N., 119° 29’ W.; depth 66 m. (86 f.); it grew on a
bottom of gravel and broken shells; the bottom temperature was 13.9° (57° F.).
A label marked ‘‘506 Tetractinellida”’ was also attached to it.

As the afferents and efferents are, with little doubt, cribriporal, and as the
skeleton is geodine in character I place this sponge in Geodia.

Its nearest allies are Geodia kiikenthali Thiele, G. (Cydonium) miilleri
Fleming, and the sponge here described as G. mesotriaena.

Geodia kiikenthali is distinguished from G. ovis by the very much smaller
size of its megascleres, by the absence of small anatriaenes and a particularly
well-developed spicule-fur, by the smaller sterrasters, and by the euasters, the
largest of which have in G. kiikenthali blunt rays while the apparently corre-
sponding largest asters of G. ovis have markedly sharp-pointed rays.

G. miilleri has, when adult, extensive praeoscular and vestibular cavities,
absent in G. ovis, and no such highly developed spicule-fur as the latter. Its
megascleres are, although the largest specimens of G. miilleri examined were
considerably larger than the specimen of G. ovis, very much shorter than those
of the latter, and also its sterrasters somewhat smaller. The dermal asters of
G. miilleri are strongylasters with cylindrical rays and measure 4-10 y in total
diameter, while the superficial asters of G. ovis are 11-24 y in total diameter
and many of them have more or less conical rays. The remarkably large oxy-
asters with stout, conic, sharp-pointed rays of G. ovis have not been observed

170 GEODIA MICROPORA.

in G. miilleri, their place here being taken by strongylasters much smaller in
size.

Apart from the minute anatriaenes the megascleres of G. mesotriaena are
in shape and size very similar to those of G. ovis. The former is, however, dis-
tinguished from the latter by the possession of praeoscular cavities, the lower
spicule-fur, and the absence of minute anatriaenes. Its sterrasters are larger,
its dermal asters much smaller, and the large oxyasters with stout, regularly
conical, and sharp-pointed rays, so abundant in G. ovis, are wanting in G. meso-

triaena.

Geodia micropora, sp. nov.

Plate 36, figs. 13-36; Plate 37, figs. 1-14.

I establish this species for a specimen from Duncan Island, Galapagos.

The radial cortical canals, both the afferents and the efferents are very
narrow and their entrances, although covered with pore-sieves, appear to the
naked eye as very small, simple, open pores. To this the specific name refers.

Shape and size. The specimen (Plate 36, fig. 32) appears as a part of a
larger lobose mass, Measures 63 mm. in maximum diameter, and is attached
to some black pebbles and pieces of coral. The surface is fairly smooth. On
its protected parts remnants of a spicule-fur are observed. With the excep-
tion of a belt about 1.5 mm. broad, which extends for a considerable distance
and in which there are only few minute pores, the whole of the surface is
occupied by the small cribriporal entrances to the radial cortical canals. On
one side of the belt the entrances to the cortical canals are 50-200 » wide and
their centres on an average 350 » apart; on the other side they are 200-500
wide and their centres on an average 600 # apart. The small canal-entrances
on the one side of the belt I consider as afferents, the wider ones on the
other side as efferents.

The colour, in spirit, is brownish white, somewhat lighter on the surface
than in the interior.

The superficial part of the body forms a cortex (Plate 37, figs. la, 2a, 3)
450-600 y thick. This is composed of three layers, an outer dermal layer (Plate
37, fig. 3a) 60-101 y thick, a middle sterraster-armour layer (Plate 37, fig. 3¢)
300-420 y thick, and an inner fibrous layer 45-100 » thick. The latter is
occupied by paratangential fibres and by cells strongly staining with haema-
toxylin. Most of these cells are elongated and extend paratangentially.

Canal-system. The afferent area of the surface is covered with pore-sieves

GEODIA MICROPORA. 1 (|

170-330 » in diameter, the centres of which are on an average 350 y apart.
The pores in these sieves (Plate 37, figs. 8, 12) are very unequal, 30-130 » wide,
and separated from each other by thin and narrow bands of dermal tissue.
Each pore-sieve forms the roof of a cavity from which a radial afferent cortical
canal, circular in transverse section and 50-200 y wide, arises. These affer-
ent canals penetrate the sterraster-armour layer and open out into a sub-
cortical cavity (Plate 37, figs. 1, 2, 3d). From the latter the narrow afferent
choanosomal canals originate.

The flagellate chambers (Plate 37, fig. 14) appear to be sphaeroidal, de-
pressed in the direction of the oral axis, 20-26 » broad, and 12-17 whigh.

The choanosomal efferents join to form wide canal-stems (Plate 37, fig. 2b)
which extend radially towards the efferent area of the surface. Some distance
below the cortex these canal-stems divide into branches leading up to the effer-
ent radial cortical canals. The latter are 200-500 y wide and covered by sieves
(Plate 37, figs. 9, 13), the pores of which are larger and more equal than the
afferent pores, and usually 80-200 wide.

The skeleton consists of large choanosomal amphioxes, small dermal rhabds,
orthoplagiotriaenes, mesoproclades, large oxyasters, large oxysphaerasters,
small strongylosphaerasters, and sterrasters. In the spicule-preparations I also
found chelotrops, which, however, are probably foreign to the sponge. In the
interior of the choanosome the large amphioxes are irregularly scattered. In its
distal part they, together with the rhabdomes of the orthoplagiotriaenes, form
radial bundles which abut vertically on the cortex (Plate 37, figs. 1, 2). The
cladomes of the orthoplagiotriaenes lie in the level of the inner limit of the
sterraster-armour layer. The rhabdomes of the mesoproclades are situated
radially, their acladomal ends are implanted in the sterraster-armour layer, their
cladomes protrude freely beyond the surface. The cladomal parts of these
spicules appear to form the bulk of the spicule-fur. A few large, slender, radial
amphioxes, however, are also found init. Small dermal rhabds, mostly situated
obliquely, occur in considerable numbers in the subcortical layer, which appears
to be their place of origin (Plate 37, fig. 3c). From here they wander rapidly
— few only are found en route in the middle layer of the cortex — up to the
dermal layer, in which they take up their final position. Here they form tufts
implanted in the sterraster-armour layer. The distal parts of these tufts pro-
trude freely beyond the surface. These spicule-tufts occupy the tracts of
poreless tissue intervening between the pore-sieves. The tufts next the pore-
sieves are inclined towards their centre so as partly to shelter them. Most of

172 GEODIA MICROPORA.

the pore-sieves are surrounded by a ring of five or six regularly distributed
spicule-tufts; the intervals between these tufts are equal.

Hardly any euasters are observed in the interior of the choanosome; distally
large oxyasters are met with in small numbers. In the subcortical layer and in
the walls of the radial cortical canals large oxysphaerasters occur. The dermal
membrane is occupied by small strongylosphaerasters. These are more numer-
ous between the pore-sieves than in the narrow bands separating the individual
pores from each other. A large number of sterrasters, chiefly young forms,
occur in the interior cf the choanosome. In its distal part there are only few
sterrasters. In the proximal part of the sterraster-armour layer the sterrasters
are quite far apart; in its distal part they are closely packed together and form
a dense mass (Plate 37, fig. 3).

Some of the large amphioxes are simple, others centrotyle. Forms with
a hardly perceptible centrai thickening connect the centrotyle forms with the
simple ones. First I thought that the centrotyle amphioxes were young stages
of the others; since, however, centrotyles are found among the thickest of these
amphioxes and simple, non centrotyle ones among the thinnest, I hardly think
that this can be so. The large amphioxes are 1.2-1.6 mm. long, in the middle
20-28 y thick, and usually slightly curved. The tyle is always situated in the
thickest, central part of the spicule. The diameter of the properly developed
central tyle is, as the following measurements show, 12 to 30 % greater than the
thickness of the adjacent parts of the spicule. The thickest tyle observed

measured 31 in transverse diameter.

Thickness of amphiox close to the Transverse diameter of the
central tyle. central tyle.
21 pz 27 p
yy Se 25
22 26m
23 « 26 «
20) 5 ole
24 © 2

The minute dermal rhabds (Plate 36, figs. 24a, 26a, 27a) are amphistrongyles
attenuated towards both ends. They are generally isoactine, rarely anisoactine,
usually slightly curved, and 125-165 » long. In the middle they are 2-3.6 yp,
at the ends 1.8-2 thick, the thickness of the ends being on an average about
59 % of the maximum thickness in the middle.

The orthoplagiotriaenes (Plate 37, figs. 4-7) have a fairly straight, conic
rhabdome 1.1-1.45 mm. long and, at the cladome, 28-47 pethick. The acladomal
end is usually blunt pointed. Sometimes irregular rhabdomes (Plate 37, fig.

GEODIA MICROPORA. 173

7) are observed. The clades are rather blunt, sometimes nearly truncate, and
curved towards the rhabdome. In the basal part of the clade this curvature is
very considerable; distally it decreases, and the ends of the clades are only
slightly curved or even quite straight. The clades rise at rather large angles
from the rhabdome, but, in consequence of their strong curvature, their chords
enclose angles of only 97-112°, on an average 104.5°, with the axis of the rhabd-
ome. The greater number of these spicules appear as plagiotriaenes (with
clade-angles over 100°), about 20 % of them as orthotriaenes (with clade-angles
90-100°). The chords of the clades are 175-240 y long.

The cladomal end of the axial thread of the rhabdome is often varicose and
in some of the young orthoplagiotriaenes a slight thickening of the rhabdome
surrounds this part of the axial thread.

In the spicule-preparations I found several chelotrops (Plate 37, figs. 10, 11).
These have conical, pointed, usually straight, more rarely angularly bent rays,
195-260 long and, at the base, 21-27 » thick. I have not found any of these
spicules in situ in the sections. Although they coincide in their dimensions with
the orthoplagiotriaenes, I think it probable that they are foreign to the sponge.

The mesoproclades (Plate 36, figs. 13-17) have a rhabdome about 1.7 mm.
long and, at the cladome, 4-9 » thick. In its central part the rhabdome is
about 20 % thicker than at the cladome. The shape of the cladome is very
variable. The epirhabd is usually well developed, straight, conical, pointed, and
25-43 «long. Sometimes it is quite short, reduced to a mere knob. Of clades
there may be three (Plate 36, fig. 13), two (Plate 36, fig. 16), or one (Plate
36, fig. 17). In the monaene forms knob-shaped rudiments of one or two
other clades, usually situated at different levels, are often present (Plate 36,
fig. 14). The clades are conic pointed or blunt, and curved, concave to the
epirhabd. Their chords are 10-30 » long and enclose angles of 32-64°, on an
average 51°, with the epirhabd.

The large oxyasters (Plate 36, figs. 24-26d, 34b) are without central thickening
and have from six to nine, most frequently seven, concentric, regularly distributed
rays. The rays are straight, conic, 8-12 long, and very slender, at the base,
only 0.4-0.9 p, usually 0.6-0.7 » thick. Everywhere, except at the base, they
bear small spines. Towards the end the ray proper becomes exceedingly thin ;
the spines, however, which are here particularly dense, make its terminal part
appear thicker and its end somewhat blunt. The total diameter of the oxyasters
is 14-20 », usually 16-17 yu.

The large oxysphaerasters (Plate 36, figs. 18b, 19b, 26c, 33b) have a spherical

174 GEODIA MICROPORA.

centrum, 4-6 7, about a quarter to a third of the whole aster, in diameter, from
which from sixteen to twenty-two concentric, regularly distributed rays arise
radially. The rays are straight, conical, sharp pointed, without the centrum
6-7.5 # long and, at the base, 1-1.6 « thick. The rays bear small spines. The
latter are sometimes so minute as to be hardly discernible as such. Often some-
what larger spines form a verticillate belt some distance below the end of the ray.
The total diameter of the oxysphaerasters is 14-20 p.

The small strongylosphaerasters (Plate 36, figs. 18a, 19a, 20, 24b, 26b, 28,
33a, 34a) have a centrum 2-3 yp, a third to a half of the whole aster, in diameter,
from which from eight to fifteen, most frequently twelve, rays arise. These are
generally regularly arranged, concentric, radial, and equal, rarely irregular.
The rays are cylindrical or cylindroconical, truncate, and often provided with a

yterminal spine. They are 2-3 long, at the base 0.6-1.3 y thick, and bear rather
large spines, some of which often form a verticil some distance below the end of
the ray. The total diameter of the small strongylosphaerasters is 6—9.2 , usually
6-8 py.

Besides these I have observed similar but smaller strongylosphaerasters,
down to 3.4 in total diameter, which I take to be young forms of the ones
described above.

The sterrasters (Plate 36, figs. 21-23, 30, 31, 35, 36) are broad, flattened ellip-
soids 72-76 , rarely as much as 82 « long, 65-74 « broad, and 55-62 «thick. The
average proportion of length to breadth to thickness is 100 :95:85. In the
centre of the sterraster there is a cluster, 2.5-5 «in diameter, of small granules.
The umbilicus is a conic pit, 11-15 # deep. Seen in profile the sides of this pit
appear somewhat concave. Its entrance is usually irregular in shape, often
considerably elongated, and measures 11-18 » in maximum diameter. The
freely protruding rays which surround the umbilicus have a transverse section
elongated in a direction radial to the centre of the umbilicus, which measures
2-2.5 4 X 4. These rays bear as many as 8-12 lateral and several terminal
spines. The other protruding rays, away from the umbilicus, have more or less
circular transverse sections, which are 2.5-3.4 m in diameter. These rays
usually bear from 5-7 lateral spines. In sterrasters not quite fully developed
the spines are slender (Plate 36, figs. 35, 36), in the fully developed ones they
are stout (Plate 36, figs. 30, 31).

This sponge was collected on April 13, 1888, at Duncan Island, Galapagos.
It was labeled F. C. 1354 and 539 Tetractinellida.

The structure of the skeleton and the cribriporal character of the afferents

GEODIA AMPHISTRONGYLA. io

and efferents show that this sponge belongs to Geodia. Its nearest allies appear
to be Geodia paupera Bowerbank, G. (Pachymatisma) inconspicua Bowerbank,
and G. (Cydonium) cooksoni Sollas.

The locality of Geodia paupera is unknown and the locality given of G.
inconspicua, “‘South Seas”’ is hardly definite enough to be of value. Both these
species have smaller sterrasters, no mesoproclades, and no centrotyle amphioxes.
The choanosomal asters of G. pawpera are strongylasters and not oxyasters as in
G. micropora, and the megascleres of G. inconspicua much larger than those of
G. micropora. 1 do not doubt, therefore, that these sponges are distinct from
Geodia micropora.

I found it much more difficult to decide the question whether or not this
species is specifically identical with Geodia (Cydonium) cooksoni Sollas, which in
many respects resembles it and which also comes from the Galapagos (Charles
Island). The description given by Sollas ' is exceedingly meagre and unaccom-
panied by illustrations. I therefore asked my friend Mr. Kirkpatrick of the
British Museum to send me a part of the type specimen of this species for exami-
nation and he had the kindness to accede to my request. I find that the sterr-
asters of G. cooksoni have nearly the same size, but are relatively thinner, than
those of G. micropora, and that the dimensions of the other spicules are
larger. This difference in size is particularly marked in the choanosomal asters,
the largest of which are in the type of Geodia (Cydonium) cooksoni fully twice
as large as the largest of G. micropora. It is for these reasons and also because
the mesoproclades have smaller clade-angles (average in G. micropora 51°, in G.
cooksoni 44°) and the large euasters relatively much larger centra in Geodia
(Cydonium) cooksoni than in G. micropora that I describe the latter as a new
species.

Geodia amphistrongyla, sp. nov.
Plate 20, figs. 1-41.

I establish this species for two fragments from Easter Island which may be
parts of the same specimen. Its choanosomal rhabds are not, as is the rule
in geodine sponges, chiefly amphioxes, but chiefly amphistrongyles, amphioxes
being absent altogether. To this remarkable peculiarity the name selected for
the species refers.

Shape and size. The larger fragment is an irregular, somewhat flattened
mass (Plate 20, fig. 31) 31 mm. long, 18 mm. broad, and 12 mm. thick. The
other, considerably smaller fragment is similar in shape. The surface is un-

1 Rept. voy. ‘‘ Challenger,” 1888, 25, p. 255, 256.

176 GEODIA AMPHISTRONGYLA.

dulating and not perforated by larger openings visible to the unaided eye. To
parts of it shells, sand, and other foreign bodies are attached. It appears to be
covered throughout with shallow pits, the centres of which are about 1 mm.
apart. There is hardly a trace of a spicule-fur, but the numerous slender spicules
broken off at the surface, seen in sections, indicate that protruding spicules were
present in the living sponge.

The colour, in spirit, is light brown on the surface and quite dark brown in
the interior.

The superficial part of the body forms a cortex composed of a thin outer
dermal membrane, thickened in the pits; a stout central sterraster-armour
layer (Plate 20, figs. 33a, 39a) about 1 mm. thick; and an inner fibrous layer,
excavated by subcortical cavities. A thin cuticular membrane was observed on
parts of the surface. Whether this belongs to the sponge or whether it was
formed by some symbiont growing on it, I cannot state with certainty.

Canal-system. Below the centre of each pit a radial cortical canal, penetrat-
ing the sterraster-armour, is situated. Above, these canals widen out to form
subdermal cavities which are covered by pore-sieves, the pores of which are
circular or broad-oval, 17-45 y wide, and separated by quite broad strands of
tissue. Below, the cortical canals are restricted by chonal sphincters (Plate
20, fig. 33b) which protrude proximally beyond the lower limit of the sterraster-
armour layer. The chonal canals passing through the sphincters, which in my
preparations are strongly contracted and often quite closed, open out into the
subcortical cavities (Plate 20, fig. 38c). From the latter the choanosomal
canals arise. ;

Skeleton. Numerous slender strands of amphistrongyles and a few styles
traverse the interior in a rather irregular manner. Towards the surface these
strands assume a radial direction and become consolidated to form bundles,
which penetrate the cortex and abut vertically on the surface. Here, in the
distal part of the choanosome and in the cortex, rhabdomes of telo-and meso-
clades are found in the bundles besides the amphistrongyles and styles. These
telo- and mesoclades are plagiotriaenes, anatriaenes, mesotriaenes, and a few
diaene, monaene, and other derivates of these spicules. The cladomes of some,
chiefly young, forms of these spicules are found in small numbers at various
levels. The cladomes of nearly all the adult plagiotriaenes and their diaene and
monaene derivates lie above the sterraster-armour, just below or in the dermal
membrane. Here also the cladomes of a few, apparently adult mesoproclades,
anatriaenes, and anatriaene-derivates are found. The slender spicules travers-

GEODIA AMPHISTRONGYLA. 17%

ing the cortex radially and broken off at the surface; are probably the proximal
parts of rhabdomes of these anatriaenes and mesoclades, which composed the
spicule-fur in the living sponge.

The microscleres are oxyasters, oxysphaerasters, small strongylosphaer-
asters, and sterrasters. The oxyasters are confined to the choanosome and are
not numerous; the oxysphaerasters occur chiefly in the walls of the cortical
canals and in the roofs of the subcortical cavities. The small strongylosphaer-
asters form a dense layer on the outer surface. The sterrasters occupy the
sterraster-armour layer of the cortex in dense masses and are scattered also in
large numbers throughout the whole of the choanosome (Plate 20, figs. 33, 39).
The choanosomal sterrasters are mostly young forms.

The large amphistrongyles (Plate 20, figs. 1-3, 17-19) are slightly or con-
siderably (Plate 20, figs. 2, 3) curved, 0.5-2.3 mm. long, and, in the middle,
18-32 » thick. Their thickness is by no means in proportion to their length,
the shortest amphistrongyles being quite as thick as or thicker than the longest.
The long amphistrongyles (Plate 20, figs. 1-3) usually taper from the centre
towards both the rounded ends, the thickness of the latter being 40-75 % of the
thickness of the former. This attenuation may be equal at both ends — then
amphistrongyles appear as isoactines; or it may be unequal — then the amphi-
strongyles appear as anisoactines. The shorter the spicules are, the less marked
‘is the attenuation towards theirends. The spicule represented on Plate 20, fig. 19,
which is 1.05 mm. long, is at the ends 17 » and 20 y, only in the middle 22 4
thick. Still shorter amphistrongyles are, if isoactine, regularly cylindrical or
slightly thickened at the ends, or, if anisoactine, club shaped, like the spicule
represented on Plate 20, fig. 18, in which the thickest part is situated at one
end. Slender amphistrongyles with knob-shaped thickenings (Plate 20, fig. 17),
which may be young forms, have been observed occasionally. It is certain
that the short, stout amphistrongyles do not develop into the long, slender ones.
That the former are cylindrical and often more or less thickened at the ends,
while the latter are never thicker and usually more slender and attenuated
towards the ends, leads me to suppose that the silicoblasts, which build up the
ends of these spicules, wander, while they perform their allotted task, either
slowly or rapidly in a distal direction. In the first case short, thick, and more
cylindrical, in the second case long and relatively slender amphistrongyles,
attenuated towards the ends, are produced.

The rare large styles (Plate 20, figs. 20, 21) have similar dimensions as the
amphistrongyles above described and may be considered as derivates of aniso-

178 GEODIA AMPHISTRONGYLA.

actine forms of such, in which the thinner end has become attenuated to a fine
point.

The plagiotriaenes (Plate 20, figs. 22, 25) have conical, usually quite straight
rhabdomes, 1.8-2.2 mm. long and 22-32 thick at the cladomal end. The
acladomal end is usually simply rounded off and 3-10 4 thick, rarely slightly
thickened. The clades are slightly curved, concave to the rhabdome, conical,
usually regularly arranged, with angles of 120° between them, and, in the same
cladome, of equal size and in nearly the same position to the rhabdome. The
clades are 155-190 » long and enclose angles of 103-120°, most frequently about
109°, with the rhabdome. Rarely the clade-rhabdome angles of the same
rhabdome are very dissimilar. In one plagiotriaene-cladome of this kind one
of the three clades was nearly upright, resembling in its position an epirhabd.
This sword-like spicule might be termed a mesorthodiaene.

Among these plagiotriaenes similar forms with reduced cladomes, which
appear as plagiodiaenes or plagiomonaenes occur. ‘The monaene forms are more
frequent than the diaene. The cladome of a normal monaene of this kind is
shown in Plate 20, fig. 23. Very rarely the rhabdome is reduced in length and
in that case cylindrical and slightly thickened at the acladomal end. One of
these spicules (Plate 20, fig. 24) has a rhabdome, only 285 » long, and a single,
bifureate clade, arising at an angle of hardly more than 90° from the rhabd-
ome. This spicule is an orthodichomonaene. Dichoclade forms of this or
any other description are, however, exceedingly rare.

Owing to the loss of nearly the whole of the spicule-fur I found but few
mesoproclades (Plate 20, figs. 7, 8), which presumably form the greater part of
this fur. None of them had an intact rhabdome, so that I cannot give its length.
Most of the mesoproclades observed were triaene, but monaene forms (Plate 20,
fig. 7) have also been met with. The rhabdomes of these spicules are, at the
cladome, 3-5 » thick; the epirhabds are straight, conical, sharp pointed, and
16-22 » long; the clades are concave to the epirhabd, usually rather blunt, and
40-60 » long. The clade-epirhabd angles of the intact cladomes observed
were 36-41°, but some broken mesoprotriaene cladomes seen indicate that
occasionally this angle is considerably larger.

The anatriaenes (Plate 20, figs. 5, 6, 10, 11) are, like the mesoproclades,
scarce in the preparations, and probably for the same reason. Anatriaenes
with intact rhabdomes were not observed. Their rhabdomes are 1.5-4 thick
at the cladomal end, the clades are conical and sharp pointed, more (Plate 20,
figs. 5, 10) or less (Plate 20, figs. 6, 11) curved, coneave to the rhabdome in their

GEODIA AMPHISTRONGYLA. 179

proximal portion and straight in their distal portion. Their chords are 26-50 px
long and enclose angles of 25-41° with the rhabdome.

I found one remarkable anatriaene-derivate (Plate 20, fig. 4) which possesses,
besides the three recurved anatriaene-clades and the rhabdome, a fifth ray,
directed obliquely upwards. This spicule appears as a mesanatriaene with
oblique epirhabd.

The choanosomal oxyasters (Plate 20, figs. 12-16, 26-30a) have no central
thickening or a hardly perceptible one, and usually straight and simple, equally
distributed rays. Very rarely one or more rays are either curved (Plate 20, figs.
13) or bifureate. The rays are conical, smooth at the base, and spined in their
distal part. In the thin-rayed, probably young, oxyasters the spines are so
small as merely to give to the rays, when examined with the highest powers, a
slightly rough appearance. In the thick-rayed, presumably adult, forms the
spines are usually large, sometimes nearly 1 » long. They arise vertically or
slightly obliquely from the rays, and are in the latter case directed upwards,
towards the tip of the ray. All the spines which were large enough to be dis-
tinctly made out, were straight. Recurved, claw-like spines were not met with.
Sometimes (Plate 20, figs. 13, 16 right upper corner) the spines are massed at the
tip of the ray, more frequently (Plate 20, figs. 14, 15, 16 middle, left) they are
sparsely scattered over its distal two thirds. The ends of stout rays are often
crowned by one particularly large, terminal spine. These oxyasters have from
five to nine rays and measure 20-30 yin total diameter. The rays are 10-15
long and at the base 0.82.1 4, usually 1.5-1.8 » thick. A correlation between
the number of the rays and the dimensions of the aster is not noticeable.

The oxysphaerasters (Plate 20, figs. 26b, 27b, 30b) have a spherical centrum
6-7 win diameter, from which from fourteen to eighteen stout, straight, conical,
and sharp-pointed, equally distributed rays arise radially. The rays are 6-11
felong and, at the base, 2-2.8 » thick. From their distal part spines, which are
usually quite large, arise. The base of the ray is always free from spines. A
rather regular verticil of particularly large spines, situated some distance below
the end of the ray, is sometimes observed. The total diameter of these spicules
is 19-28 ». A correlation between the number of their rays and the dimen-
sions of the aster is not discernible.

The small strongylosphaerasters (Plate 20, figs. 28c, 29¢, 34-36) consist of
a spherical or somewhat irregular centrum, 2.2-3 y4, rarely as much as 4 # in diam-
eter, from which from seven to twelve rays arise. These are regularly distributed
and in the same aster equal in size, or, more rarely, unequal in position and

180 GEODIA AMPHISTRONGYLA.

dimensions. The rays are cylindrical or slightly thickened distally and have a
flat or convex terminal face from which a bunch of stout spines arises. The spines
forming this bunch are sometimes nearly parallel to each other and to the axis
of the ray, sometimes they diverge, occasionally so much so that the outermost
ones become nearly vertical to the axis of the ray. Generally the sides of the
rays are quite smooth. Sometimes, however, I have observed small spines on
them. The rays are 1-2 »long and 0.8-1.8 4, usually about 1 «thick. The total
diameter of the strongylosphaerasters is 4.8-8 y, usually 5-6.5 ». The strongylo-
sphaerasters with numerous rays are on the whole smaller than those with few
rays. Thus those with ten or more rays measured were 4.8-6.5 y, those with
from seven to nine rays 6-8 4, in diameter.

Besides these normal strongylosphaerasters I found a few sphaerasters
similar in size, with much more numerous (from fourteen to nineteen) and more
slender, conical rays. Perhaps these were foreign, or — which, however, does not

seem very probable — young forms of the oxysphaerasters.

The sterrasters (Plate 20, figs. 9, 28d, 30d, 32, 37, 38, 40, 41) are flattened
ellipsoids and measure 100-110 # in length, 87-94 » in breadth, and 72-78 4
in thickness. The average proportion of length to breadth and to thickness is
100:87:74. The young forms, which are very abundant in the choanosome,
are enclosed in endothelial capsules (Plate 20, fig. 9) and composed, as usual, of
slender rays the proximal parts of which coalesce to form a solid mass (Plate
20, figs. 9, 28d, 30d). In the centre of this a few irregularly distributed granules
are observed. The distal parts of the rays projecting freely over the surface of
the adult sterrasters (Plate 20, figs. 37, 38, 40, 41) are everywhere, except close
to the umbilicus, 3-3.5 thick and provided with terminal verticils of from
three to five stout, conical, and straight lateral spines, 1.5-2 » long and at the
base 1.5 # thick. The rays surrounding the umbilicus have a transverse sec-
tion, elongated in a direction radial to the centre of the umbilicus, which meas-
ures 3-3.5 win breadth and 4-4.5 » in length. These rays are provided with a
greater number, some with as many as nine, terminal, lateral spines. The
spines of these rays pointing inwards, towards the umbilicus, are larger than the
others, up to 2.5 long, and more or less curved.

This sponge was collected on the shore at Easter Island on December
20, 1904.

On account of its cribriform pores and its spiculation this sponge must be
placed in Geodia. I have compared it with the known species of Geodia and of
Sidonops. The only species of these genera with similar sterrasters, in which

GEODIA LOPHOTRIAENA. 181

the choanosomal rhabds are not, as usual, for the most part amphioxes, but, as
in Geodia amphistrongyla for the most part amphistrongyles, is Sidonops (Geodia)
flemingii (Bowerbank).’ This species resembles Geodia amphistrongyla not
only in respect to the rhabds and sterrasters, but also in respect to the euasters.
Since, however, in Geodia amphistrongyla the plagioclades (orthoclades) are simple
triaenes and the anatriaene-cladomes very small, while in Sidonops (Geodia)
flemingii the former are dichotriaenes and the latter quite large; since there
appears to be a difference in the efferent pores which necessitates the placing
of the two in two distinct genera; and since the one comes from the south coast
of Australia and the other from Easter Island, there can, I think, be no doubt
that they should be kept specifically distinct.

Geodia lophotriaena, sp. nov.

Plate 47, figs. 9-36; Plate 48, figs. 1-34.

I establish this species for seven spirit specimens which came probably
from New Zealand. They possess, besides ordinary dichotriaenes, a good many
lophotriaenes with more than two end clades on one or two or all three main
clades and to this the name refers.

The seven specimens are all cushion shaped and cut off from the surface
on which they grew and to which they appear to have been attached by broad
bases. Their upper surface is convex, their contour irregularly circular,
rounded polygonal, or elongated. The smallest of them is 7 mm. long, 6 mm.
broad, and 2.5 mm. high; the largest, which is penetrated in the middle (Plate
47, fig. 22), 15.5 mm. long, 11 mm. broad, and 3 mm. high. The surface is
entirely destitute of a spicule-fur and appears, when viewed with a lens,
shagreened. It is quite continuous, neither depressions nor apertures, visible
with the unaided eye, occur in it.

The colour, in spirit, is brownish, lighter on the surface than in the interior.

The superficial part of the body is differentiated to form a cortex (Plate 47,
fig. 24a), in which an outer dermal layer (Plate 47, fig. 23a), free from
sterrasters, rich in sphaerasters, and 30-65 thick, and an inner sterraster-
armour layer (Plate 47, fig. 23b) 60-125 yp thick, can be distinguished.

The cortex is penetrated by numerous radial canals which are 150-250 4
apart. Those observed were strongly contracted, only 5-15 « wide. These

17, S. Bowerbank. Contributions to a general history of the Spongiadae. IV. Proc. Zool. soc.
London, 1873, p. 3, pl. 1, figs. 1-8, W, J. Sollas. Tetractinellida. Rept. voy. “Challenger”, 1888, 25,

p- 252,

182 GEODIA LOPHOTRIAENA.

-anals are surrounded by mantles of tissue, about 20-30 » thick, which are free
from sterrasters.

Skeleton. In the internal (basal) part of the choanosome loose, irregular
strands of amphioxes occur. These mostly extend more or less paratangentially.
In this region also large, smooth oxyasters and smaller, spined sphaerasters,
mostly strongylosphaerasters, are met with. The distal, subcortical part of
the choanosome is traversed by radial bundles composed of amphioxes and the
rhabdomes of plagiotriaenes, dichotriaenes, other lophotriaenes, anatriaenes,
and mesoprotriaenes. Asters, similar to those of the interior, and minute,
radially arranged, mostly amphiox-rhabds also occur in this region. The
sterraster-armour is oecupied by not very closely packed sterrasters lying from
three to five deep. In the dermal layer large masses of sphaerasters, which are
particularly densely packed just below the surface, in the dermal membrane,
are met with. More or less radially arranged minute dermal rhabds, for the
most part amphioxes, are implanted in the cortex. Many of these spicules
traverse the whole of it, their proximal ends being imbedded in the distal part
of the choanosome and their distal ends protruding freely beyond the surface.
A few minute dermal anatriaenes also occur in this region. The cladomes of
the plagiotriaenes, the dichotriaenes, and the other lophotriaenes lie just below
the sterraster-armour (Plate 47, fig. 23). At a slightly lower level a good many
anatriaene-cladomes and a few mesoprotriaene-cladomes are met with. The
relative frequency of the plagiotriaenes and the dichotriaenes (lophotriaenes)
is different in different specimens. In some the simple plagiotriaenes, in others
the dichotriaenes (lophotriaenes), appear to form the majority.

Most of the large choanosomal amphiozes (Plate 48, figs. 3-6) are quite stout,
1.2-1.8 mm. long and 25-42 » thick. The great majority of those thinner than
25 ware shorter than 1.2 mm. and appear as young forms of the stout amphioxes.
Occasionally, however, long and very thin amphioxes (Plate 48, fig. 7) are met
with, which can hardly be considered as young stages of the stout ones. One
of these slender spicules was 2 mm. long and only 10 » thick.

Most of the minute dermal rhabds (Plate 48, figs. 8, 9, 10a) are rather blunt
pointed, fairly isoactine amphioxes, but anisoactine forms, with one end more
slender and more sharply pointed than the other, also occur. In some of these
spicules (Plate 48, fig. 10a) the axial thread (canal) is greatly widened towards
the thicker end. The minute dermal rhabds are 110-200 » long and usually
3-6 p thick.

The plagiotriaenes (Plate 47, figs. 17, 21) usually have a conical rhabdome

GEODIA LOPHOTRIAENA. 183

pointed at the cladomal end (Plate 47, fig. 17), 0.6-1.2 mm. long and, at the
cladome, 25-35 y» thick. In some the rhabdome is greatly reduced in length,
cylindrical, and terminally rounded. A spicule of this kind (Plate 47, fig. 21)
has a rhabdome only 0.2 mm. in length. The clades are usually fairly equal,
conical, pointed, curved, concave to the rhabdome throughout their entire
length, and 140-195 » long. Rarely one clade is reduced in length and termi-
nally rounded. The cladome has a breadth of 240-370 4. The chords of the
clades enclose angles of 102-114°, on an average 107.4°, with the axis of the
rhabdome.

The dichotriaenes and the other lophotriaenes (Plate 47, figs. 9-16, 18-20,
25-33) usually have a simple, fairly straight rhabdome the cladomal half of
which is nearly cylindrical, the acladomal half conical and sharp pointed. Just
below the cladome the rhabdome is often slightly constricted. Very rarely a
downwardly directed branch-ray arises from the central part of the rhabdome
(Plate 47, fig. 20). The rhabdome is 0.8-1.2 mm. long and, at the cladome,
35-59 yw thick. The three main clades of the same cladome are usually equal,
straight, cylindrical, and 70-140 » long. They enclose, with the rhabdome,
angles of 105-130°, on an average 120.8°. The end clades are conical and blunt
pointed. When, as is the case in the dichotriaenes, there are only two end
clades on one main clade, the end clades are curved more or less, concave towards
each other (Plate 47, figs. 25-29). When, as is the case in the other lophotriaenes,
there are more than two end clades on one main clade, one end clade, or rarely
two (Plate 47, fig. 31) are straight and appear as continuations of the main clade,
the other end clades being curved more or less, concave towards these (Plate 47,
figs. 30-33). The end clades of the same cladome are usually unequal. The
true dichotriaenes are much more numerous than the other lophotriaenes. In
the latter one (Plate 47, figs. 30, 33), two (Plate 47, fig. 31), or, more rarely, all
three (Plate 47, fig. 32) main clades bear from three to five, instead of only two,
end clades. In the most regular dichotriaenes the end clades extend in a plane
nearly vertical to the rhabdome or are directed obliquely upwards (Plate 47,
figs. 9, 10, 13, 18). In other dichotriaenes and in most of the other lopho-
triaenes not all of the end clades are in this position, some being directed
downward (Plate 47, figs. 11, 12, 14, 19, 20). The length of the end clades is
70-180 y, the breadth of the whole cladome 300-500 yu. The axial threads of
the end clades of the dichotriaenes arise nearly at right angles from the end of
the axial thread of the main clade and then bend outward to assume the direc-

tion of the end clade.

184 GEODIA LOPHOTRIAENA.

The mesoprotriaenes (Plate 47, fig. 34) have a fairly straight rhabdome,
about 1.3 mm. long and, at the cladome, 4-9 y thick; in its thickest part, near
the middle of its length, it measures 7-11 in transverse diameter. The clades
are fairly equal, slender, conical, pointed, and curved, concave to the epirhabd.
Their chords are 44-80 long and enclose angles of 32-41°, on an average 35°,
with the axis of the epirhabd. The epirhabd is straight, conical, and 38-60
long.

The large anatriaenes (Plate 47, figs. 35, 36). I found no large anatriaenes
with the rhabdome intact, so that I am unable to give its length. At the cla-
dome the rhabdome is 8-13 im thick. The cladome is without apical protuber-
ance. The clades are fairly equal, slender, conical, and sharp pointed. Their
proximal parts are quite strongly curved, concave to the rhabdome, their distal
parts straight. The chords of the clades are 60-85 » long and enclose angles
Mot 41—50°, on an average 44.5°, with the axis of the rhabdome.

The minute dermal anatriaenes have a fairly straight rhabdome, 170-210
long. At the cladome the rhabdome is 1-3 yp, at its thickest point, near the
middle, 2-4 » thick. The acladomal end is pointed. The fairly equal clades
are conical, blunt, and not very strongly curved, concave to the rhabdome.
Their chords are 4-9 » long and enclose angles of 49-67°, on an average 56°,
with the axis of the rhabdome.

The large oxyasters (Plate 48, figs. 1, 2, 11, 12b, 16-18, 21, 32b) are either
without a central thickening or have asmall centrum up to 4 y, that is, one eighth
to one tenth of the whole aster in diameter. There are from four to eleven,
rather regularly distributed and usually equal, straight, conical, sharp pointed
rays. The rays are perfectly smooth, 8-22 long and, at the base, 1—2.2 y thick.
Very rarely one or two of the rays are reduced in length and terminally rounded.
The total diameter of the aster is 15-41 yw. A correlation (inverse proportion)
between ray-number and total diameter is indicated, the oxyasters with from
four to seven rays being 23-41 w, those with from eight to eleven rays only
15-30 ys, in diameter.

The sphaerasters (Plate 48, figs. 10b, 12a, 13-15, 19, 20, 22-26, 32a) have a
centrum 2-8 y, from one sixth to nearly one half of the whole aster, in diameter.
From this from seven to twenty-two, rarely as many as twenty-eight, quite
regularly distributed rays arise radially. The rays are usually cylindroconical,
and attenuated more (Plate 48, figs. 23-26) or less (Plate 48, figs. 20, 22, 23)
towards their ends, the end itself being truncate or rounded and crowned by a
terminal spine; much more rarely the rays are conical and pointed (Plate 48,

GEODIA LOPHOTRIAENA. 185

fig. 19). The rays are as a rule covered with spines which increase in size
towards the end of the ray, the largest often forming a verticil at or just below
its end; smooth rays (Plate 48, fig. 19) are exceedingly rare. The rays are,
without the centrum, 3-8 y long, and, at the base, 0.8-2.4 » thick. The diam-
eter of the whole aster is 7-22 4. The few-rayed of these asters are larger,
have smaller centra, and more slender rays than the many-rayed. In the cen-
trifugal spicule-preparations I have found a few smaller strongylosphaerasters,
the total diameter of which was only 4.5 ». These rare asters may be young
forms of the ones described above, or foreign.

The sterrasters (Plate 47, fig. 23; Plate 48, figs. 27-31, 33, 34) are sphaeroids
or relatively very broad flattened ellipsoids. They usually measure 30-45 y in
length, 30-44 » in breadth, and 27-35 in thickness. Often the sterraster has,
when viewed from above (with the umbilicus in the centre), a regularly circular
outline, but even when this outline is oval the differences between the two axes
is quite insignificant. The average proportion of length to breadth to thickness
of the sterrasters is 100 :99 :87. In the spicule-preparations I found, besides
these sterrasters, a few larger ones, similar in shape, but 57-58 y» long (Plate 48,
fig. 28). These may, very likely, be foreign spicules.

The umbilicus is 6-9 » broad and about 6 #4 deep. The rays away from it
are about 2 y thick and have circular or polygonal transverse sections (Plate
48, figs. 33, 34); the rays surrounding the umbilicus are 3-5 y thick and have
transverse sections more or less elongated in a direction radial to the umbilicus
(Plate 48, figs. 30, 31). The ends of the rays bear from seven to eleven small
spines. From one to three of these arise from the terminal face, the others are
vertical to the axis of the ray and form a verticil round its end.

The seven specimens of this sponge are labeled 6312. In the list of the
specimens sent, 6312 does not occur, nor, except 6311, any number near it. There
is therefore some probability that 6312 should be 6311, the locality of which is
New Zealand, so that New Zealand may be the habitat of these sponges.

Although I have not been able to make out quite clearly the nature of the
entrances and exits of the canal-system, I think there can be little doubt that
both are cribriporal. For this reason and because the skeleton is distinctly
geodine in character, I place these sponges in Geodia. The sponges described
by Kieschnick ' as Cydonium sphaeroides, by Lindgren * as Geodia arripiens, and

10. Kieschnick. Silicispongiae von Ternate. Zool. anz., 1896, 19, p. 529.
2.N. G. Lindgren. Beitrag zur kenntniss der spongienfauna des Malayischen Archipels. Zool.
jahrb. Syst., 1898, 11, p. 346, plate 18, figs. 10, 18, plate 20, fig. 5.

186 GEODIA LOPHOTRIAENA.

by Thiele ' as G. sphaeroides seem to be allied to them. These sponges I consider”
one species, Geodia sphaeroides.
The chief characters of Geodia sphaeroides and G. lophotriaena are tabulated

below.
Geodia sphaeroides Geodia lophotriaena
small, massive, oval or spherical; 4 t 5
- Se ae ene small, inerusting, cushion shaped;
Shape in one ease at least with praeoscu- = :
; without praeoscular cavity.
lar cavity.
Cortex 720-900 thick. 90-190 » thick.

Large choanosomal am-
phioxes

2.2 or 2.4 mm. by 36 or 40 y.

up to 1.8 mm. by 42 ” and 2 mm.
by 10 p.

Minute dermal amphi-
oxes

230 by 5

up to 200 by 6 1.

Plagiotriaenes

not mentioned.

always present, in some specimens
more numerous than the dichotri-
aenes; rhabdome up to 1.2 mm.

by 35 »; clades up to 195 p.

Dichotriaenes
(lophotriaenes)

rhabdome 2.35 or 3 mm. by 60 or
70 yt; main clades 120 or 220;
always two end clades 150 or 180 1.

rhabdome up to 1.2 mm. by 59 p;
main clades up to 140»; two or
more end clades up to 180 4.

Proclades

protriaenes; rhabdome 2.5 or 3
mm. by 16 or 20 ; clades stout,
frequently irregular, 70 or 80/4.

mesoprotriaenes; rhabdome 1.3
mm. by up to 9 (11) #; clades
slender, regular, up to 30 p; epi-
rhabd up to 60 1.

Large anatriaenes

rhabdome up to 3.3 or 3.5 mm. by

| 18 or 20 4; clades stout, 50 or 60 p.

rhabdome up to 13 » thick; clades
slender, up to 85 p.

Small dermal or sub-
cortical anatriaenes

thabdome 340 or 360 by 2-3 p;
clades 8 ys.

rhabdome up to 210 by 3 (4) 4;
clades up to 9 y4.

Large oxyasters (oxy-

36 or 50 » in diameter; with large

15-41 » in diameter; without cen-
trum or with small centrum, not

sphaerasters ant 6 é ; Tays spi 5

sphaerasters) centrum (6 or 15 #4); rays spined morethan 4 rays pectoctieateeren
Small sphaerasters

(mostly strongylo- 8-15 in diameter. 7-22 y in diameter.

sphaerasters)
ia eee Sar 88 alone 30-45 1, exceptionally perhaps up

to 58 » long.

Habitat Cochin China; Ternate. probably New Zealand.

1 J. Thiele. Wieselschwimme von Ternate I. Abhandl. Senckenb. gesellsch., 1900, 25, p. 41,

plate 2, fig. 14.

2 R. v. Lendenfeld. Tetraxonia. Tierreich, 1903, 19, p. 110.

GEODIA LOPHOTRIAENA. 187

This table shows that these two species, although very similar in many
respects, differ considerably in others. The difference in the shape of the body,
the thickness of the cortex, and the length of some of the megascleres might be
considered as due to differences of age or to differences in individual adaptation.
That plagiotriaenes are present in G. lophotriaena and not mentioned as occurring
in G. sphaeroides would, in itself, also hardly be sufficient for systematic distinc-
tion, because their relative abundance varies in the different specimens of G.
lophotriaena and because according to Topsent’ only ortho- plagio-triaenes
or only dichotriaenes or both these kinds of spicules may be present in another
species of Geodia, G. conchilega. That Thiele and Lindgren describe the pro-
clades of G. sphaeroides (G. arripiens) as protriaenes, while they are mesopro-
triaenes in G. lophotriaena, is also of but little importance, since it is known that
mesoprotriaenes have often been described as protriaenes. That lophotriaenes
with more than two end clades occur in G. lophotriaena, besides the ordinary
dichotriaenes, while only true dichotriaenes with two end clades have been ob-
served in G. sphaeroides, might be a more important difference, if one could only
be convinced, as I am not, that such lophotriaenes are really altogether absent in
the latter. Greater importance than to this is, in my opinion, to be attached to
the differences in the clades of the mesoprotriaenes (protriaenes) and anatriaenes,
which are very stout and, in the mesoprotriaenes (protriaenes), often partly
reduced in length and irregular in G. sphaeroides, but regularly developed to their
full length and slender in G. lophotriaena. Still more important than these
differences are those of the sterrasters and large euasters. The sterrasters are in
G. sphaeroides twice as large as in G. lophotriaena, while there is no corresponding
difference in the size of the other spicules and the whole body. The large euasters
are in G. sphaeroides oxysphaerasters, 36-50 in diameter, with a large centrum
and stout, spined rays; in G. lophotriaena oxyasters, 15-41 in diameter, with a
very small centrum or no central thickening at all, and perfectly smooth, slender
rays. These differences, particularly the last named, appear to be germinal
in nature and, particularly when taken together with the others, in themselves
unimportant, quite sufficient for specific distinction.

1B. Topsent. Etude mongraphique des spongiaires de France I. Arch. zool. expér., 1894, ser. 3,
2, p. 326.

18S GEODIA ACANTHTYLASTRA.

Geodia acanthtylastra, sp. nov.

Plate 45, figs. 1-39; Plate 46, figs. 1-21; Plate 47, figs. 1-8.

I establish this species for five spirit specimens from the coast of Lower
California (Station 2829). The choanosomal asters are chiefly acanthtylasters
and to this the name refers.

Apart from differences in age which find their expression in differences of
size, the five specimens are identical in structure. They are (Plate 465, figs. 16,
29) irregularly spherical, oval, or tuberous, the smaller ones being more regular
than the larger. The smallest specimen measures 10 by 8 mm., the largest 20 by
13 mm. In sheltered places, chiefly in the vicinity of the base of attachment,
remnants of a spicule-fur are observed. The surface is minutely pitted. The
pits, which mark the positions of the entrances to the radial cortical canals,
* appear in some specimens to be uniformly distributed and everywhere a little
under 0.5 mm. apart. In other specimens the pits are larger and more distant,
in one or two restricted areas on an average 1 mm. apart. Larger apertures,
visible to the unaided eye, do not occur on the surface. One of the specimens
is partly overgrown by a thin crust of a monaxonid sponge.

The colour, in spirit, is brownish white.

The superficial part of the body is differentiated to form a cortex (Plate 45,
fig. 16). This is composed of three layers: a dermal layer (Plate 45, fig. 39a;
Plate 46, fig. 20b) 40-160 » thick; a middie sterraster-armour layer (Plate 46,
fig. 39b; Plate 46, fig. 20c) 150-350 y» thick; and a fibrous inner layer. This
inner layer is very thin and quite inconspicuous. In many places the chamber-
bearing choanosome extends right up to the sterraster-armour; in these the
inner layer cannot be made out at all.

Canal-system. The dermal membrane is perforated by groups of pores.
These pore-groups occupy the pits mentioned above. The pores themselves
are oval, in some places 10-40 y, in others (Plate 46, fig. 21) 20-70 » wide. The
smaller appear to be the afferent, the wider the efferent pores.

The pores of each group lead into a system of wide, lacunose canals (Plate
45, fig. 39d) excavated in the dermal layer, which converge and join to form a
radial cortical canal. The radial cortical canals observed were all either strongly
contracted or quite closed.

The flagellate chambers (Plate 46, fig. 28b) appear oval in sections and have
a maximum diameter of 20-25 ». Numerous rather wide canals traverse the
choanosome (Plate 465, fig. 16).

GEODIA ACANTHTYLASTRA. 189

Skeleton. In the inner proximal part of the choanosome rather irregularly
disposed amphioxes, sterrasters, and acanthtylasters occur. In the distal part
of the choanosome and the inner layer of the cortex similar amphioxes and the
rhabdomes of plagio-, mesopro- and ana-triaenes form loose radial bundles.
The cladomes of the plagiotriaenes and of many of the anatriaenes lie just below
the sterraster-armour. Most of the mesoprotriaenes observed penetrate the
cortex, their cladomes protruding freely beyond the surface (Plate 46, fig. 20d).
Minute dermal rhabds, mostly disposed radially, small strongylosphaerasters,
and large oxysphaerasters are also found in this region besides acanthtylasters
and sterrasters similar to those of the interior. The sterrasters are not nearly
so abundant here as in the proximal part of the choanosome (Plate 46, fig. 20).
The sterraster-armour layer is occupied by sterrasters not very closely packed,
a few small strongylosphaerasters, radial minute dermal rhabds, and the parts
of the rhabdomes of the mesoprotriaenes which penetrate the cortex. In the
walls of the radial cortical canals, which traverse this layer, large oxysphaeras-
ters are met with. The dermal layer is occupied by groups of more or less
radially disposed, minute dermal rhabds, which are nearly all amphiox. The
proximal ends of these spicules are deeply implanted in the sterraster-armour;
their distal ends protrude freely beyond the surface (Plate 45, fig. 39c; Plate 46,
fig. 20a). The superficial part of the dermal layer, that is to say, the dermal
membrane, is occupied by a thin but dense layer of small strongylosphaerasters.
Similar spicules are also found in small numbers in the walls of the dermal canals.

The spicules to which the following descriptions refer were taken from the
largest specimen.

The choanosomal amphioxes (Plate 46, figs. 17-19a) are slightly curved in a
simple or irregular manner, 0.7-2.2 mm. long, and 14-40 y thick. A smaller
and a larger kind of amphiox, which are connected by relatively few transi-
tions can be distinguished. The smaller ones measure 0.7-1.4 by 14-25 y, the
larger 1.5-2.2 mm. by 23-40 yp.

The minute dermal rhabds (Plate 45, figs. 20-22, 39c) are nearly all amphi-
oxes. Styles have also been observed among them, but they are exceedingly
rare. The minute dermal amphioxes are 150-300 » long and 3-15 y, usually
8-13 thick. The thickness is usually proportional to the length. Those over
265 «long are 12-15 y thick, while those less than 200 long are less than 7.5 4
thick. Among those intermediate in length both thick and thin ones were
observed. Styles, although, as mentioned, very rare, were observed both
among the thick and the thin rhabds. They are about 250 long and 4-10
thick.

190 GEODIA ACANTHTYLASTRA.

In the centrifugal spicule-preparations I found some very minute amphioxes
41-53 long and 1-1.2 » thick. These may belong to the sponge, but it is more
probable that they are foreign.

The plagiotriaenes (Plate 45, figs. 18b, 19b, 30-35, 38) have a fairly straight
rhabdome. Its cladomal half is usually nearly cylindrical, its acladomal half
conic and pointed (Plate 46, figs. 18b, 19b, 31-33). Rarely the rhabdome is
nearly cylindrical throughout and rounded at the acladomal end (Plate 46, fig.
38). The rhabdome is 1.2-2.5 mm., in fully developed plagiotriaenes usually
about 2.1 mm. long and, at the cladome, 40-77 » thick. The cladome is usually
simple and composed of three fairly equal stout, conic, and blunt-pointed clades,
curved, concave to the rhabdome quite strongly in their proximal part. The dis-
tal part of the clades is slightly curved in the same direction, or straight (Plate

y45, figs. 18b, 19b, 30-82, 34, 35, 38). The clades of these regular plagiotriaenes
are 160-260 y long, their chords enclosing angles of 100-116°, on an average
106°, with the axis of the rhabdome. Besides these spicules, which form the
great majority, similar ones with irregular cladomes have been observed in
small numbers. In these spicules either a fourth clade, situated some distance
below the cladome, is added to the three ordinary terminal ones, or one (Plate
45, fig. 33) or more clades are bifurecate or provided with a short, irregular,
upward-directed branch. In the spicule-preparations also some true dicho-
triaenes (Plate 45, figs. 36, 37), with shorter rhabdomes and smaller clade-angles
than those of the regular plagiotriaenes were seen. As, however, I failed to find
regular dichotriaenes in the sections they may very likely not belong to the sponge.

The mesoprotriaenes (Plate 46, figs. 1-7; Plate 46, fig. 20d) have a straight
or slightly curved rhabdome, 2.3-3.3 mm. long. At the cladome it is 13-22 p,
farther down, at its thickest point near the middle, 21-31 », usually not quite
half again as thick as at the cladomalend. This thickening towards the middle is
exemplified by the following measurements of five mesoprotriane-rhabdomes : —

THICKNESS OF THE RHABDOME.

At its thickest point

At the cladome iene tho anidale

ft i
13 21
14 22
17 20

GEODIA ACANTHTYLASTRA. 191

The clades are conic, pointed or blunt, and slightly curved, concave to the
epirhabd. They are fairly equal (Plate 45, figs. 1, 3) or considerably unequal
(Plate 45, figs. 2,6) inlength. The longest clade of the cladome is 55-130 p long.
The chords of the clades enclose angles of 31-53°, on an average 39°, with the
axis of the epirhabd. The epirhabd is straight, conic, and 30-85 » long. Its
length usually is from a third to a half of the length of the longest clade.

The anatriaenes (Plate 45, figs. 8-15, 19c) have a rhabdome 3-5.4 mm., mostly
about 4.7 mm. long and, at the cladome, 18-28 » thick. The cladome is without
an apical protuberance. The clades are stout, fairly equal, conic, considerably
curved, concave to the rhabdome, in their proximal part, and slightly curved in
the same direction or straight in their distal part. Their chords are 50-110 yu
long and enclose angles of 38-56°, on an average 45.3°, with the axis of the rhabd-
ome. In the small, probably young, anatriaenes, in which the rhabdome is,
at the cladome, only 5-13 » thick and the clades only 30-46 » long, the clade-
angles are much larger, 52-58° wide (Plate 45, fig. 9).

Besides the microscleres mentioned above which have been observed in situ
in the sections, oxyasters and a large strongylosphaeraster were found in the
centrifugal spicule-preparations. These, particularly the large strongylosphaer-
aster, may be foreign to the sponge, but since this is doubtful I think it better
to describe them here, together with the asters, which undoubtedly form part of
the skeleton proper.

The oxyasters are of two kinds, smaller and larger ones. The smaller oxy-
asters, which are, in the centrifugal spicule-preparations, relatively much more
numerous than the larger ones, are without centrum and have from five to ten
straight, conic rays, 12-15 » long and, at the base, 0.7-3 yw thick. The distal
two thirds of the rays are covered with spines. The total diameter of these asters
is 22-29 ». The larger oxyasters are similar to the small ones, have six or
seven rays, about 20 » long and 3 » thick at the base, and measure 36-38 in
total diameter.

The large oxysphaerasters (Plate 46, figs. 10-13) are connected with the
small strongylosphaerasters by transitional forms. Some might also be consid-
ered as transitions between the oxysphaerasters and the oxyasters described
above. Before dealing with these transitional forms I will describe the true
oxysphaerasters (Plate 46, figs. 10-13). These have a spherical centrum, 3.5-
5 w, from a quarter to a third of the whole aster, in diameter. From this from
twelve to twenty-six, rather regularly distributed concentric rays arise radially.
The rays are straight, conic, sharp pointed, without centrum, 3.7—7 » long, and,

192 GEODIA ACANTHTYLASTRA.

at the base, 1-2.2 » thick. The distal parts of the rays bear a few rather large,
vertically arising spines, which usually form, some distance below the end, a more
or less pronounced verticil. Besides the spines forming this verticil, others,
situated more proximally, are not infrequently observed. The total diameter
of the oxysphaerasters is 12-16.5 ». A correlation between size and ray-
number could not be detected.

The oxysphaerasters which I consider as transitions to the small strongylo-
sphaerasters are similar to the true ones described above, have from ten to twenty-
five rays, and measure 6-9.5 in total diameter. Their centrum is 2-3 y, about
a third of the whole aster, in diameter. The conic, spined, and pointed rays are,
without the centrum, 1.5-3 long and, at the base, 0.6-1.3 y thick.

The oxysphaerasters which might be considered as transitions to the oxy-
asters are also similar to the true ones, have from sixteen to twenty rays, measure
about 28 in total diameter and have a centrum 5-6 yp, from a sixth to a fifth of
the whole aster, in diameter.

The acanthtylasters which are the chief choanosomal microscleres (Plate 45,
figs. 23-27a; Plate 47, figs. 1, 2, 3b, 4-6, 7b, 8) have from four to twelve
regularly distributed rays and usually a central thickening 1.7—3 y in diameter.
The rays are straight, cylindroconical, attenuated distally, and, at the base, 0.5-1.3
pethick. Rays over 1 « thick are found only in the larger, few-rayed acanthtyl-
asters. The rays are truncate; often a small and slender terminal spine arises
from their terminal face. The rays bear a terminal verticil of protuberances,
which together form a conspicuous tyle 0.9-1.6 in transverse diameter. The
individual protuberances forming this tyle, are so minute that it is difficult to
make out their shape. They are always more or less recurved in a claw-shaped
manner. Sometimes they appear as rounded knobs, sometimes as stout-pointed
spines and sometimes as more slender and longer branches, strongly curved
backward like the clades of anatriaenes. A few protuberances (spines) similar
to those forming the terminal verticil (acanthtyl) are often observed farther down
on the rays. The total diameter of the acanthtylasters is 11-22 ». It is in
inverse proportion to the number of the rays. Four- to five-rayed acanthtyl-
asters are 15-22 p, six- to seven-rayed 13.5-20.5 m, eight- to nine-rayed 11.5-
18 , and ten- to twelve-rayed 10.5-15 in diameter.

The small strongylosphaerasters (Plate 45, figs. 24b, 25b, 27b; Plate 46,
figs. 1-7; Plate 47, figs. 3a, 7a) have a spherical centrum 1.5-2.7 yp, rarely as
much as 3.4 , from about one third to two fifths, seldom as much as one half
of the whole aster, in diameter. From this from fourteen to twenty-two quite

GEODIA ACANTHTYLASTRA. 193

regularly distributed rays arise radially. The rays are cylindrical or, more
rarely, conical, truncate, without the centrum 1-2 » long, and 0.3-0.6, rarely up to
0.8 « thick. Their distal part is covered with minute spines. These may either
all be so small as merely to render the ray rough in appearance, or some of them,
which usually form a verticil at or just below the end of the ray, may be larger
and clearly distinguishable as spines. Sometimes the ray appears terminally
thickened. Such asters have a somewhat acanthtyl appearance and may, if
large, be considered as transitional to the acanthtylasters described above.
The total diameter of the small strongylosphaerasters is 4.3-6.1 , usually 5-6 pu.

The large strongylosphaeraster found in a centrifugal spicule-preparation,
which is probably a foreign spicule, has seventeen conic, truncate rays, at the
base 5 y: thick. The centrum is 13 y, the whole aster 23 in diameter. The con-
vex terminal faces of the rays are densely covered with small spines. All the
other parts of the aster are smooth.

The sterrasters (Plate 45, figs. 28a, 39b; Plate 46, figs. 8, 9, 14-19) are
flattened ellipsoids, 65-76 » long, 55-68 broad, and 42-46 thick. The average
proportion of length to breadth to thickness is 100 : 85 : 64.

In young sterrasters of a certain developmental stage, in which the rays are
still terminally rounded, from two to five small rudiments of spines, standing
close together, arise from the summit of each ray. In the adult sterrasters the
protruding rays away from the umbilicus are 2.5-3 y thick and bear terminal
verticils of usually five or six lateral spines, which are about 1.5 ” long and
remarkably stout. The transverse sections of the rays surrounding the umbilicus
are only slightly elongated in a direction radial to the umbilicus, and measure
about 3 by 4.5 . These rays usually bear from six to eight spines. The spines
directed towards the umbilicus are considerably larger than the others.

The surface of the umbilical pit is uneven and often covered with numerous
conspicuous protuberances (Plate 46, fig. 19).

The five specimens of this species were caught with the tangles at Station
2829 on May 1, 1888, off Lower California, in 22° 52’ N., 109° 55’ W.; depth 56
m. (31 f.); they grew on a rocky bottom; the bottom temperature was 23.4°
(74.1° F.). They were labeled F. C. 1342.

The cribriporal nature of the afferents and efferents and the character of the
skeleton show that these sponges belong to Geodia. The spiculation differs
from all the other species of this genus as well as from those of Sidonops to
such an extent that a new species had to be established for it.

194 GEODIA MEDIA.

Geodia media BowreRBANK.

Plate 16, figs. 1-21; Plate 17, figs. 1-22.
Proc. Zool. soc. London, 1873, p. 13, pl. 2, figs. 24-29.
Synops (2) media Souuas, Rept. voy. “Challenger,” 1888, 25, p. 266.
Sidonops media (Bowerbank) LENDENFELD, Tierreich, 1903, 19, p. 103.

There are in the ‘Albatross’ collection eight specimens in spirit from
Panama, which, as a comparison with a part of the type of Geodia media Bower-
bank in the British Museum, kindly placed at my disposal for examination by
Mr. Kirkpatrick, shows, belong to this species.

One of the eight ‘Albatross’? specimens is digitate, the others are, like
Bowerbank’s type, irregularly massive. The former differs slightly in regard
to the dimensions of its spicules and the character of its canal-system from the
latter. As will be seen from the description given below, these differences are
not sufficient, however, for the establishment of subspecies or varieties I con-
sider them as two different forms of one species.

Shape and size. Of the seven specimens of the massive form one, the
largest, is fairly complete, the other six are more or less fragmentary. The
complete massive specimen (Plate 16, fig. 16) is an irregular mass 85 mm. long,
41 mm. broad, and 29 mm. high. It is attached at several points. Between
these points of attachment the lower surface is considerably raised, and thus forms
the roof of rather high tunnels which undermine the sponge. On the upper
side flat-bottomed depressions, irregular in outline and 9-21 mm. in diameter,
are observed. These are surrounded by conspicuous elevated borders, which
here and there rise to form higher, rounded protuberances. The surfaces of the
depressions are occupied by small, shallow pits, the centres of which are about
1 mm. apart. These pits are separated from each other by minute ridges which
form a network. The convex parts of the upper side and the unattached parts
of the lower side are smooth. Here and there a few spicules protrude beyond
the surface, but there is nowhere a trace of a true spicule-fur. Larger open-
ings, oscula or openings of uniporal cortical canals, are absent. The smaller,
fragmentary massive specimens are quite similar to the large one. They
measure 26-38 mm. in maximum diameter. One has depressed pit-bearing
areas, up to 35 mm. long.

The single specimen of the digitate form (Plate 16, fig. 17) is a curved irregu-
lar cylinder, 65 mm. long, 6-12 mm. thick, and attached at several points
along one side. Its transverse section is throughout more or less circular. The
surface is smooth, slightly undulating, and destitute of a spicule-fur. There

GEODIA MEDIA. 195

are no clearly circumscribed pit-bearing areas as in the massive form, but pits
similar to those of the latter, are found singly and in groups, scattered irregu-
larly over the concave parts of the surface. Oscules or uniporal openings of
cortical canals are absent.

The colour of the massive form is, in spirit, brownish white. Bowerbank’s
dry specimen was pale buff-yellow. The surface of the elevations and the lower
side are lighter in colour than the depressed pit-bearing areas and the interior.
The digitate form is dirty white, a little darker in the interior.

Various space-symbionts, some specimens of Donatia, a desmacidonid
sponge with exotyle spicules, Serpulae, and small composite ascidians, are at-
tached to the specimens of the massive form.

The superficial part of the body forms a cortex composed of three layers: an
outer dermal layer (Plate 17, fig. 21a), a middle sterraster-armour layer (Plate
17, figs. 21c, 22c), and an inner fibrous layer (Plate 17, fig. 22e). The dermal
layer is free from sterrasters. In the pits above described it attains a consider-
able thickness and here it is excavated by extensive dermal lacunose canals.
Everywhere else it is but a thin membrane. The sterraster-armour layer,
which forms the largest part of the whole cortex, is filled with sterrasters. The
inner fibrous layer is thin and consists chiefly of paratangential fibres. It con-
tains only a few scattered sterrasters. The thickness of the cortex is very
different under different parts of the surface. In the pit-bearing depressions of
the massive form it is 0.5-1 mm., on the convex parts of the upper side about 1.5
mm. thick. On the lower side it is much thinner. It attains the greatest thick-
ness in the elevated borders surrounding the pit-bearing depressions, and is in
some parts of these 2.3 mm. thick, a great part of the whole border consisting
of cortical tissue. In the digitate form the cortex is 0.5 mm. (under the con-
cave parts of the surface), 1.2 mm. (under the convex parts of the surface)
thick.

Canal-system. The pits congregated in the depressed areas of the surface
of the massive form and scattered over the concave parts of the surface of the
digitate form are covered by pore-sieves, which appear as nets of slender strands
with broad oval meshes (pores). In many places these pore-sieves (nets) consist
of a primary network of stouter, pigmented strands, in the meshes of which a
secondary network of more slender, mostly unpigmented strands is spread out.
The meshes of the primary network are 80-100 y, those of the secondary network,
the pores proper, 15-30 » wide. Where primary and secondary nets cannot be
distinguished the pores are also 15-30 y« wide. In the depressed areas of the

196 GEODIA MEDIA.

massive form the pits are so close together and the pore-sieves covering them so
extensive that the latter often come in direct contact with each other, so that
these depressions appear covered throughout by a nearly continuous pore-sieve.
In the digitate form such a junction of pore-sieves has not been observed. The
pores of these sieves lead into cavities up to 100 » wide, excavated in the dermal
tissue occupying the pits. The cavities of each pit join to form a radial cortical
canal about 250 « wide, which penetrates the sterraster-armour and opens out
below into a subcortical cavity. These cavities are mostly 300-500 y in radial
diameter. The proximal opening of each cortical canal is restricted by a ring-
shaped chonai sphincter, composed of an annular strand of contractile tissue
about 90 » thick, which protrudes into the subcortical cavity. In the sections
studied nearly all these chonal sphincters were more or less dilated, the lumen of
some being as much as 140 » wide. The cortical canals leading down from the
scattered pits of the digitate form are similar. As one radial cortical canal
belongs to each pit, these canals are in the depressions of the massive form about
1 mm., in the digitate form farther apart.

The elevated borders surrounding the depressions of the massive form, and
parts of the convex portions of the surface of the digitate form appear to be
entirely destitute of pores. On other convex portions of the surface both of the
massive and the digitate form, pores, arranged in groups and forming pore-
sieves, are met with. These pore-sieves are much less extensive than the ones
covering the pits, never composed of a primary and secondary network, and
pierced by much larger holes. The latter are circular or oval and measure 30-
70 «in diameter. The pores (holes) of each group lead into cavities, joining to
form a canal 200-300 » wide, which traverses the cortex radially and is constricted
below by a chonal sphincter. In some places these radial cortical canals are
quite close together, their centres being only 1 mm. apart, in other places they
are much farther apart.

In the basal part of the massive form very large canals, some as much as
6 mm. wide, are observed Just below the cortex. In the digitate form I have not
met with any wide canals of this kind. In the choanosomal canals leading up
to the pit-bearing parts of the surface transverse sphincter-membranes are spread
out at frequent intervals.

The chonal sphincters of the cortical canals leading down from the pits
being wide open, the choanosomal canals leading up to them being traversed by
sphincter-membranes, and the pits themselves being situated on concave (de-
pressed) parts of the surface, one would suppose that the pores in the sieves

GEODIA MEDIA. 197

covering the pits are the efferents. The fact that the pores in these sieves are
smaller than those on the smooth convex parts of the surface is on the other
hand in favour of the view that they are the afferents.

The skeleton of the interior consists chiefly of amphioxes and large asters
(oxyasters and a few strongylasters). Besides these spicules also styles, various
angularly bent or branched amphiox- and style-derivates, and sterrasters occur
in small numbers. The amphioxes, styles, amphiox- and style-derivates are
arranged rather irregularly in the interior, but assume a regular radial arrange-
ment towards the surface. Here, just below the cortex, plagiotriaenes, meso-
clades (mesomonaenes), and a few slender, amphiox-like mesoclade-derivates are
added to these megascleres. The clades of the plagiotriaenes lie at the limit
between the middle and inner layers of the cortex, the sterraster-armour resting,
as it were, on the plagiotriaene-cladomes. Just below the cortex small oxysphaer-
asters, which are very numerous in some places, to a great extent replace the
large asters of the interior, and here also scattered sterrasters and small dermal
styles occur. The latter are situated radially, obliquely, or paratangentially (Plate
17, fig. 22d). The whole of the cortex, with the exception of the dermal layer,
contains sterrasters. These are scattered in small numbers in the thin, inner-
most, fibrous layer, and form a dense mass in the thick middle (sterraster-armour)
layer. Between the sterrasters a few small sphaerasters, dermal styles, and pro-
truding spicules occur. Small sphaerasters form a dense and continuous layer
on the outer surface of the dermal layer. Over the pits, where the latter is
greatly thickened (Plate 17, fig. 21), these sphaerasters are not confined to the
surface but are very numerous also for some distance below it. Tufts of small
dermal styles are met with on all parts of the surface. These styles are more
numerous in the thickened parts of the dermal membrane in the pits (Plate 17,
fig. 21b) than elsewhere. They are situated radially or obliquely; their rounded
ends lie proximally and are deeply imbedded in the sponge, their pointed ends
abut on the surface or protrude more or less beyond it (Plate 17, fig. 21). The
fact that these styles are abundant on the surface and rather plentiful also in
the subcortical layer, but comparatively rare in the sterraster-armour layer,
indicates that they move up from the distal parts of the choanosome, where
they are formed, to the surface, at first, till they reach the cortex, slowly, and
then, when they reach the sterraster-armour, rapidly, traversing the latter at
such a rate that at any time only a few are found in the act of passing
through it.

Besides the spicules above described several other forms, not met with in

198 GEODIA MEDIA.

situ in the sections, have been observed in the spicule-preparations. One of
these, an anatriaene, is, in all probability, proper to the sponge.

Among the large choanosomal amphioxes, two forms, a slender and a stout
one, can be distinguished. These are, it is true, connected by transitions, but as
the latter are not at all numerous and as the slender amphioxes are longer than
most of the stout ones, and therefore cannot be considered as young forms of the
latter, I am inclined to consider the slender and the stout amphioxes as distinct
forms.

The slender, chounosomal amphioxes (Plate 16, figs. 3-5a) are quite frequent
in the digitate form but rare in the massive form. They are slightly curved,
fairly isoactine, not very sharp pointed, 1.3-1.5 mm. long, and 23-80 p thick.

The stout choanosomal amphioxes (Plate 16, figs. 1-5b, 6z, 8b, z, 11z) are
very numerous in both forms. They are slightly curved, fairly isoactine, and
gradually attenuated to the not very sharp-pointed ends. In the digitate form
they are 1.2-1.55 mm. long and 33-51 y thick, in the massive form 1.3-1.7 mm.
long and 35-51 yp thick.

In Bowerbank’s type, reexamined by me, I found choanosomal amphioxes
1-1.5 mm. by 18-50 ». Sollas gives their dimensions as 1.51 mm. by 32 p.

The large choanosomal styles (Plate 16, figs. 11f, 13f) are not numerous.
They are relatively more frequent in the massive than in the digitate form.
These styles are 0.9-1.3 mm. long and 30-50 thick. Some are for the greater
part of their length cylindrical, and simply rounded off at one end, others some-
what attenuated towards, others again slightly thickened at, the rounded end.
In four styles of the massive form the thickness was: —

In the centre 36 4, at the rounded end 24 p

“ “ “ 40 “« « “cc “ “ 43 *
“ “ “ 43. 6 “ “ce “c 50 “«
“ “c “ 50 “ec “<c “ec 40 “

These spicules are not mentioned by Sollas; I found a good many, however,
in Bowerbank’s type.

Both the digitate and the massive forms are exceedingly rich in irregular,
angularly bent or branched derivates of the amphiozes and styles described above.
In their dimensions these spicules agree with the regular forms from which they
are derived. The simplest form of amphiox-derivate is an amphiox very slightly
angularly bent near one end. This bend may be in the same direction as the
curvature of the spicule (Plate 16, fig. Ly), or in a direction opposite to it (Plate
16, fig. 9y). In other amphiox-derivates of this kind the angular bend is much
more pronounced, the angle between its two limbs being smaller. Such a
spicule is represented on Plate 16, fig. 9c.

GEODIA MEDIA. 199

Derivates of styles, angularly bent in such a manner, are also met with
(Plate 16, fig. 13g). Occasionally amphioxes angularly bent at two places
have been observed (Plate 16, fig. 9w). In the simplest forms of the branched
amphiox- and style-derivates a small straight branch is observed arising near one
end of the spicule. This branch is generally directed upwards and encloses,
with the epirhabd-like continuation of the shaft of the spicule, a smaller (Plate 16,
figs. 10d, 13h) or a larger (Plate 16, fig. 10x) angle. Rarely it is directed down-
wards (backwards) (Plate 16, fig. 11h). Generally there is only one branch,
but spicules with a cluster of two or more branches have also been observed
(Plate 16, fig. 131). Some of these derivates are both branched and angularly
bent. Such forms have been found among the style-derivates (Plate 16, fig. 13k)
as well as among the amphiox-derivates (Plate 16, fig. 2}). Occasionally I have
observed spicules thickened and rounded at both ends which can be considered
as amphityle derivates of amphioxes or styles. These spicules have not been
mentioned by Bowerbank and Sollas, and in the former’s type they are indeed
very scarce.

The small dermal styles (Plate 17, figs. 21b, 22d) are straight or slightly
curved and taper towards the distal pointed, and also towards the proximal
rounded end. The transverse diameter of the rounded end is a little more than
half the transverse diameter of the thickest central part, the average ratio be-
tween these two dimensions being about 10:6. In the digitate form these
styles are 140-190. », usually 180-190 » long and 2-5 yp, usually about 4 y thick.
In the massive form they are slightly larger, 150-265 y long and 3-6 y thick.
These spicules are not mentioned either by Bowerbank or by Sollas, but they
are very abundant in the former’s type, and there measure 150-200 by 2-4

Among the plagiotriaenes, as among the amphioxes, two forms, a slender and
a stout one, can be distinguished. As the intermediate forms connecting these
are not numerous and as the slender plagiotriaenes have longer clades than
most of the stout ones, showing that the former cannot be young forms of the
latter, I consider them as two distinct kinds of spicules. Both these types of
plagiotriaenes have been found also in Bowerbank’s type.

The slender plagiotriaenes (Plate 16, figs. 1q, 5r, 12q, 14q), are met with
in both forms but are not very numerous in either. Their rhabdome is straight
or slightly curved, conic, pointed, 1-1.6 mm. long and 17-30 y thick at the
cladomal end. The clades are generally considerably curved, concave to the
rhabdome, particularly in their distal part, and 160-260 y long. The chords
of the clades enclose angles of 107-114° with the axis of the rhabdome. Similar

slender triaenes were also found by me in Bowerbank’s type.

200 GEODIA MEDIA.

The stout plagiotriaenes (Plate 16, figs. 16s, 11s, 12s) occur in large num-
bers in both forms. Their rhabdome is straight or very slightly curved, and
usually pointed, rarely rounded at the acladomal end. It is either regularly
conical throughout or slightly constricted just below the cladome. The rhabd-
omes of five of the plagiotriaenes of the massive form thus constricted, which —
I measured, were :—

60 » thick at the cladomal end and 62 » thick at the thickest point a little farther down.
60 ae “ “ “ce ae oe ae 64 ae oe oe ae oe oe ‘ “ee ia oe
60 “ Gb wm (id i gl a “ ee tip “ “

“ “ 68 “ “ “ “i “ ci “i “ “ ‘ce

“ “ “ nL, 7) a ee ce WG “ cree et may “ “
70

65“
67“

In observing a thin transverse splinter of a plagiotriaene-rhabdome with a
high power I found its axial rod to be triangular in transverse section.

In the digitate form the rhabdomes of the stout plagiotriaenes are 0.8-1.4
mm. long and 30-65 , usually 50-60 » thick at the cladomal end. The clades
are only slightly curved, concave to the rhabdome, conical, not sharp pointed,
and 110-260 long. Their chords enclose angles of 107—118° with the axis of the
rhabdome. The stout plagiotriaenes of the massive form are similar in shape
but considerably larger. Their rhabdomes are 0.9-1.7 mm. long and at the cla-
domal end 30-80 , usually 50-70 » thick. Their clades are 160-310 long and
the chords of their clades enclose angles of 105-120° with the axis of the rhabd-
ome. The length of the rhabdome and the clades is not in proportion to their
thickness, the very thick ones having by no means particularly long rhabdomes
or clades.

In Bowerbank’s type, examined by me, the dimensions of the stout plagio-
triaenes are: rhabdome 0.81.2 mm. by 40-60 y, clades 200-260 p long, clade-
angles 105-120°. Sollas gives their measurements as follows: rhabdome 1.12
mm. by 450 (printers error for 45) y, clades 254 long, clade-angles (according
to the relation of the clade-length to the transverse diameter of the cladome,
320 p, given by him) large.

Besides the regular plagiotriaenes described above a few irregular derivates —
of them occur. In some of these the clades are reduced in length and rounded
at the end; in others two of the clades are single and one bifureate (Plate 16,
fig. 12t); in others again the clades are curved convex to the rhabdome (Plate
16, fig. 5u). I have also observed a few mesoclade plagiotriaene-derivates in
which the rhabdome is much shortened, rounded at the end, and continued be-
yond the cladome in the shape of a long conical epirhabd (Plate 16, fig. 1v).

The mesomonaenes (Plate 16, fig. 7m, n, 0, p; Plate 17, figs. 8-12), which

GEODIA MEDIA. 201

occur in both forms but are numerous in neither, have more or less curved rhabd-
omes 1.7—2.5 mm. long and 4-10 y thick at the cladomal end. The epirhabd
is 40-75 y long, while the clade attains a length of 17-50 ». The latter encloses
an angle of 32-89° with the epirhabd. Those in which this angle is small appear
as mesopromonaenes, those in which it is large as mesorthomonaenes, while the
intermediate forms are mesoplagiomonaenes. The clades are generally rounded
off terminally, more rarely pointed. The axial thread of the rhabdome and
epirhabd bears at the point of origin of the clade, besides the axial thread of the
clade itself, a number of small, rudimentary branchlets, resembling the struc-
tures observed by me in the anaclades of Thenea valdiviae.' The axial thread
of the clade is rounded off at the end and often shows indications of being split
up terminally into small branches. Neither Bowerbank nor Sollas mentions
these spicules. I found several, however, in the former’s type similar in every
respect, to those described above.

I have found a few long and slender amphioxes of dimensions similar to
those of the rhabdomes (and epirhabds) of the mesomonaenes. In some of
these spicules (Plate 17, fig. 7) a slight thickening near one end, enclosing a few
rudimentary branchlets of the axial thread, indicates clearly that they are
mesoclade-derivates, while in a few others hardly any trace of such a thickening,
or no thickening at all, can be detected. Although these spicules appear as
true amphioxes I am inclined to consider them as mesoclade-derivates. Such
spicules are not mentioned either by Bowerbank or by Sollas. I found several,
however, in the former’s type.

The scarce anatriaenes have a long rhabdome, at the cladome about 12 y
thick, and clades about 23 » long and curved concave towards the rhabdome
throughout their whole length. The clade-angle is about 44°. Bowerbank
says that there are “‘very slight indications of the presence of recurvo-spicula,”
Sollas, however, does not mention anatriaenes. In the spicule-preparations of
Bowerbank’s type, I observed a good many anatriaenes with rhabdomes
10-15 p thick, clades 25-35 p long, and clade-angles of 50-52°. I found in it
also a mesanatriaene with similar clades, an epirhabd 165 long, and a short-
ened, terminally thickened, club-shaped rhabdome, 0.8 mm. long and, at the
thickened, acladomal end, 27 y thick.

The large oxyasters and strongylasters (Plate 17, figs. 4-6b, 15b, 18b, 19b, 20)
are destitute of a central thickening and have from four to eleven, most fre-

‘R. v. Lendenfeld. Die Tetraxonia. Wissensch. ergebn. deutschen Tiefsee-Expedition, 1898-1899,
1907, 11, p. 200 ff.

202 GEODIA MEDIA.

quently six rays. The rays are straight, and in the many-rayed forms concentric
and fairly regularly distributed. In the few-rayed forms the rays are frequently
not uniformly distributed and sometimes not quite concentric; many of these
spicules have a somewhat metastrose character accordingly. The rays are conic
and fairly pointed, or cylindroconical and blunt, or cylindrical and truncate.
The asters with rays of the latter kind which appear as true strongylasters, are
however rare. Quite at the base the rays are, for a short distance, smooth.
Apart from this they are wholly covered with spines standing rather close
together. In size the spines are subject to considerable variation, but they are
never very minute. In some asters with particularly large spines I noticed
that the spines are directed backwards, towards the centre of the spicule. The
blunt, cylindroconie rays usually bear a terminal spine. In the digitate form
these asters are 21-41 y, in the massive form 21-45 y, in total diameter. The
basal thickness of the rays is in the former 1-2.5 y, in the latter 1-3.5 w. An
inverse proportion between the size and ray-number is clearly expressed in
regard to the thickness of the rays but not so well marked in regard to the
total diameter of the spicule. In Bowerbank’s type, these asters have from
four to seven rays, 1.2-2.3 4 thick, and measure 20-37 » in total diameter.
Sollas gives their diameter as 26 p.

The subcortical oxysphaerasters have a spherical centrum 3-4.5 in diameter,
from which from fifteen to twenty rays arise radially. These are regularly dis-
tributed, 4-9 y long, straight, conic, sharp pointed, and 0.8-1.7 » thick at the
base. <A verticil of large vertically arising spines is observed some distance
below the end of each ray. Besides these large verticil-spines smaller ones,
distributed irregularly over the distal part of the ray, are often present. The
total diameter of these oxysphaerasters is 9-18 4. I have found similar spicules
in Bowerbank’s type, and he appears to have observed them, but they are not
mentioned by Sollas. In the centrifugal spicule-preparations F have met with a
few spineless oxysphaerasters with only six rays, about the size of the largest
spined oxysphaerasters. It seems to me doubtful whether these spineless oxy-
sphaerasters, which I have not observed in situ in the sections are proper to the
sponge.

The small dermal sphaerasters (Plate 17, figs. 1-3, 4-6a, 13, 14, 16, 17, 18a,
19a) have a spherical centrum 1.7-5 » in diameter, from which from seven to
eighteen rays arise radially. These rays are quite regularly distributed 1.3-
3.5 long and 0.8-1.6 thick at the base. The total diameter of these asters
is 4.6-8, usually 5.5-6.5 4. The shape of the rays and the relative size of the

GEODIA MEDIA 203

centrum are subject to great variations. In the dermal sphaerasters most
frequent in the digitate form the diameter of the centrum isalways more than
half and usually less than two thirds of the diameter of the whole aster. The
rays of these asters are cylindrical or cylindroconical, truncate or rounded off at
the end, and covered with spines, chiefly in their distal portions. The size of the
spines is proportional to the thickness of the rays. In the dermal sphaerasters
most frequent in the massive form the centrum is relatively a little smaller, its
diameter in not a few being less than one third of the total diameter of the spicule.
The rays of these asters are conical or cylindroconical, rounded off or, more
rarely, pointed at the end (Plate 17, figs. 1-3, 13, 14, 16, 17), and covered with
spines, chiefly in their distal parts. Of course both the oxyastrose and the
strongylastrose kinds of dermal sphaerasters occur in both forms. The dif-
ference lies in the fact that the strongylastrose forms are much more prevalent in
the digitate than in the massive form. In Bowerbank’s type, these spicules
have from six to eighteen rays, which seem, on the whole, somewhat stouter

‘“ Albatross”? specimens.

than those of the small strongylosphaerasters of the
The total diameter of the aster is 5.5-8 , that of the centrum 2-3.5 yw. Sollas
gives their diameter as 6 p.

Besides these a few other forms of asters, which may be considered as
sphaeraster-derivates, are found in the dermal membrane. One of these aster-
forms has a very large centrum and exceedingly short and thick rays, the latter
appearing as rounded knobs on the surface of the relatively overgrown centrum.
In another form the centrum is very much reduced and there are only five or
six cylindrical rays thickened at the end. The basal and central, cylindrical
parts of the rays of these asters are smooth, their terminal thickening covered
with numerousspines. Finally, there are sphaerasters the rays of which are not
concentric. These appear as metasters or ataxasters.

The normal sterrasters (Plate 16, figs. 15, 18, 19; Plate 17, figs. 21¢, 22c) are
flattened ellipsoids, broad oval, sometimes nearly circular, in outline. The
sterrasters of the digitate form are 84-98 long, 73-86 y broad, and 60-65
thick; those of the massive form larger, 89-105 y long, 83-90 y broad, and
about 62 » thick. The proportion of the length to the breadth to the thickness
is in both on an average about 10:9:7. With the exception of those surround-
ing the umbilicus, the rays protruding over the surface of the sterrasters are
2.5-3 thick and about 2 apart. They bear terminal verticils of usually from
four to six stout lateral spines, many of which are curved. The rays surround-
ing the umbilicus (Plate 16, fig. 15) have a somewhat oval transverse section

204 GEODIA MEDIA.

2.5-3 » broad and 4-5 long, the longer diameter being situated radially to the
umbilicus. These periumbilical rays bear from seven to ten lateral and usually
also some terminal spines. In Bowerbank’s type the sterrasters measure 100—
110 by 82-94 by 70-74 p. Sollas gives their dimensions as 110 by 80 p.

Abnormal sterrasters, sferroids, with fewer and more distant protruding rays
are met with quite frequently. In these spicules (Plate 16, figs. 20, 21) the
parts of the surface destitute of rays bear spines similar to those forming the
verticils on the rays of the normal sterrasters. Some of these spines are
seattered singly, others arranged in rosette-like groups. The protruding rays
of the sterroids attain a thickness of 12-15 » and a length of 10-12 » Their
sides are quite smooth; the rounded end is densely spined.

All the eight specimens of this species were collected on October 26, 1904, in
the Gulf of Panama, off Panama, on the shore of the islands of the Station
"Pacific Mail Steamship Company or of the Taboga Islands. Bowerbank states
that his specimen came from ‘‘Mexico.” Thinking that the Gwf of Mexico
must be meant, when the locality of a marine organism was given as Mexico,
in 1903 I gave the Gulf as its locality. Now that this species has been found
in the Gulf of Panama, it seems much more probable that Bowerbank’s type
‘ame from the Pacific coast of Mexico and not from the Gulf of that name.

Whichever of the two kinds of dermal apertures are the afferents and which
the efferents, there can be no doubt, that both are cribriporal. For this reason
I now place these sponges in Geodia and not in Sidonops as, following Sollas’s
suggestion, I formerly (1903) did.

GEODINELLA ROBUSTA. 205

GEODINELLA LENDENFELD.

Without regular triaenes. The tetraxon megascleres are monaene or diaene
teloclades with reduced clades, and occur not only in the superficial part of the
sponge but also in the interior.

There are in the ‘‘Albatross” collection four specimens of Geodinella.
These belong to one new species, which is divided into three varieties.

Geodinella robusta, sp. nov.

megasterra, var. nov.
Plate 1, figs. 1-4, 16, 18-24; Plate 2, figs. 1, 3; Plate 3, figs. 3, 4, 7, 9; Plate 4, figs. 1, 4-7, 13, 21, 22.
carolae, var. nov.
Plate 1, figs. 5-12, 17; Plate 2, figs. 4-7, 9-11; Plate 3, fig. 1; Plate 4, figs. 2, 3, 8-12, 14-20.

megaclada, var. nov.
Plate 1, figs. 13-15; Plate 2, figs. 2,8; Plate 3, figs. 2, 5, 6, 8; Plate 4, figs. 23-25.

I establish this species for four spirit specimens, three fairly complete, one
fragmentary, collected at Stations 2946, 4199, and 4228 on the Pacific slope of
North America. On account of their having larger spicules than the only other
known species of this genus I have named the species robusta. Although these
sponges are similar enough to be considered as representatives of the same
species, only two are really systematically identical, while the other two differ
to some extent from these two and also from each other. It therefore seems
advisable to consider these four sponges as three separate varieties. The most
conspicuous differences between them are found in the shape and size of their
sterrasters and the clades of their teloclades. In two specimens, one taken
in Charlotte Sound, Vancouver Island, the other in Naha Bay, Behm Canal, 8. E.
Alaska (Stations 4199, 4228), the teloclades have short clades and the sterr-
asters are smaller and nearly always regularly ellipsoidal. In the other two
specimens the sterrasters are larger and more diversified in shape, three-lobed
ones not infrequently occurring besides the ellipsoidal ones. The variety I
establish for the specimens from Stations 4199 and 4228 with short clades and
small sterrasters, I name var. carolae after the locality, Queen Charlotte Sound,
where one of them was found. In one of the two remaining specimens the
sterrasters are not so large and the clades of the teloclades considerably longer
then in the other; the variety with longer clades which was also found at
Station 4228 I name var. megaclada. The third variety, from Station 2946, the
sterrasters of which are larger than those of the other two, larger in fact than

those of any sponge in the collection, | name var. megasterra.

206 GEODINELLA ROBUSTA.

The specimens of var. carolae (Plate 1, figs. 12, 17) are both incrusting
cushion shaped and nearly of the same size, about 45 mm. long, 30 mm. broad,
and 15 mm. high. From the margin of the base of the specimen from Naha
Bay (Plate 1, fig. 12) several processes arise. One is large and digitate, 23 mm.
long and 7 mm. thick at the base; the others are small. The surface is slightly
uneven, covered with low rounded elevations. On the summit of some of the
more pronounced elevations a circular aperture, up to 700 in diameter, is ob-
served. This leads into a rather wide cavity, constricted below by a chonal
sphincter. The appearance of some of the elevations, not thus provided with an
apical vent indicates that these also possessed such an aperture which has,
however, been quite closed by contraction. The specimen of var. megasterra
(Plate 1, fig. 16) is somewhat irregularly finger shaped, slightly bent, and 43
mm. long. It is at the base 16 mm. thick; towards its upper end it tapers to a
diameter of 10mm. The surface has the same character as that of the specimens
of var. carolae, the only difference being that the apertures on the elevations at-
tain a larger size, the widest measuring as much as 1.5 mm. in diameter. The
specimen of var. megaclada is 13 mm. in diameter and appears to be a fragment
(the tip) of a finger-shaped sponge like the specimen of var. megasterra. ‘The
largest apertures on its surface are 300 wide. The specimens of var. mega-
sterra and of var. carolae from Naha Bay are partly covered by thin crusts of
a monaxonid sponge. In both these varieties examination of the surface with a
lens shows here and there indications of hirsute spicules, broken off short;
since, however, I have been unable to find in the sections any spicules penetrating
the cortex, it is doubtful whether these broken, apparently hirsute, spicules
are proper to the sponge.

The colour is white or white with a brownish tinge on the surface and dirty
white or brownish in the interior. The surface of the two specimens from Naha
Bay is partially covered with a darker brown coating which may be foreign to
the sponge.

A hard cortical layer, composed of dense masses of sterrasters, a sterraster-
armour (Plate 1, figs. 21a, 24a; Plate 2, fig. 6a), surrounds the choanosome.
This cortex is 1-2 mm. thick in var. megaclada on the whole somewhat thicker
than in the other two varieties.

Canal-system. The sterraster-armour (cortex) is perforated by round or
oval holes 300 #-1.5 mm. wide. Each perforation is traversed by a radial canal
situated centrally. This canal may be wide and extend nearly to the sterrasters
surrounding the perforation through which it passes, or it may be contracted,

GEODINELLA ROBUSTA. 207

- narrow, or altogether closed, and separated from the nearest sterrasters by a
thick intervening layer of soft tissue. The canals passing through the large
perfcrations of the sterraster-armour are usually dilated and open out on the
surface of the sponge with a wide aperture. These apertures are the vents,
which as stated above, crown the summits of some of the elevations. The
canals passing through the smaller perforations are usually contracted. On
account of the hardness of the cortex and also because I wished to preserve such
valuable material, I have not been able to ascertain exactly how these narrow
radial canals open out on the surface. So far as I could make out they divide
distally into several oblique nearly paratangential canals, which lead to small
groups of round pores, 50-80 wide, above the perforation of the armour through
which the radial canal passes. Below, these radial canals are provided with
chones composed of brownish tissue. These chones are about 200 » broad and
protrude to a distance of 400 4 into the choanosome.

The choanosome (Plate 1, figs. 21b, 24b; Plate 2, fig. 6b) is remarkably
soft and traversed by numerous wide canals up to 1.5 mm. in diameter. The
widest of these canals usually extend paratangentially just below the surface
(Plate 1, figs. 21, 24). From the canal-walls numerous low transverse ridges
protrude into the interior, partially dividing the canal-lumina into rows of
chambers. Such ridges occur not only in the wide, but also in the narrow canals.
The flagellate chambers (Plate 2, figs. 4b, 11a) are longitudinally compressed,
the length of the main axis, passing through the chamber-mouth, usually being
only from two thirds to three fourths of the length of the transverse diameter.
In the specimen of var. carolae from Charlotte Sound, which is better preserved
than the others, the flagellate chambers are 20-35 y long and 30-45 y broad.

In the characters of their canal-system these sponges appear on the whole
to approach rather closely the species of Sidonops.

Histological structure. Many of the flagellate chambers have lost their collar-
cells altogether, in others collar-cells are present, but they are never numerous
(Plate 2, figs. 4, 11). Whether this sparseness of the collar-cells is due to some
having dropped off post mortem, or whether it is a natural condition of the living
sponge, I cannot say. “‘Sollas’s membranes” about 9 y distant from the
chamber-walls are frequently observed. The nuclei of the collar-cells are about
2 in diameter. The dermal membrane is thin and composed of stout, spindle-
shaped elements, extending paratangentially. The proximal part of the corti-
cal tissue, just below the sterraster-armour, contains coarse paratangential
fibres and granular oval or spherical cells, 15-30 » in diameter. In the walls

208 GEODINELLA ROBUSTA.

of the canals stout bands of longitudinal, spindle-shaped, finely granular ele-
ments, 3-5 thick in the centre, are met with. Occasionally more or less iso-
lated, spindle-shaped elements, attaining a maximum transverse diameter of
10 , are seen in the sections. I have often noticed that the substance composing
these elements differs from the surrounding tissue, not only by being less stain-
able, but also by having a higher refractive index, so that it to a certain extent
resembles the spongin of the fibres of horny sponges. In sections of the choano-
some of var. carolae I have observed groups of highly stainable, oval cells 20—
30 . long and 12-17 » broad (Plate 2, fig. 7a). These cells are imbedded in
an alveolar tissue, composed of a mesh-work of elongated elements (Plate 2,
fig. 7b), and free from flagellate chambers. I am inclined to consider these large
cells as young ova.

Skeleton. The sterraster-armour and the microscleres generally are well
developed, the megascleres somewhat rudimentary; not the megascleres of
the choanosome, but the hard sterraster-armour of the cortex, forms the main
support of the sponge.

The megascleres are rhabds, simple teloclades, and a few multicladomal
and mesoclade derivates of the latter. The rhabds are for the most part blunt,
spindle-shaped, isoactine amphioxes. These pass on the one hand into pointed
amphioxes, on the other into cylindrical amphistrongyles or even amphityles.
Anisoactine blunt or pointed diactines and occasionally also styles and tylo-
styles are met with likewise. Some rhabds are rendered irregular by the presence
of small protuberances. Most of the teloclades are monaenes. Besides these
monaenes some diaenes occur in var. carolae. These are more numerous in the
specimen from Charlotte Sound than in the specimen from Naha Bay. In var.
magaclada and var. megaslerra monaenes only, the clades of which are much
longer in var. megaclada than in var. megasterra, are met with. Most of these
spicules are plagioclade, some orthoclade.

These teloclades (teloclade-derivates) and rhabds are met with in all parts
of the choanosome. In its central parts the teloclades are relatively less nu-
merous than in its superficial parts. In the axial part of the choanosome of the
finger-shaped var. megasterra these spicules form a rather loose longitudinal
column or strand (Plate 1, fig. 21), from which single spicules and small spicule-
bundles extend towards the surface. These are arranged in a somewhat plumose
manner, abut obliquely on the sterraster-armour, and extend distally to the
limit between the cortex and the choanosome, or a little beyond it, in which
case they penetrate the proximal part of the sterraster-armour and terminate
within it.

GEODINELLA ROBUSTA. 209

In the spicule-preparations I have seen a few fragments which appeared to
be parts of spicules much thinner and longer than the rhabds and the rhabdomes
of the ordinary teloclades. It has been stated above that on the surface indi-
cations of hirsute spicules, broken off short, are observed. It is possible that
these slender spicule-fragments are parts of those hypothetical hirsute spicules.
If hirsute spicules are really present they are probably teloclades with long
and slender cladomes. Since, however, I have not been able to find a single
cladome of such a spicule in all the numerous preparations made, I doubt
whether the sponge possesses any teloclades of this kind and rather incline to
the view that these fragments are parts of foreign spicules.

The microscleres are larger oxyasters passing into smaller oxysphaerasters,
small strongylosphaerasters, and sterrasters. The strongylosphaerasters are very
numerous on the outer surface and also occur in small numbers in the walls of
the cortical and choanosomal canals. The oxyasters (and oxysphaerasters)
are quite numerous in the choanosome, chiefly in the canal-walls; a few
also occur in the cortex. The sterrasters are densely packed in the sterraster-
armour and are, in radial sections, frequently also found scattered in the choano-
some. Whether they naturally occur there, or whether those observed there
have been brought to this position in cutting, I cannot say. Among the ordi-
nary sterrasters a few sterrasters with abnormal (hypertrophic) rays, sterroids
oceur. These appear to be irregularly scattered among the others.

The isoactine rhabds (Plate 1, figs. 22, 23; Plate 4, figs. 8,9, 11, 12, 16, 24)
are usually straight or slightly curved and spindle shaped, tapering towards
the ends; more rarely shortened and cylindrical. The ends of the spindle-
shaped ones are usually blunt, less frequently sharp pointed; the ends of the
cylindrical ones simply rounded off or slightly thickened. All these forms, the
spindle-shaped, sharp-pointed, and blunt amphioxes, and the shortened cylindri-
eal, simple amphistrongyles, and terminally thickened amphityles are connected
by transitional forms to such an extent that it appears advisable to describe
them together. These spicules attain a maximum length of 2.8 mm. Within
this limit their longitudinal dimension is variable, particularly so in the specimen
of var. carolae from Charlotte Sound, where a good many of the isoactine rhabds
are very much reduced in length. In this specimen the isoactine rhabds are
0.37-2.5 mm., usually 1.62.2 mm. long and 40-65 thick. As will be seen from
the measurements tabulated below the thickness is not in proportion to the
length, the shortest spicules being nearly, if not quite, as thick as the longest.

210 GEODINELLA ROBUSTA.

DIMENSIONS OF ISOACTINE RHABDS OF VAR. CAROLAE FROM CHARLOTTE
SOUND.

Thickness in

Length the middle

0.37 mm. 48 ps
0.43 =“ 47 ph
2 47
1 8r> ES 45 ps
2 sO Us 50 4
Dia ee 52 ft

There is, however, a correlation between the length and the shape of these
spicules, the long rhabds being spindle shaped, the short ones cylindrical and
usually thickened at the ends, amphityle in character. The terminal thickenings
of these amphityles are more or less spherical. One of the short amphityles is
represented on Plate 4, fig. 16. This spicule is 0.4 mm. long and 65 yp thick in
the middle. Its terminal thickenings measure 80 in diameter. In other
amphityles this terminal thickening is still more pronounced. In one such, 0.42
mm. long and 56 y thick, the terminal thickenings measured 73 yp. In these
shortened amphityle spicules the axial thread terminates in the centres of the
spherical terminal thickenings. I think that these short amphityles may have
been produced by some cause preventing the axial thread from attaining its full
length, but not proportionately reducing the vital energy of the silicoblasts
which deposit the silica, intended, as it were, for enclosing the (missing) terminal
parts of the axial thread, at the ends of the (short) axial thread present. The
influences regulating the production (growth) of the axial threads seem accord-
ingly to be distinct from, and to a certain extent independent of, those regulating
the action of the silicoblasts. In var. megasterra and var. megaclada the iso-
actine rhabds are as a rule spindle shaped, 1.6—-2.8 mm., mostly 1.9-2.3 mm.
long, and 33-60 y, mostly 40-55 # thick. Cylindrical rhabds reduced in length
and thickened at the ends, like those described above from var. carolae, are
exceedingly rare in var. megasterra and seem to be absent altogether in var.
megaclada. As will be seen from the measurements of these spicules of var.
megasterra tabulated below, the terminal attenuation is very variable, and the
thickness of the ends correlated to the length and the thickness in the middle only
in so far as the two longest spicules in the list both have slender pointed ends.

GEODINELLA ROBUSTA. 211

ISOACTINE RHABDS (AMPHIOXES) OF VAR. MEGASTERRA.

Thickness 40 Thickness in the

E ;
ma below each end middle Benet
pointed 15 4 57 1.8 mm.

“ 17 * 48 « 27 ‘

“ nee 55 | 2.6
rounded off 226 40 “ DED

“ee “eo 99 ce 53 “ce | 2 E 1

a ce 23 “ce 48 “ce 1 Ss

“oe “ 28 oe 53 “e 2.2

“e “ 32 “oe 57 “ce A) D)

The anisoactine rhabds (amphioxes and amphistrongyles) and the true
styles (Plate 4, fig. 17) are slightly shorter and (at the stouter end) a trifle thicker
than the isoactines. They measure 1.1—2.3 mm. in length and 40-80 yp in thick-
ness. In the true styles, which represent, as it were, the end of the series of
increasingly anisoactine rhabds, the thickness of the rounded end is generally
speaking in inverse proportion to the length of the spicule. Some of the styles
of both the specimens. ° var. carolae are somewhat thickened at the end and
appear as subtylostyles —

Irregular rhabds (Plate 4, figs. 6, 7, 10). Not a few of the rhabds of var.
carolae have a slightly undulating surface which renders their contour percep-
tibly wavy. Other rhabds, both of this variety and of var. megasterra, possess
on one side a small, rounded, well-defined protuberance 5-10 y high, which is
usually nearer to one of the ends than to the centre of the spicule. Below the
protuberance the silica-layers forming the spicule conform to the outer surface,
this disturbance (upheaval) reaching right down to the axial thread and thus
showing that the cause of the formation of the protuberance acted before the
silica-layers were produced. Sometimes more than one such protuberance is
observed on a rhabd. In the spicule of var. megasterra (Plate 4, figs. 6-7) quite
a cluster of such protuberances rises from each end of the spicule. In some cases
the protuberance is not, as in the spicules described above, confined to one side
but goes nearly or quite round it, forming a more or less complete annular thicken-
ing. Among the irregular, blunt amphioxes of var. megasterra I have observed
some with an annular thickening of this kind below each of the ends. One of
these spicules was 1.7 mm. long and 60 » thick in the middle. The two rounded
ends were respectively 30 and 36 thick. One of the annular thickenings was
quite complete and situated 230 » below the more slender end; the other was
not quite complete and situated 90 p below the stouter end of the spicule. The

PAM GEODINELLA ROBUSTA.

former had a diameter of 50 » and the part of the spicule from which it arose
was 42 4 thick. The latter had a diameter of 43 4 and the part of the spicule
from which it arose was 40 # thick. Occasionally, but very rarely, rhabds with
a clade-shaped protuberance, resembling anamonaenes (Plate 4, fig. 10) occur.

In the rhabds with undulating surface the irregularity is probably caused
by some inequality in the action of the silicoblasts during growth. The monaene-
like forms just referred to are altogether abnormal, probably pathological. The
rhabds with the rounded and annular protuberances I am inclined to consider
as spicules transitional between regular rhabds and teloclades.

The teloclades and teloclade-derivates (Plate 1, figs. 1-11, 13-15, 18-20;
Plate 4, figs. 23, 25). The cladome of the teloclades is always reduced. This

reduction is different in degree and in kind in the three varieties. In var.
carolae the teloclades have entirely lost one or two of the triaene-clades; in the
two others invariably two. In var. megaclada the single remaining clade is often
quite long. In the two other varieties the clades are always very short. Thus
var. carolae possesses some diaenes besides the monaenes, both with short
clades, var. megaclada only monaenes, many of which have a rather long clade,
and var. megasterra only monaenes, which always have a short clade. In all
three the cladome is usually simple and situated at or near the cladomal end
of the rhabdome. Besides these ordinary teloclades, teloclade-derivates with
more cladomes than one, and with clades arising some distance from the end
or reduced to insignificant protuberances, are met with.

The diaenes and monaenes of var. carolae (Plate 1, figs. 5-11) have the
same dimensions. The monaenes (Figs. 6, 9-11) are much more numerous
than the diaenes (Figs. 5, 7, 8), particularly in the specimen from Naha Bay.
The rhabdome is 1.1-1.7 mm. long and at the cladomal end 26-40 y thick;
it is generally straight or slightly curved, rarely (Fig. 6) angularly bent, and
usually attenuated towards the acladomal, blunt or, more rarely, pointed end.
Sometimes (Fig. 5) this attenuation is so slight that the rhabdome appears
nearly cylindrical. In such spicules it is simply rounded off at the end. The
cladome is generally quite terminal (Figs. 5-9, 11), rarely situated a little below
the end of the rhabdome (Fig. 10). The clades are 30-70 yp, usually 40-55 p long,
generally quite straight, irregularly conical, and pointed (Figs. 5, 6, 8-10),
or, more rarely, cylindrical and rounded terminally (Fig. 11). In the diaenes
a pointed clade may be associated with a rounded one (Fig. 7). The clades
enclose angles of 93-130° with the rhabdome, so that some of these spicules ap-
pear as orthodiaenes or orthomonaenes (Fig. 10), others as plagio- or pro-diaenes

GEODINELLA ROBUSTA. Die

or plagio- or pro-monaenes (Figs. 5-9, 11). The latter are much more numerous
than the former. The monaenes of var. megasterra (Figs. 1-4, 18-20) are in
every respect similar to those of var. carolae, the only difference being that their
rhabdomes, which measure 1.4—2.1 mm. in length, are on an average slightly
longer. The monaenes of var. megaclada (Plate 1, figs. 13, 14; Plate 4, fig. 23)
have rhabdomes similar to those of the monaenes of var. megasterra but thicker,
sometimes attaining a thickness of 42 4. The clade which is terminal (Plate 1,
fig. 14), or, more rarely, situated a little below the end of the rhabdome (Plate 1,
fig. 13; Plate 4, fig. 23), is conie, pointed, 80-105 long, and straight or slightly
bent upwards at the end. It encloses an angle of 87-135° with the rhabdome.

The teloclade-derivates (Plate 1, fig. 15; Plate 4, fig. 25). The teloclade-
derivates with more than one cladome are rare. I have observed them only
among the monaenes, and they never seem to have more than one secondary
cladome. The primary (terminal) clade is similar to that of ordinary mo-
naenes, the secondary clade is situated a considerable distance below the
cladomal end and smaller than the primary clade (Plate 1, fig. 15). The
mesoclades are likewise rare. They are always monoclade and appear as rhabds,
attenuated towards both ends or towards one end only with a short and stout
clade arising a considerable distance from either end (Plate 4, fig. 25). The
teloclades with terminal clades reduced to mere rounded protuberances pass,
by further cladome-reduction, into tylostyles and styles. They appear as
transitional forms connecting the teloclades with the rhabds. In the teloclade-
derivate tylostyles the tyle is often irregular and the axial thread of the rhabd-
ome becomes tortuous on entering the tyle. In one of these spicules I noticed
that the short, tortuous part of the axial thread lying in the tyle was not con-
nected with the axial thread of the rhabdome. Mesoclade and multicladomal
teloclade-derivates with clades further reduced are rare. They pass into the
rhabds with one or more protuberances or annular thickenings (Plate 4, figs.
6, 7). These resemble the rhabds proper so closely, that I have thought it
better to describe them above together with the regular rhabds.

The shape, size, and arrangement of the megascleres of these sponges and
the closeness of their connection by transitional forms lead to the conclusion
that the rhabds are more closely related to the teloclades than is generally
assumed.

The fragments of long and slender spicules, which, as stated above, are met
with occasionally in the spicule-preparations, are 8-12 y thick. The longest
one observed was over 2 mm. in length. Most of them are broken off at both

214 GEODINELLA ROBUSTA.

ends. In a few one end was intact and pointed. Their slenderness would
incline me to believe that they are parts of rhabdomes of teloclades, but the fact
that, in spite of the most careful search, I have failed to find any cladomes
belonging to them, is against this view. As stated above, these spicules may
be foreign to the sponge.

The large oxyasters and small oxysphaerasters (Plate 2, figs. 3a, 10a, b; Plate
4, figs. 1-5, 21, 22) are so closely connected by transitional forms that it is
advisable to describe them together. In var. carolae these spicules are 11-36
in diameter and have from six to eighteen rays. In many a central thicken-
ing 2-8 » in diameter is clearly distinguishable, others are without such a
centrum. The rays are straight, conic, 1-3.5 » thick at the base, and (without
the central thickening) 3-17 » long. They are usually simple, but occasion-
ally such asters are observed in which one.or more of the rays are bifurcate,
the two branches extending in a nearly parallel direction and lying close
together. The rays are pointed (Plate 4, fig. 3) or, rarely, somewhat blunt (Plate
4, fig. 2). Their distal part is covered with spines, the size, number, and
arrangement of which are variable. In some (Plate 4, fig. 2) the spines are
so small that even with the 280 yy light no distinct image of them can be pro-
cured; as a rule, however, they are large enough to be clearly shadowed on the
photographic plate by these u. v. rays (Plate 4, fig. 3). The number of rays
and the development of the central thickening are, roughly speaking, in inverse
proportion to the size of the spicule. Oxyasters (oxysphaerasters) under 20
in diameter have from ten to eighteen rays and a well-developed central thick-
ening, the diameter of which is from one fifth to nearly one half of the diameter
of the whole spicule. Oxyasters (oxysphaerasters) over 20 in diameter usually
have only from six to nine rays, and either no central thickening at all, or only a
small one, never more than a quarter of the whole spicule in size. Also in the
spines a certain (inverse) proportion between size and number is discernible;
when the spines are numerous, they are very small, and the smaller their num-
ber is the larger they become. When, as is most frequently the case, the spines
are few in number and large in size, some, generally the longest, form a verticil a
little below the end of the ray, so that the spicule becomes somewhat acanthtyl-
aster in character (Plate 4, fig. 3). The oxyasters and oxysphaerasters of var.
megasterra are similar to those of var. carolae. They measure 9-38 s in diameter
and have from six to seventeen rays (without the central thickening) 2.5-21 px
long, and 0.64 thick at the base. The central thickening is small, never over
5 in diameter. In many of these spicules (Plate 4, fig. 4) the spines of the rays

GEODINELLA ROBUSTA. 215

are particularly large, and very regularly arranged in verticils. These large
Spines arise vertically from the ray and often appear to be bent down at the
end so that they become claw shaped. Also in the oxyasters and oxysphaer-
asters of this variety an inverse relation between the number of rays and the
size of the spicule is discernible; the asters under 20 in diameter having from
nine to seventeen, the asters over 20 » in diameter, from six to eleven rays.
The oxyasters and oxysphaerasters of var. megaclada are smaller than those of
var. megasterra, only 11-30 y in diameter, and have on an average more rays,
the small ones (under 20 in diameter) up to twenty, the large ones (over 20 p
in diameter) eleven to fifteen.

The strongylosphaerasters (Plate 2, figs. 3b, 8, 9, 10c; Plate 4, figs. 18-20) of
var. carolae consist of a spherical, central thickening, from which from eleven
to twenty-seven radial rays arise. The whole aster is 7-12 yin diameter. The
diameter of the central thickening is usually from one half to two thirds of the
diameter of the whole spicule and measures 3.5-7 p. The rays are cylindrical
and arise from the central thickening with trumpet-shaped basal extensions.
They are 2-3 » long, 1-1.7 « thick, and terminally rounded. The distal parts
of the rays are covered with small spines which often form a conspicuous terminal
verticil (Plate 4, fig. 19). A correlation between the size of the spicule and the
number of rays is not discernible. In var. megaclada the maximum dimensions
of these spicules are similar, but the minimum dimensions greater. The
strongylosphaeraster of this variety measured were 10-12 y in diameter, and
had from nineteen to twenty-five rays and a central thickening 5-7 y in
diameter. Among the strongylosphaerasters of this variety I have observed
many in which the verticillate arrangement of the ray-spines was particu-
larly well marked, and I noticed that in many of these the spines of the verticils
are recurved. The rays of these spicules, particularly when viewed from above,
closely resemble sterraster-rays. In var. megasterra the strongylosphaerasters
are also similar, but here they attain a somewhat larger size, measure 8-13 yin
diameter, and have fewer, only from twelve to nineteen, rays; the central thick-
ening is 3.5—7 in diameter.

Sometimes sphaerasters, similar to these strongylosphaerasters, but with
rays distinctly tapering towards the distal end, are observed. These spicules
are transitional to the small oxysphaerasters. I have noticed such sphaeras-
ters particularly in var. carolae.

The normal sterrasters (Plate 2, figs. 1, 2,5; Plate 3, figs. 1-3, 7, 9; Plate 4,
fig. 13). As stated above the size and shape of the normal sterrasters are

216 GEODINELLA ROBUSTA.

different in the three varieties, var. carolae having the smallest, and var. mega-
sterra the largest; var. megaclada being in this respect intermediate between the
other two. In var. carolae all or nearly all the sterrasters are flattened ellip-
soids, the proportion between the three axes being about 5:7:9 (Plate 2, fig.
5). In the other two varieties most of the sterrasters have a similar shape
(Plate 2, figs. la, 2a, e); but we find in these, among the ordinary, ellipsoidal
sterrasters, also a good many flattened, three-lobed ones (Plate 2, figs. 1b, 2b).
The ellipsoidal sterrasters are in var. carolae 180-195 y long, 130-160 « broad,
and 80-115 # thick; in var. megaclada 190-217 » long, 160-190 » broad, and
105-125 » thick; in var. megasterra 220-237 » long, 165-200 p broad, and 120-
130 » thick. The three nearly equal maximum diameters of the three-lobed
sterrasters of the two last-named varieties are nearly or quite as long as the

longest diameter of their ellipsoidal sterrasters. In the specimen of var. carolae
from Naha Bay I found two tetra-lobed sterrasters.

The centre of the sterraster is, as Thiele has already noticed in another
species, Geodinella (Geodia (?)) cylindrica,’ surrounded by granules the refrac-
tive index of which differs from that of the silica in which they are imbedded.
These granules form a hollow, spherical cluster 6-8 » in diameter (Plate 3,
figs. 7a, 9a). Rather to my surprise I found that in the three-lobed sterr-
asters the position of these pericentric granules is the same as in the ellip-
soidal ones. The siliceous substance surrounding this cluster of granules
shows the usual radial structure. The individual granules often appear to be
onion shaped and produced distally in a radial process, forming one of the radial
lines which give the radially striated appearance to the siliceous substance of
the sterraster. Sections, optical and other, through the sterrasters show that
many of them are not only radially striated but also paratangentially stratified,
one or two, very conspicuous limits (Plate 3, figs. 7, 9b) between the super-
posed zones being distinctly visible. These limits are concentric and parallel
to the outer surface. The radial striations pass continuously through them.

The umbilicus (Plate 3, figs. 1-3, 7c, 9c; Plate 4, fig. 13a) lies in the
centre of one of the broad faces of the sterraster. It is generally a caliculate pit
15-25 deep; its circumference (mouth) is oval, 17-20 «4 broad and 23-30 p long.
Proximal continuations of the rays surrounding the umbilicus project into the
umbilical pit and form longitudinal (radial) ridges on its flanks. The bottom
of the pit appears rough. The remainder of the surface of the sterraster is
covered by freely projecting rays (Plate 2, figs. 1, 2, 5; Plate 3, figs. 1-3, 7, 9;

Zoologica, 1898, 24, p. 13.

GEODINELLA ROBUSTA. 217

Plate 4, fig. 13). These rays are cylindrical, 3 long and about as broad, and
rather uniformly distributed over the surface, their axes being 5 # apart. Those
portions of the surface of the central mass of the sterraster which lie between the
rays are more or less roughened. At the end each ray bears a verticil of stout,
conic, lateral spines, 1-2 » long and broad, the axes of which are vertical to the
axis of the ray. The average number of the spines ina verticil is six; but there
may be as few as one or two or as many as eight or ten.

Among these ordinary sterrasters, forming, as stated, the great majority
in all the three varieties, a few sterroids occur, which are similar to these in shape
and size, but have different rays. Two kinds of sterroids can be distinguished.

In the one, which is observed more frequently, the rays are, as in the normal
sterrasters, quite uniformly distributed and close together, but wholly or in
part much larger and crowned with a much greater number of spines. In these
sterroids (Plate 3, figs. 4, 6; Plate 4, figs. 14, 15) the rays attain a thickness
of 7-10 p, the spine-verticils are composed of 15-20 spines, and the convex
apical ends of the rays also bear several, usually 4-8, spines equalling in size
the verticil-spines (Plate 4, figs. 14,15). In these sterroids the verticil-spines
are usually directed slightly downwards, the whole verticil appearing as the
serrated and somewhat recurved margin of a terminal, shield-like expansion of
the ray, from the distal face of which several spines arise.

The other kind of sterroid (Plate 3, figs. 5, 8), which is very rare, consists
of a central mass of the usual ellipsoidal or a more spherical shape, from which
rather sparse and irregularly distributed rays arise. These rays are cylindro-
conical, 22-27 » long, 11-17 y thick at the base, and covered with numerous
small spines; on the parts of the surface of the central mass free from rays such
spines also occur. On the rays the spines form extensive patches within which
they stand quite close together. On the central sphere they are for the most
part farther apart and irregularly scattered. Here and there well marked,
smooth, channel-like zones separate adjacent spine-patches.

The four specimens of this species were obtained on the Pacific slope of
North America. One specimen of var. carolae was trawled at Station 4199 on
June 25, 1903, in Queen Charlotte Sound off Fort Rupert, Vancouver Island,
B. C.; centre of Round Island 8. 46° W., 11.5 km. (6.2 miles), drift 8. 85° E. ;
depth 124-196 m. (68-107 f.); it grew on a bottom of soft green mud and vol-
canic sand; the bottom temperature was 7.7° (45.9° F.). The specimen of var.
megaclada and one specimen of var. carolae were trawled at Station 4228 on July
7, 1903, in the vicinity of Naha Bay, Behm Canal, 8. E. Alaska; Indian Point, N.

218 GEODINELLA ROBUSTA.

18° E., 1.7 km. (0.9 miles); drift N.2° W.; depth 75-245 m. (42-134 f.); they
grew on a bottom of gravel and sponge spicules; the bottom temperature was
8.8° (47.8° F.). The specimen of var. megasterra was trawled at Station 2946 on
February 6, 1889, off southern California, in 33° 58’ N., 119° 30’ 45” W.; depth
274 m. (150 f.); it grew on a bottom of coarse gray sand; the bottom tempera-
ture was 13.6° (56.5° F.).

TABLE SHOWING THE VARIETAL DIFFERENCES IN GEODINELLA ROBUSTA.

Varieties carolae megaclada megasterra

incrusting, cushion

Shape haved finger shaped. finger shaped.
amphioxes; amphistron-
vles; amphityles; F pee 3
Rhabds ace Ser amphioxes, 1.6-2.8 mm. | amphioxes, 1.6-2.8 mm.

0.37-2.5 mm. long, 40- long, 33-60 » thick. long, 33-60 » thick.

65 p» thick.

monaenes and oecasion-
ally diaenes; rhabdome
1.1-1.7 mm. long, 26—
40 » thick; clades 30-
70 «long; cladal angles
93-130°.

monaenes only; rhabd- | monaenes only; rhabd-
ome 1.4—-2 mm. long, | ome 1.4-2.1 mm. long,
25-42 » thick; clades | 26-40 » thick; clades
80-150 » long; cladal | 30-70 » long; cladal
angle 85-135°. angles 93-103°.

Teloclades.

11-36 » in diameter;| 11-30 js in diameter; | 9-38 » in diameter; with
Oxyasters and oxy- | with 6-18 rays; central | with 11-20 rays; cen-| 6-17 rays; central thick-

sphaerasters. thickening usually well | tral thickening usually | ening usually small or
developed. small or absent. absent.

7-12 « in diameter;| 10-12 y in diameter; | 8-13 pin diameter; with
with 11-27 rays; central | with 19-25 rays; cen-| 12-19 rays; central
thickening 3.5-7 yp in| tral thickening 5-7 | thickening 3.5-7 in
| diameter. in diameter. diameter.

Strongylosphaerasters.

ellipsoidal or, more | ellipsoidal or, more
ellipsoidal; 180-195 | rarely, three-lobed; the | rarely, three-lobed; the

Sterrasters. long, 130-160 » broad, | ellipsoidal ones 190-217 | ellipsoidal ones 220-
80-115 » thick. jelong, 160-190 » broad, | 237 « long, 165-200
105-125 p thick. broad, 103-120 » thick.

In 1898 Thiele (Zoologica, 24, p. 12) described a geodid sponge with reduced
and irregularly arranged, partly axially situated teloclades and large ellipsoidal
sterrasters from the northwestern Pacific (Japan). As the specimen at his
disposal was merely a small fragment, he, although convinced that it did not
belong to any of the geodid genera then known, refrained from establishing a
new genus for it, and named it Geodia (?) cylindrica.

When I was preparing the systematic account of the Tetraxonia for the

GEODINELLA ROBUSTA. 219

Tierreich I found it quite impossible to place this sponge in any of the then
existing genera, and, carrying out Thiele’s suggestion, established the new genus
Geodinella for it.’

There can be no doubt that the sponges above described belong to this genus.
From the only species of it hitherto known, Geodinella (Geodia ?) cylindrica, they
differ in having a much stouter body, megascleres twice as large, and different
euasters. Geodia cylindrica has oxysphaerasters (pycnasters) 7-8 in diameter
only; in G. robusta besides the strongylosphaerasters 7-13 yin diameter, which
appear to correspond to the pyenasters of G. cylindrica, oxyasters attaining a
diameter of 30-38 y are present in large numbers. (eodia cylindrica has been
found only in Japan, G. robusta on the Pacific coast of North America.

'R. v. Lendenfeld. Ticrreich, 1903, 19, p. 117.

III. GENERAL SYSTEMATIC ACCOUNT OF THE GENERA, SPECIES,
AND VARIETIES OF GEODIDAE FROM THE PACIFIC OCEAN.

Geodidae.

Tetraxonia with rhabd, teloclade and usually also mesoclade megascleres,
and a superficial armour composed of massive, spheroidal, or ellipsoidal sterr-
asters. Euasters are always, ataxasters or microrhabds sometimes, present.
Without desme megascleres and without thin, dise-shaped sterrasters.

This family, as now limited, comprises the genera Caminella Lendenfeld,
Pachymatisma Johnston, Caminus O. Schmidt, Isops Sollas, Sidonops Sollas,
Geodia Lamarck, and Geodinella Lendenfeld.

All of these, with the exception of Pachymatisma, occur in the Pacifie Ocean.

Ninety-four species of Geodidae are known; forty-six occur in the Pacific

Ocean, and five of the Pacific species are further divided into fifteen varieties.

CAMINELLA LEeNpreNrFeLD.

Among the megascleres are regular triaenes. The tetraxon megascleres
are confined to the superficial part of the sponge and arranged radially. The
dermal microscleres are asters. The afferents are uniporal; the efferents larger
oscula.

Two species are known; one of which, C. nigra (Lindgren), occurs in the
Pacific Ocean.

Caminella nigra (LinpGREN).

LENDENFELD, Tierreich, 1903, 19, p. 90.
Isops nigra LinDGREN, Zool. anz., 1897, 20, p. 486. Zool. jahrb. Syst., 1898, 11, p. 352, plate 18, fig. 11,
plate 20, fig. 7a—e.

Ellipsoidal. Black.

Large amphioxes: 900 by 20 ». Minute amphiores: 72 by 2 p. Plagio-
protriaenes: rhabdome 960 by 20 »; clades 96 » long; cladome 180 broad
and 60 high; clade-angles about 135°.

ISOPS CONTORTA. 221

“=

Choanosomal oxyasters: Rays not numerous, smooth, pointed; centrum
small; total diameter 24 4. Oxysphaerasters: numerous conical rays; centrum
very large; total diameter 20 y. Sterrasters: 62 y long and broad, 52 y thick.

Western Pacific. Java Sea; Gaspar Strait.

CAMINUS 0. Scum.

Among the megascleres are regular triaenes. The tetraxon megascleres
are confined to the superficial part of the sponge and arranged radially. The
dermal microscleres are asters. The afferents are cribriporal, the efferents larger
oscula.

Four species are known, one of which, C. chinensis Lindgren, occurs in the
Pacific Ocean.

Caminus chinensis LinpGREN.

Zool. anz., 1897, 20, p. 485. Zool. jahrb. Syst., 1898, 11, p. 339, plate 17, fig. 16, plate 20, figs. 2a-e, e.

LENDENFELD, Tierreich, 1903, 19, p. 92.

Spherical or ellipsoidal, erect. Brown. Cortex very hard.

Amphistrongyles: 720 by 24 ». Orthotriaenes: rhabdome 460-600 by 36 4,
blunt; clades 325-540 » long; clade-angles, according to the figure 90-100°.

Oxyasters: centrum small; total diameter 24-32 ». Sphaeres: 2-5 win diam-
eter. Sterrasters: 136 by 108 by 90 p.

Northwestern Pacific. China Sea; Strait of Formosa.

ISOPS Sottas.

Among the megascleres are regular triaenes. The tetraxon megascleres
are confined to the superficial part of the sponge and arranged radially. The
dermal microscleres are asters. The afferents are uniporal; the efferents uni-
poral.

Of the seventeen species known four occur in the Pacifie Ocean.

Isops contorta (BowrRBANK).

Souuas, Rept. voy. “Challenger,” 1888, 25, p. 271. LeNDENFELD, Tierreich, 1903, 19, p. 97.
Pachymastina contorta BOWERBANK, Proc. Zool. soc. London, 1873, p. 327, plate 31, figs. 7-11.

Branching, branches anastomosing. Dry: light brown.

Large amphioxes: 1.838 mm. by 35 p. Small amphioxes: about 0.5 mm.
long. Styles: 1.3 mm. long. T'riaenes: rare.

Oxyasters: total diameter 32 p. Small sphaerasters: rays conical; total
diameter 70 . Sterrasters flattened, ellipsoidal, 160 y long.

Western Pacific. Fiji Islands.

999 SIDONOPS.

Isops imperfecta (BOWERBANK).

Sotzas, Rept. voy. “Challenger,” 1888, 25, p. 269. LenDENFELD, Tierreich, 1903, 19, p. 97.
Geodia imperfecta BoWERBANK, Proc. Zool. soc. London, 1874, p. 299, plate 46, figs. 6-13.

Massive tuberous. Dry: white.

Amphioxes and amphistrongyles: 1.75 mm. by 39 p. Teloclades: triaene,
diaene, or monaene; clades usually rounded terminally.

Sphaerasters: rays slender, cylindrical, truncate; centrum small; total
diameter 21 4. Dermal sphaerasters: rays terminally divided into numerous
spines; centrum large; total diameter 12-19 y. Sterrasters: ellipsoidal, de-
pressed, or cylindrical, 110 by 80 4.

? Southern Pacific. South Sea.

Isops obscura THIELP.

Zoologica, 1898, 24, p. 6, plate 2, fig. 2; plate 6, figs. 2a—k. LENDENFELD, Tierreich, 1903, 19, p. 96.

Trregularly massive. Dark brown.

Amphioxes: 1.2-1.6 mm. by 15-40 ». Plagiotriaenes: rhabdome curved,
1.25 mm. long; clades stout, 80-120 » long; clade-angles large.

Large thin-rayed oxyasters: rays slender, 20 » long. Small thick-rayed
oxyasters: total diameter 75 ». Oxysphaerasters: diameter the same as in the
oxyasters. Sterrasters: spherical, 60 » in diameter. Sterroids: rare, similar
to the sterrasters but with much thicker rays.

Northwestern Pacific. Japan.

Isops sollasi LeENDENFELD.

Descriptive catalogue sponges Australian museum, 1888, p. 34. Tierreich, 1903, 19, p. 97.

Cup shaped, peduncular, or lamellar. Brown.

Choanosomal rhabds: 800 by 16 . Plagiotriaenes: rhabdome 1 mm. by
25 w; clades blunt 260 » long.

Euasters: 16-30 pin diameter; smaller many-rayed and larger few-rayed to
be distinguished. Sterrasters: 48 yin diameter.

Southwestern Pacific. East coast of Australia; Port Jackson.

SIDONOPS Soutas.

Among the megascleres are regular triaenes. The tetraxon megascleres
are confined to the superficial part of the sponge and arranged radially. The
dermal microscleres are asters. The afferents are cribriporal; the efferents
uniporal.

SIDONOPS PICTETI. 223

Twenty species are known; nine occur in the Pacific Ocean.
The species with anatriaenes found in the Pacific Ocean are S. lindgreni Ldf.,
S. picteti (Tops.), S. californica Ldf., S. alba (Kieschnick), S. angulata Ldf.
(var. megana Ldf., var. microana Ldf., var. orthotriaena Ldf.), S. oxyastra Ldf.,
S. reticulata (Bwk.); those without anatriaenes are S. bicolor Ldf., S. nitida
(Soll.).
Sidonops lindgreni LENDENFELD.

Tierreich, 1903, 19, p. 102.

Sidonops picteti LINvGREN (non Topsent 1897), Zool. anz., 1897, 20, p. 486. Zool. jahrb. Syst., 1898, 11,
p- 349, plate 18, figs. 17, a, b, plate 20, fig. 6, a-h, e’-e”, d’. Kirxparrick, Proc. Zool. soe. London,
1900, p. 130.

Massive irregular, sometimes with digitate processes. Cortex brown,
choanosome grayish.

Large choanosomal amphioxes: 2.5 mm. by 40 y. Minute dermal styles:
240 by 5 yw. Orthotriaenes: rhabdome 2.4 mm. by 54 y; clades slender, strongly
concave to the rhabdome, 756 ys long. Proclades (mesoproclades): triaene,
diaene, or monaene; rhabdome 4.6 mm. by 24 y; clades 100 » long; cladome
140 » broad and 70 # high. Anatriaenes: rhabdome 3 mm. by 12 yw; clades
stout, 68 «long; cladome 80 y broad and 60 y high.

Choanosomal oxyasters: rays rough, pointed, very numerous; centrum
small; total diameter 20-48 p. Small strongylosphaerasters: centrum large ;
total diameter 4 4. Sterrasters: 160 by 120 pu.

? Western Pacific. Java.

Sidonops picteti TorsEnt.

Revue Suisse zool., 1897, 4, p. 431, plate 18, fig. 2. LENDENFELD, Tierreich, 1903, 19, p. 103.

Massive tuberous, elongated. In spirit: reddish gray.

Amphiozes: 500-600 by 30 ». Orthotriaenes; rhabdome of similar dimen-
sions as the amphioxes; clades slender, 90 » long; cladome 185 ys broad.
Protriaenes (? mesoprotriaenes): rhabdome 10 yp thick; clades 70 yw long.
Anatriaenes: rhabdome 10 » thick; cladome 70 « broad and 53 y high.

Oxyasters: usually from seven to twelve slender, spined, conical rays, 2 4
thick; centrum small; total diameter 35-40 ». Strongylosphaerasters: centrum
large; total diameter 4-6 ». Sterrasters: 97 by 85 p.

Topsent, 1897, states that in another specimen the sterrasters measured
140 by 115». T should say that this sponge probably belongs to another species.

Western Pacific. Bay of Amboyna.

224 SIDONOPS ALBA.

Sidonops californica LENDENFELD.

Ante, p. 18.

Elongate, tuberous. In spirit: yellowish white.

Large choanosomal amphioxes: 1.2-2 mm. by 30-48 uu. Large styles: rare;
55 «thick. Minute dermal styles: 175-290 by 3-7 p. Very minute amphiozes:
50 by 1 y, possibly foreign. Orthoplagiotriaenes: rhabdome 0.9-1.45 mm. by
20-78 yw; clades concave to the rhabdome, 160-400 » long; clade-angles 104-
120°. Mesoplagioclades: rhabdome 6-15 y thick; from one to three clades,
20-42 » long; clade-angles 102-118°. Anatriaenes: rhabdome 10-17 4 thick;
clades stout, 22-45 y long, clade-angles 45-66°. Anadiaenes and anamonaenes:
of similar dimensions; rare.

Large choanosomal oxyasters: six to fourteen rays, 1.7-3 y thick, covered
everywhere, except at the base with spines; total diameter 22-48 yw; size in
inverse proportion to ray-number. Small oxysphaerasters: 7-9 in diameter.
Small strongylosphaerasters: from six to seventeen spined rays, rounded termi-
nally and 0.8-1.5 » thick; centrum 2-3.5 4, whole aster 4.5-9 #, in diameter.
Sterrasters: 116-130 by 97-105 by 70-90 pu.

Eastern Pacific. West coast of North America; 22° 52’ N. ‘‘Albatross”’
Station 2829.

Sidonops alba (KiescHNIck).

LENDENFELD, Tierreich, 1903, 19, p. 100.

Synops alba Krescunick, Zool. anz., 1896, 19, p. 529.

Sydonops alba (Kieschnick) Tarete, Abhandl. Senckenb. gesellsch., 1900, 25, p. 46, plate 2, fig. 16a—h.
Irregular, depressed, with attached foreign bodies. Dirty brownish white.
Large choanosomal amphioxes: 2.5 mm. by 30 yp. Large styles: rare, of

similar dimensions as the large amphioxes. Minute dermal styles: 250 by

5 wu. Plagiotriaenes: rhabdome 2 mm. long; clades 450 » long. Mesoproclade-

derivates: Monaene or diaene; clades reduced, short; total length of spicule

3mm.; its thickness 14 4; epirhabd 80 » long. Large anatriaenes: rhabdome

2.5 mm. by 14 #; clades, strongly recurved, about 20 » long. Minute dermal

anaclades (Thiele, 1900, exotyles): triaene, or, more rarely, diaene or monaene;

rhabdome 170 # long and about as thick as the minute dermal styles.

Oxyasters: a smaller kind with numerous rays, not quite up to 15 p long;
and a larger kind with few, rough rays, up to 30 «long. Small strongylosphaer-
asters: rays short and stout; centrum large; total diameter about 8 yp. Sterr-
asters: 110-90 p.

Western Pacific. Ternate.

SIDONOPS ANGULATA VAR. MICROANA.

bo
bo
on

Sidonops angulata LENDENFELD.
Ante, p. 24.

Massive, irregularly spherical or lobose. In spirit: yellowish in the interior
and white to reddish or purplish brown on the surface.

Stout chaonosomal amphioxes: straight, slightly curved or angularly bent;
1.6-3.7 mm. by 20-72 p. Styles and style-derivates: rare; 2.1-2.5 mm. by 60-
110 »; the derivates with a branch-ray. Slender dermal amphioxes: slightly
curved or angularly bent; 2.9-9.5 mm. by 5-34 4. Orthoplagiotriaenes: rhabd-
ome 1.5-2.8 by 47-82 y, rarely shortened and thickened, up to 105 y thick;
clades concave to the rhabdome, curvature increasing distally, 330-700 long,
clade-angles 89-112°. Anaclades: mostly triaene, sometimes diaene; rhabd-
ome up to more than 9 mm. in length and 7-39 y» thick; clades 30-210 y long;
clade-angles 27-66°.

Oxyasters and oxysphaerasters: one to twenty-three perfectly smooth, conic
rays 1.6-5 » thick; centrum, when present, up to 12 win diameter; total diam-
eter 11-64 y; size on the whole in inverse proportion to the ray-number.
Strongylosphaerasters: from ten to twenty, equal or, more rarely, unequal,
truncate, distally spined rays, 2-6 » thick; centrum 7-14 », whole aster 16-28
fein diameter. Sterrasters: 85-122 by 75-113 by 57-86 yp.

Northeastern Pacific. West coast of North America; off southern Cali-
fornia. ‘Albatross’ Stations 2945, 2975, 4417.

Sidonops angulata var. megana LENDENFELD.
Ante, p. 24.

Stout choanosomal amphioxes up to 72 4 thick. Slender dermal amphiozes,
not numerous, up to 34 » thick. Anaclades, triaene, rhabdome up to 39 y thick,
clades up to 210 » long. Strongylosphaerasters with nearly cylindrical rays.
Sterrasters up to 122 y long, all ellipsoidal.

Northeastern Pacific. West coast of North America; 34° 1’ 30” N.
“Albatross” Station 2975.

Sidonops angulata var. microana LeENDENFELD.
Ante, p. 24.
Stout choanosomal amphioxes up to 52 py thick. Slender dermal amphiozes,
very abundant, up to 22 «thick. Anaclades, triaene and, less frequently, diaene,
rhabdome up to 18 y thick, clades up to 50 «long. Strongylosphaerasters with

conical rays. Sterrasters up to 97 jp: long, all ellipsoidal.

226 SIDONOPS OXYASTRA.

Northeastern Pacific. West coast of North America; near Santa Barbara

Islands. ‘‘Albatross”’ Station 4417.

Sidonops angulata var. orthotriaena LeENDENFEéLD.

Ante, p. 24.

Stout choanosomal amphioxes up to 70 # thick. Slender dermal amphi-
oxes, not numerous, up to 17 y thick. Anaclades, triaene, rhabdome up to
18 » thick, clades up to 80 «long. Strongylosphaerasters with nearly cylindrical
rays. Stlerrasters up to 111 #4 long; besides the ellipsoidal also rhomboidal ones
occur.

Northeastern Pacific. West coast of North America; 34°N. ‘Albatross”’

Station 2945.

Sidonops oxyastra LENDENFELD.

Ante, p. 40.

Lobose. In spirit: brownish white to purplish brown.

Large choanosomal amphioxes: 1.1-1.55 by 10-32 4. Large amphistrongyies :
0.8-1 mm. by 18-23 ». Large styles: rare, 850 by 38 4. Minute dermal styles:
130-230 by 3-5.5 #. Plagiotriaenes: rhabdome 1—1.65 by 24-40 »; clades equal,
or some of them reduced, concave to the rhabdome, the distal part of long ones
often curved in the opposite direction, when fully developed, 250-285 y long;
clade-angles 100-118°. Anaclades: mostly triaene, sometimes diaene or mon-
aene, rhabdome over 1 mm. long, 5-12 « thick; a protuberance on the apex
of the cladome, sometimes elongated to form an epirhabd 56-75 long; clades
15-30 y long; clade-angles 40-65°.

Large choanosomal oxyasters: from four to ten, usually seven rays, 0.5-
2 » thick and covered everywhere, except quite at the base, with spines; cen-
trum very small; total diameter 18-45 p. Large subcortical oxysphaerasters:
from sixteen to twenty-three conical, spined rays 1.1-1.4 m thick; centrum
4.2-6.5 4, whole aster 16-22 », in diameter. Small dermal oxyasters and
oxysphaerasters: nine to eighteen spined rays, 0.7-1.5 « thick, centrum, when
present, up to a third or more, of the whole aster in diameter; total diameter
6-13.5 pw. Sterrasters 76-85 by 66-73 by 50-64 pu.

Eastern Pacific. Galapagos; Duncan Island.

SIDONOPS BICOLOR.

bo
i)
“I

Sidonops reticulata (BowreRBANK).

LENDENFELD, Tierreich, 1903, 19, p. 102.
Geodia reticulata BOWERBANK, Proc. Zool. soc. London, 1874, p. 300, plate 46, figs. 14-20. Sounas, Rept.
voy. “Challenger,” 1888, 25, p. 253.

Massive. Dry: white.

Amphioxes: 1.75 mm. by 19 p. Orthoplagiotriaenes: rhabdome 2 mm.
by 31.2 y; clades 237.5 » long; cladome 450 » broad. Protriaenes: (mesopro-
triaenes ?): clades 250 » long; cladome 120 » broad and 206 » high. <Ana-
triaenes: clades 144 » long; cladome 150 » broad and 125 » high.

Oxyasters: no centrum; total diameter 25.4 yp. Small strongylasters:
total diameter 8.5 . Sterrasters: spherical, small.

? Eastern Pacific. Mexico.

Sidonops bicolor LENDENFELD.

Ante, p. 46.

Trregularly tuberous, elongated or flattened. In spirit: whitish to reddish
or purple-brown, some parts of the surface usually much darker than others.

Stout amphioxes: rather blunt; 2.3-5.6 mm. by 36-105 py. Stout styles:
rare, not in all specimens; 4 mm. by 100-200 yp. Slender amphioxes: 3.5—9 mm.
by 15-40 ». Plagiotriaenes: rhabdome usually 2.1-4 mm. by 82-110 y at the
cladome, and 3-10 » more a little below, at the thickest point; sometimes re-
duced in length, cylindrical and rounded; clades, concave to the rhabdome,
nearly straight, or concave to the rhabdome proximally and convex to it distally,
blunt, 280-700 » long; clade-angles 103-122°.

Large choanosomal oxyasters: from one to twelve, most frequently from
seven to nine, distally spined rays, 1-2.8 « thick; centrum 2.8-6 4, whole aster
19-34 yw, in diameter. Ozysphaerasters: from twelve to twenty-five distally
spined rays, 0.7-2.5 » thick; centrum 4-10 y, whole aster 10-23 y, in diameter.
Strongylosphaerasters: usually from nine to thirty, very rarely only one, distally
spined rays, 0.7-3.5 » thick; centrum 4-12 », whole aster 9-21 y, in diameter.
Sterrasters: 130-170 by 100-133 by 77-97 pu.

Northeastern Pacific. West coast of North America; from 33° 18’ N. to
Monterey Bay. ‘‘Albatross’’ Stations 2958, 2981, 3168, 4420, 4531, 4551.

D9 GEODIA.

aa

Sidonops nitida (SoLuas).
Proc. Roy. Dublin soe., 1889, 6, p. 277. Jeske Tierreich, 1903, 19, p. 104.
Synops nitidus Souias, Proe. Roy. Dublin soe., 1886, 5. p. 198. Rept. voy. ‘“‘Challenger,” 1888, 25, p. 231,
plate 22, figs. 1-18.

Lamellar, attenuated towards the margin. In spirit: faint brownish white.

Amphioxes: 1.25 mm. by 26 ». Orthoplagiotriaenes: rhabdome 1.07 mm.
by 28.7 w; clades 183 long; cladome 358 broad, 50 # high.

Choanosomal oxyasters: usually about seven, sometimes as few as two,
stout blunt rays with large spines; centrum small; total diameter, 43.4 4.
Strongylosphaerasters: rays thick, terminally rounded; total diameter 13.5 y.
Sterrasters: spherical, 51.6 in diameter.

Southwestern Pacific. East coast of Australia; Port Jackson.

GEODIA Lamarck.

Among the megascleres are regular triaenes. The tetraxon megascleres
are confined to the superficial part of the sponge and arranged radially. The
dermal microscleres are asters. The afferents are cribriporal, the efferents
cribriporal.

Forty-four species are known, twenty-nine of which occur in the Pacific
Ocean.

SUMMARY OF THE SPECIES FOUND IN THE PACIFIC OCEAN.

A, Fully developed triaene mesoproclades (proclades) present.
A, The large choanosomal asters are oxyasters, rarely strongylasters, never
acanthtylasters.
A; With sphaerasters up to 90 in diameter.
G. nux (Selenka).
B, The largest sphaerasters under 60 in diameter.
A, With sphaerasters up to 31 yin diameter, with very large centrum
and numerous very short and thick rays.
A; The rays of the sphaerasters with large centrum are conic.
G. eosaster (Sollas). G. globostella Lendenfeld.
B; The rays of the sphaerasters with large centrum are cylindrical,
truncate, and bear terminal spines.
G. distincta Lindgren.
B, Without large sphaerasters with very large centrum and very short
rays.

GEODIA. 299

A; The dermal asters small, usually 5-6, the largest never over

8.3 in diameter. Regular or irregular (ataxastrose).
G. erinaceus (Lendenfeld). G. variospiculosa Thiele.
var. typica Lendenfeld var. clavigera Thiele. var.
intermedia Lendenfeld. var. micraster Lendenfeld.
G. reniformis Thiele. G. japonica (Sollas). G. cook-
soni (Sollas). G. hilgendorfi Thiele. var. typica
Lendenfeld var. granosa Thiele. G. ataxastra Lenden-

feld.
B,; The dermal asters large, usually 8-12, the largest never under

8 # in diameter. Usually regular.
Ag Sterrasters over 300 long.
G. hirsuta (Sollas).
B, Sterrasters up to 70-125 yp long.
A, The large choanosomal rhabds are chiefly amphioxes.
G. mesotriaena Lendenfeld. var. pachana Lendenfeld.
var. megana Lendenfeld. var. microana Lendenfeld.
G. agassizii Lendenfeld. G. mesotriaenella Lendenfeld.
G. breviana Lendenfeld. G. ovis Lendenfeld. G.
sphaeroides (Kieschnick). G. micropora Lendenfeld.
G. berryi (Sollas). G. kiikenthali Thiele.
B, The large choanosomal rhabds are chiefly amphistron-
gyles.
G. amphistrongyla Lendenfeld.
C, Sterrasters under 60 y long.
G. lophotriaena Lendenfeld.
B, The large choanosomal asters are acanthtylasters.
G. acanthtylastra Lendenfeld.

B, Regularly triaene promesoclades (proclades) absent. Promesoclades (pro-
clades) with more or less reduced cladomes (only two or one clades) some-
times present.

G. nigra Lendenfeld. G. media Bowerbank. G.
magellani (Sollas). G. exigua Thiele. G. incon-

spicua (Bowerbank).

230 GEODIA EOSASTER.

Geodia nux (SELENKA).
Stelletta nux SELENKA, Zeitschr. wiss. zool., 1867, 17, p. 569, plate 35, figs. 11-13.
Cydonium nux Sous, Rept. voy. “Challenger,” 1888, 25, p. 260.

Spherical with wart-shaped protuberances. In spirit: brown on the surface,
yellowish in the interior.

Amphistrongyles: 1.83 mm. by 38.7 yp. Dichotriaenes: (Sollas, 1888).
Plagiotriaenes (Selenka, 1867) clades short.

Large oxysphaerasters: rays numerous, conical smooth; centrum 51.6 4,
whole aster 96 “in diameter. Small tylosphaeraster (Sollas, 1888): total diameter
16 4. Sterrasters (Sollas, 1888): 90 by 77.4 p.

This is a very doubtful species. Ridley ’ was inclined to consider it as a
monaxonid (Tethya, that is, Donatia), while I did not include it in my synopsis
of the Tetraxonia (Tierreich, 1903, 19). Of course, if all the spicules found by
Sollas in the spicule-preparation examined by him, really belong to it, it is a
geodine tetraxonid, but as one frequently finds foreign spicules in such prepa-
rations, and as Selenka himself does not mention sterrasters, the status of this
sponge must remain doubtful. I therefore place Geodia nux here with all
reserve.

Western Pacific. Samoa Islands.

Geodia eosaster (SoLAs).

LENDENFELD, Tierreich, 1903, 19, p. 110.
Cydonium eosaster Souuas, Rept. voy. “Challenger,” 1888, 25, p. 225, plate 20, fig. 22, plate 21, figs. 15-29.
Non Geodia eosaster TOPSENT, 1904.

Spherical. In spirit: yellowish white.

Large choanosomal amphioxes: 2.856 mm. by 32 p. Minute dermal am-
phioxes: 250-300 by 3.5 4. Dichotriaenes: rhabdome 3.57 mm. by 47 »; main
clades 110 4, end clades 210 wlong. Protriaenes (? mesoprotriaenes): rhabdome
5 mm. by 19 «at the cladome, and 26 at the thickest point near the middle;
clades 190 # long; cladome 190 » broad. Anatriaenes: rhabdome 8.21 mm.
by 29 p; clades 95 long; cladome 190 » broad and 48 y high.

Large oxyasters: four to numerous rays; total diameter 27.6-39 ». Large
oxysphaerasters: rays exceedingly short and broad, appearing as low conical
protuberances of the very large centrum; total diameter 19.8-31 p. Small
strongylosphaerasters: rays conical or cylindrical, truneate; total diameter 10 p.
Sterrasters: spheroidal; 64-70 in diameter.

‘'S. O. Ridley. Spongiida. Rept. voy. “Alert,” 1884, p. 472, foot-note.

GEODIA GLOBOSTELLA. 231

Topsent ' has identified a number of sponges from the Agores as Geodia
eosaster (Sollas). Since, however, the rays of the large strongylosphaerasters
with very large centrum are in these sponges short, thick, cylindrical, and spined
on their terminal face, and since they possess, besides the dichotriaenes, also
orthotriaenes, I hardly think that Topsent’s identification is correct.

Southwestern Pacific. East coast of Australia; Port Jackson.

Geodia globostella, nom. nov.

Geodia globostellifera Ripuey, Rept. voy. “ Alert.,” 1884, p. 480, plate 43, fig. b.
Non Geodia globostellifera CARTER, 1880.

In spirit: gray with a crimson tinge in places.

Large choanosomal amphioxes: 3 mm. by 38 p. Minute dermal amphioxes:
160 by 5 w. Orthotriaenes: rhabdome 70 » thick; cladome 580 y broad. Pro-
triaenes: rhabdome 16 » thick; clades 1 mm. long.

Large choanosomal oxyasters: rays few in number, often curved; total
diameter 38 ». Large oxysphaerasters: rays conical, very short and stout; cen-
trum very large; total diameter 28 y. Small strongylasters: rays numerous;
total diameter 6.3 4. Sterrasters: 90 long.

In 1880, H. J. Carter described * a sponge as Geodia globostellifera (globo-
stellata *) from the Gulf of Manaar. In 1884, 8. O. Ridley * had occasion to study
a sponge from Port Darwin, which, although distinguished from G. globostellifera
Carter, 1880, by the presence of minute dermal amphioxes, by the much greater
thickness of the orthotriaene-rhabdomes, the much greater length of the pro-
triaene-clades and the larger size of the spicules generally, particularly the small
strongylasters, he assigned to this species of Carter. Sollas* and myself °
doubted the correctness of this identification, and in consideration of the differ-
ences between the specimens of Carter and of Ridley, I think it advisable to
distinguish them specifically.

Southwestern Pacific. North Australia; Port Darwin.

1B. Topsent. Spongiairs des Agores. Result Monaco, 1904, 25, p. 67, plate 4, fig. 7; plate 9, fig. 5.

2 H. J. Carter. Report on specimens * * * from the Gulf of Manaar. Ann. mag. nat. hist., 1880,
ser. 5, 6, p. 134 (Sep., p. 488), plate 6, fig. 38a—h (err. f).

3H. J. Carter. Loc. cit., p. 154 (Sep., p. 508).

‘S$. O. Ridley. Spongiida. Rept. voy. “Alert,” 1884, p. 480, plate 43, fig. b.

5 W.J.Sollas. Tetractinellida. Rept. voy. “Challenger,” 1888, 25, p. 261.
5 R. v. Lendenfeld. Tetraxonia. Tierreich, 1903, 19, p. 111.

932 GEODIA ERINACEUS.

Geodia distincta LinDGREN.

Zool. anz., 1897, 11, p. 486. Zool. jahrb. Syst., 1898, 11, p. 343, plate 17, fig. 15, plate 18, fig. 19,
plate 20, fig. 3a-k, a’, d’. LENDENFELD, Tierreich, 1903, 19, p. 111.

Tuberous.

Large choanosomal amphioxes: 1.5-1.8 mm. by 32 yp. Minute dermal
amphioxes: 290 by 12 4. Orthotriaenes: rhabdome 1.8-2.5 mm. by 48 4; clades
concave to the rhabdome, 240 long, sometimes one or two bifureate. Pro-
triaenes (probably mesoprotriaenes): rhabdome 2.33-3 mm. by 12 p; clades
135 « long; cladome 120 y broad and 130 » high. Anatriaenes: rhabdome
3.4mm. by 12 p; clades 72 pw long; cladome 84 # broad and 60 » high.

Large choanosomal oxyasters: rays blunt, roughened; centrum small;
total diameter 44 4 Small subcortical oxyasters: rays spined; centrum 4 p, whole
aster 16 y, in diameter. Large strongylosphaerasters: rays numerous, short, and
thick, with spines on their terminal faces; centrum very large; total diameter
28 yw. Small strongylosphaerasters: centrum 2.5 4, whole aster 8 4, in diameter.
Sterrasters: spheroidal, 68 by 56 yp.

Western Pacific. Java Sea; Java.

Geodia erinaceus (LENDENFELD).

Tierreich, 1903, 19, p. 107.
Cydonium erinaceus LENDENFELD, Descriptive catalogue sponges Australian museum, 1888, p. 36.

The following description is based on an examination of part of the type
specimen in the British Museum.

Massive flattened, margin mostly lobose, or digitaté; up to 20 em. long.
In spirit: dirty white to brown. Spicule-fur chiefly of large orthotriaenes.

Large choanosomal rhabds: attenuated towards both ends, one or both ends
usually rounded; 3.5-4.6 mm. (Ldf., 1888, 2 mm.) by 23-49 ». Minute dermal
rhabds: mostly amphiox; 200-300 (Ldf., 1888, 320 #) by 10-12 ». Orthotriaenes:
rhabdome 3.5-4.2 mm. (Ldf., 1888, 1-3 mm.) by 40-50 »; clades concave to the
rhabdome throughout their whole length, 400-550 long; clade-angles 90-100°.
Mesoproclades: rare; rhabdome 4.8-5.2 mm. by 10-20 4; clades usually irregu-
lar and one or two often reduced, fully developed ones 50-60 long; clade-
angles 49-56°; epirhabd 30-70 » long (not mentioned by Ldf., 1888). Large
anatriaenes: rare; rhabdome 13 thick; clades 30 » long; clade-angles 63°
(not mentioned by Ldf., 1888). Minute dermal anatriaenes: rhabdome termi-
nally rounded, about 280 by 2-3.5 at the cladome, and 3-5.5 vein the middle;
clades 6-9 long; clade-angles 39-54° (not mentioned by Ldf., 1888).

GEODIA VARIOSPICULOSA. 233

Large oxyasters: from five to fourteen, conical, in small ones smooth, in large
ones spined, rays, 1-1.5 » thick; total diameter 16-30 4. Large oxysphaerasters:
connected with the former by transitions; from fifteen to twenty rays, 1.2-3 »
thick, with sparse, stout spines near their end; centrum 3-8 y, whole aster 22-28
pin diameter. Small strongylosphaerasters: from eight to twenty, usually equal,
rarely unequal, truncate or rounded, spined rays, 0.8-1.3 » thick; centrum 1.6-—
3 #, whole aster 5.5-7.3 » (Ldf., 1888, 3-5 #), in diameter. Sterrasters: 140-160
by 127-144 by 100-108 yp.

Southwestern Pacific. East coast of Australia.

Geodia variospiculosa THIELE.

Zoologica, 1898, 24, p. 10, plate 6a-l, 7a, b. LENDENFELD, Tierreich, 1903, 19, p. 107. Ante, p. 55.

Massive tuberous. In spirit: dirty white to light brown.

Large choanosomal amphioxes: 1-3.9 mm. by 20-50 yp. Large choanosomal
tylostyles or styles: 1.35 mm. by 25-50 yp; tyle 380-70 » (observed by me in vars.
intermedia and micraster, not mentioned by Thiele, 1898, in vars. typica and clavi-
gera). Large dermal tylostyles: 1.6 mm. by 11 p, tyle 18 » (present only in var.
clavigera). Minute dermal styles: 200-320 by 3-7 yp. Orthoplagiotriaenes:
rhabdome 1.25-3 mm. by 30-65 » at cladome and 43-70 y at thickest point, a
little below; clades 220-760 » long; clade-angles 99-111°. Dichotriaenes:
not numerous; rhabdome 1.25-2.6 mm. by 30-75 y at cladome and 45-90
at thickest point, a little below; main clades 150-340, end clades 140-400
long; main clade-angles 90° or a little over. Mesoprotriaenes: rhabdome 2.5—
3.2 mm. by 7-20 ; clades 60-220 » long; clade-angles 30-63°; epirhabd 25—-
95 w long. Large anatriaenes: rhabdome 3.6-5.2 mm. by 12-46 p; clades 30-
180 plong; clade-angles 22-70°. Minute dermal anaclades; rhabdome 205-560
by 1-4 pat cladome, and 2-7.5 p at thickest point below the middle; clades 3-
13 w long; clade-angles 38—54°.

Large choanosomal oxyasters: from one to eleven conical, distally spined rays,
9-135 p» long, 1-8 » thick; total diameter in var. intermedia and var. micraster
17-180 p. Oxysphaerasters: from fourteen to twenty-two rays, 1-2 yp thick;
centrum 5-6 y, whole aster 14-30 p, in diameter. Small strongylosphaerasters:
from ten to nineteen equal or unequal rays 0.5-2 p thick; centrum 2.4 4, whole
aster 5-8 p, in diameter. Sterrasters: 80-133 p long, 65-116 broad, in vars.
intermedia and micraster 70-90 y« thick.

Northwestern Pacific. Japan, off Honshu Island. “Albatross” Stations
3746, 3758; westward of Yogashima.

234 GEODIA RENIFORMIS.

Geodia variospiculosa var. typica LeENDENFELD.

Geodia variospiculosa THIELE, Zoologica, 1898, 24, p. 10, plate 6, fig. 6.

Without dermal tylostyles. Orthoplagiotriaene-clades 400-460 y long.
Anatriaene-clades up to 180” long. Oxyaster-rays up to 135 w long. Sterr-
asters up to 115 u long.

Northwestern Pacific. Japan; westward of Yogashima.

Geodia variospiculosa var. clavigera THisLe.

Zoologica, 1898, 24, p. 11, plate 6, fig. 7a—b.

With large dermal tylostyles. Orthoplagiotriaene-clades 250-300 y long.
Anatriaene-clades only 30-40 long, one or two frequently absent (anadiaenes,
anamonaenes). Sterrasters 100 long.

Northwestern Pacific. Japan.

Geodia variospiculosa var. intermedia LenpeENFELD

Ante, p. 55.

Without large dermal tylostyles. Orthoplagiotriaene-clades 220-550 yp
long. Anatriaene-clades up to 135 » long. Oxyaster-rays up to 90 # long.
Sterrasters up to 125 y long.

Northwestern Pacific. Japan; off Honshu Island. “ Albatross” Station
3746.

Geodia variospiculosa var. micraster LenpeNFELD.

Ante, p. 55.

Without large, dermal tylostyles. Orthoplagiotriaene-clades 240-760 yp
long. Anatriaene-clades up to 130 « long. Oxyaster-rays up to 72 y long.
Sterrasters up to 133 » long.

Northwestern Pacific. Japan; off Honshu Island. ‘ Albatross” Station
3798.

Geodia reniformis Tu 1e.e.

Zoologica, 1898, 24, p. 9, plate 1, fig. 3, plate 6, fig. 5a-h. LeNDENFELD, Tierreich, 1903, 19, p. 108.

Kidney shaped. Dry: light brown.

Large chounosomal amphioxes: 3.3 mm. by 45 4. Minute dermal amphiozes:
170 long. Orthotriaenes: rhabdome 2.8 mm. by 90 4; clades 500-600 « long.
Mesoproclades: mostly triaene, rarely diaene; rhabdome 1.5 mm. by 25 ; clades

GEODIA JAPONICA. 235

70 w, epirhabd 30-40 y long. Anaclades: mostly triaene, rarely diaene or mon-
aene; rhabdome 4-4.5 mm. long, clades 50 y long.

Large oxyasters: rays few, 40-70 y long. Small oxyasters: rays more
numerous, 15-20 long. Oxysphaerasters: rays stout and short; total diameter
12 ~ Small strongylosphaerasters: regular or irregular (ataxastrose); total
diameter 5 yu. Sterrasters: 130 by 113 yp.

Northwestern Pacific. Japan; Enoshima.

Geodia japonica (Souuas).

THIELE, Zoologica, 1898, 24, p. 7, plate 2, fig. 1, plate 6, fig. 3. LeNDENFeLD, Tierreich, 1903, 19, p. 111;
Ante, p. 72.

Cydonium japonicum SOLAS, Rept. voy. “Challengér,”’ 1888, 25, p. 256.

The following description is also based on an examination of part of the
type specimen in the British Museum.

Cup shaped, outside lobose, large, up to nearly 50 em. high. Dry: white.

Stout choanosomal amphioxes: 2-3.3 mm. by 30-51 . Large styles: 2.1-
2.8 mm. by 40-43 y in the middle and 10-31 y at the rounded end. Slender
amphioxes: 1—2.2 mm. by 12-22 p. Minute dermal rhabds: mostly amphiox,
rarely style; 195-280 by 3.5-7 yw; often irregularly curved. Orthoplagio-
triaenes: rhabdome 2.3-3.2 mm. by 50-85 4; clades 180-380 y long, distal part
straight; clade-angles 90-102°. Mesoproclades: rhabdome 2.8—4.3 mm. by 11-
21 w; clades 65-125 » long; clade-angles 22-48°; epirhabd 40-105 y long (not
mentioned by Sollas, 1888, and Thiele, 1898). Large anaclades: mostly triaene,
rarely monaene; rhabdome 2.4—5 mm. by 8-23 »; clades 70-130 » long; clade-
angles 23-45°. Munute dermal anaclades: triaene, diaene, or monaene; rhabd-
ome 235-310 by 1-2 wat the cladome, and 2.8-5 » in the middle; clades 3-10
pe long; clade-angles 30-54° (not mentioned by Sollas, 1888, and Thiele, 1898).

Large oxyasters: from three to seven straight, conical, blunt, spined rays,
1.2-2.8 » thick; total diameter 21-46 » (Thiele, 1898, gives the ray-length as
6-14 pw). Oxysphaerasters: from fifteen to twenty-one straight, conical, pointed,
spined rays 1.4-2 y» thick; centrum 5-7.5 », whole aster 15-22 y in diameter.
Small strongylosphaerasters: mostly regular, rarely irregular (ataxastrose);
from six to twenty-two truncate or terminally rounded, spined rays, 0.5-1.3 u
thick; centrum 1.2-5 in diameter; the regular forms 4-6 y, the ataxastrose ones
5.8-7.3 in total diameter. Sterrasters: 80-92 by 65-80 by 55-61 yp (Thiele,
1898, 75 by 65 yp).

Northwestern Pacific. Japan; (Thiele’s specimen, near Enoshima).

GEODIA HILGENDORFI.

bo
oo
co

Geodia cooksoni (Souas).
LENDENFELD, Tierreich, 1903, 19, p. 115.
Cydonium cooksoni Souuas, Rept. voy. “Challenger,” 1888, 26, p. 255.

The following description is based on an examination of part of the type
specimen in the British Museum.

Probably lobose. In spirit: light brown.

Choanosomal amphioxes: 1.2-2 mm., mostly 1.7-1.9 mm. by 18-36, mostly
25-32 pe (Sollas, 1888, 41 2). Minute dermal rhabds: mostly amphiox, rarely
with one or both ends blunt; 150-190 y (Sollas, 1888, 129 2) by 2-5 yp. Ortho-
plagiotriaenes: rhabdome 1.2-1.9 mm. by 40-60 y (Sollas, 1888, 64.5 y); clades
coneave to the rhabdome throughout their length, 200-340 » long; clade-angles
88-1042, on an average 97°. Mesoproclades (Sollas, 1888, protriaenes): rhabd-
ome 1.6-2 mm. (Sollas, 1888, 2.38 mm.) by 7-13 y (Sollas, 1888, error, 75 4)
at cladome and 10-18 jin the middle; rarely three fully developed clades, usually
one, two, or all three reduced or absent altogether, fully developed clades 27—
45 wlong; clade-angles 34-64°, on an average 44°; epirhabd 30-70 p long.

Large choanosomal oxyasters, with from two to thirteen usually strongly
spined, straight, conical rays, 1.6-2.3 y thick at the base; centrum small or ab-
sent; total diameter 22-42 » (not mentioned by Sollas, 1888). Sollas, 1888,
describes chiasters 19.7 # in diameter with slender, cylindrical, truncate rays,
which were not observed by me. Ozysphaerasters: with from eleven to thirty
conical, distally sparsely spined rays, 0.5-1.6 » thick at the base; centrum
1.3-3 y, whole aster 12-18 yp (Sollas, 1888, up to 19 ) in diameter. Strongylo-
sphaerasters: with from ten to thirty cylindrical or cylindroconical, truncate rays,
0.5-1.1 » thick; centrum 1.5-2.5 ys, whole aster 4-6.5 p (Sollas, 1888, 1-6 #4), in
diameter. Sterrasters: flattened ellipsoids, 75-80 pp by 70-75 p» by 57-60 yp
(Sollas, 1888, 77.4 by 66 #).

Eastern Pacific. Galapagos; Charles Island.

Geodia hilgendorfi Tuieue.
Zoologica, 1898, 24, p. 8, plate 1, fig. 4, plate 6, fig. 4a—k. LenDENFELD, Tierreich, 1903, 19, p. 112.

Massive elongate, with broad and low lobose protuberances. Dry; whitish.

Large amphioxes: 1.2-1.6 mm. long. Minute dermal amphioxes: blunt,
140-180 » long. Orthoplagiotriaenes: rhabdome 1.6-1.7 mm. long sometimes
reduced and rounded; clades 250-300 long; cladome sometimes very irregular.
Mesoprotriaenes: rhabdome 1.2-1.6 mm. long; clades as long as or longer than
the epirhabd. Anatriaenes.

GEODIA ATAXASTRA 237

Large choanosomal oxyasters: with or without centrum; total diameter
15-40 ». Subcortical oxysphaerasters: rays short and stout, centrum very large;
total diameter 12 yw. Irregular sphaerasters (ataxasters): rays very short;
total diameter 3-5 yx. Sterrasters: 53-80 by 45-60 pu.

Northwestern Pacific. Japan.

Geodia hilgendorfi var. typica, var. nov.

Geodia hilgen lorfi THIELE, Zoologica, 1898, 24, p. 8, plate 1, fig. 4, plate 6, fig. 4 a—-h.

Oxyasters: with centrum, 15-20 # in diameter. Without particularly small,
irregular sphaerasters in the interior. Sterrasters: 80 by 60 p.
Northwestern Pacific. Japan; probably from the vicinity of Enoshima.

Geodia hilgendorfi var. granosa THIELE.

Zoologica, 1898, 24, p. 9, plate 6, fig. 4 i, k.

Oxyasters: without centrum, 40 # in diameter. Small irregular sphaer-
asters only 3 yin diameter in the interior. Sterrasters: 53 by 45 pu.
Northwestern Pacific. Japan; probably from the vicinity of Enoshima.

Geodia ataxastra LENDENFELD.

Ante, p. 79.

Spherical or irregularly massive, tuberous or lobose. In spirit: white to
lilac-gray.

Large choanosomal rhabds: mostly amphioxes, but also some amphistron-
gyles and styles; 0.6-2.8 mm. by 12-43 4. Minute dermal rhabds: mostly am-
phioxes, but also some styles, 120-215 by 3-7 yx. Orthoplagiotriaenes: rhabdome
1.3-2.3 mm. by 29-70 y; clades 130-290 » long; clade-angles 85-116°. Meso-
proclades: mostly triaene; rhabdome 1.6-3.4 mm. by 7-10 y; clades 25-80 yu
long; clade-angles 25-58°; epirhabd 28-73 » long. Large anatriaenes: rhabd-
ome: 2-3.1 mm. by 3-12 y; clades 17-68 y long; clade-angles 20-55°.
Minute dermal anaclades: observed only in var. angustana; rhabdome 190-
340 by 0.5-2 pw; clades 2-6 p» long; clade-angles 33-57°.

Large oxyasters: from two.to eleven rays, 0.6—2.6 y thick; total diameter
15-50 p, in inverse proportion to the ray-number. Oxysphaerasters with slender
rays: from eighteen to twenty-eight conical, pointed, distally spined rays 0.7-
1.3 u thick; centrum 2.4-5.5 y, whole aster 8-14.4 y, in diameter. Oxysphaerasters
with smooth, thick rays: observed only in var. latana; about eighteen conical,
blunt, smooth rays, 2 4 thick; centrum about 4.5 #4, whole aster about 13 y, in

238 GEODIA ATAXASTRA VAR. LATANA.

diameter. Small strongylosphaerasters: from seven to twenty straight or conical,
distally often thickened rays, 0.2-0.8 1 thick; centrum 0.6-3 y, whole aster 2.6-
6.4 w, in diameter. Ataxasters: from one to eight cylindrical, or cylindroconical,
truncate, simple or rarely bifureate, rough or spined rays, 0.3-2.8 « long and
0.4-1.5 » thick; the rays are very irregularly distributed and often very un-
equal in size; they arise from a spherical or irregularly tuberous centrum 1.4—
4.5 in diameter; total diameter 4-8.3 4; connected by transitions with the
small strongylosphaerasters. Acanthtylasters: rare, from ten to fifteen cylin-
drical, simple or rarely branched rays, 1-2 » thick; their distal ends bear dense
clusters of large divergent spines and appear thickened; no centrum; total
diameter 8-16 y. Irregular sterraster-derivates (sterroids): observed only in
var. latana; a simple or lobose central mass with extensive tufts of ray-like
spines 4-8 » long; total diameter 21-50 yp. Sterrasters: 55-78 by 50-67 by
47-o7 pL.

Eastern Pacific. Gulf of Panama; Perico Island.

Geodia ataxastra var. angustana LENDENFELD.

Ante, p. 79.

Generally whitish. Large amphioxes up to 2.8 mm. long; orthoplagio-
triaene-rhabdomes up to 2.8mm. by 70 #. Average anatriaene-clade angle 34°.
Minute dermal anaclades present, their rhabdomes 190-340 4 long. Two-rayed
oxyasters present, 40-50 » long, the three- to seven-rayed 21—40 » in diameter.
Oxysphaerasters with stout smooth rays and sterroids not observed. Sterras-
ters up to 78 long.

Eastern Pacific. Gulf of Panama; Perico Island.

Geodia ataxastra var. latana LenpeENFELD.

Ante, p. 79.

Lilac-gray. Large amphioxes up to 1.9 mm. long; orthoplagiotriaene
rhabdomes up to 1.7 mm. by 45. Average anatriaene-clade angle 47°. Minute
dermal anaclades not observed. Two-rayed oxyasters not observed, the three-
to seven-rayed 17-28 » in diameter. Oxysphaerasters with stout, smooth rays
present, 13 in diameter. Sterroids present, 21-50 in diameter. Sterrasters
up to 65 » long.

Eastern Pacific. Gulf of Panama; Perico Island.

GEODIA MESOTRIAENA. 239

Geodia hirsuta (Souuas).

LENDENFELD, Tierreich, 1903, 19, p. 106.
Cydonium hirsutus Souuas, Proce. Roy. Dublin soe., 1886, 5, p. 197. Rept. voy. “Challenger,” 1888,
25, p. 218, plate 21, figs. 30-42.

Irregular, lobose with digitate processes. In spirit: grayish white.

Amphioxes: shorter and stouter, 4.462 mm. by 60 y, and longer and more
slender, 9mm. by 31.6 4. Tylostyles: with large spherical tyle. Dichotriaenes:
rhabdome over 4.46 mm. long, 84 thick, attenuated at first very rapidly;
main clades 127 , end clades 350 » long. Protriaenes (? mesoprotriaenes)
rhabdome at cladome 20 #, at thickest point 29 » thick; clades 127 » long;
cladome 143 ys broad and 99 yp high. Anatriaenes: rhabdome long, 18 1
thick; clades 36 long.

Choanosomal oxyasters: rays pointed or truncate; total diameter 19.7 py.
Cortical oxysphaerasters: rays spined; centrum 12 4, whole aster 32 y, in diam-
eter. Strongylosphaerasters: total diameter 11.8 . Sterrasters: 306 by 245 by
161 y, outline oval or somewhat hexagonal.

Central Pacific. Ki Islands; 5° 49’ 15”S., 132° 14’ 15” W.

Geodia mesotriaena LeNDENFELD.

Ante, p. 96.

Massive, cake shaped, horizontally expanded, large, up to 23 em. in maxi-
mum diameter, with praeoscular cavities, scarce in the smaller, numerous in
the larger specimens. Dry and in spirit: yellow to brown.

Large choanosomal amphioxes: 4.3-8.2 mm. by 50-105 p. Large styles: rare;
3-4 mm. by 70-110 ». Minute dermal amphioxes and styles: 380-680 by 9-19 pu.
Orthoplagiotriaenes: rhabdome 4.6—7.2 mm. by 85-120 1; clades 200-670 y long,
concave to the rhabdome proximally, straight or slightly curved in the opposite
direction distally; clade-angles 85-117°. Mesoprotriaenes: rhabdome 6-14
mm. by 15-40 » at the cladome, and 38-70 y at the thickest point near the
middle; clades 90-310 » long; clade-angles 29-56°; epirhabd 95-330 y long.
Anatriaenes: rhabdome 11-16 mm. by 8-40 y; clades 70-270 y long, clade-
angles 34-58°. Mesanaclade anatriaene-derivates: rare; dimensions as in the
regular anatriaenes.

Large oxyasters: from five to fifteen rays, conical throughout or nearly cylin-
drical at the base, distally spined, 1-4 thick; total diameter 19-54 Small
oxyasters: from eight to twenty rays 0.9-3 y thick; total diameter 11-20 y.
Large oxysphaerasters: from fifteen to twenty-five very spiny rays, 1-3 y thick;

940 GEODIA MESOTRIAENA VAR. MICROANA.

centrum 3-10 4, whole aster 19-32 , in diameter. Small strongylosphaerasters:
from six to twenty distally spined rays, 0.5-2.5 y thick; centrum 2-6, whole
asters 6-14.5 4, in diameter. Large strongylosphaerasters: transitional to sterr-
oids; rare; 16-33 y in diameter. Sterroids: rare; observed only in var.
meguna, numerous rays 4-6 # thick, with spined terminal face; 39-58 pin total
diameter. Sterrasters: 92-125 by 78-107 by 67-82 yu.

Northeastern Pacific. West coast of North America; from 33° 38’ 45”
to 34° 22’N. ‘Albatross’ Stations 2909, 2942, 2958.

Geodia mesotriaena var. pachana LENDENFELD.

Ante, p. 96.

Large amphioxes up to 105 » thick. Among the minute dermal rhabds
styles are numerous. Average angle of orthoplagiotriaene-clades 99.4°.
Rhabdome of mesoprotriaenes and anatriaenes up to 40 y thick. Anatriaene-
clades thick and short, up to 170 » long. Oxyasters up to 37 p in diameter.
True sterroids absent. Average proportion of length to breadth of sterrasters
100: 91.

Northeastern Pacific. West coast of North America; 34° 22’N. ‘‘Alba-
tross”’ Station 2909.

Geodia mesotriaena var. megana LENDENFELD.

Ante, p. 96.

Large amphioxes up to 105 y thick. Styles among the minute dermal
rhabds rare. Average angles of orthoplagiotriaene-clades 91.9°. Mesopro-
triaene-rhabdomes up to 40 4, anatriaene-rhabdomes up to 38 yp thick. Ana-
triaene-clades thick and up to 270 » long. Oxyasters up to 54 p in diameter.
Sterroids in small numbers present, 39-58 « in diameter. Average propor-
tion of length to breadth of sterrasters 100 : 90.

Northeastern Pacific. West coast of North America; 34° 4’ N. ‘“‘Alba-
tross” Station 2958.

Geodia mesotriaena var. microana LENDENFELD.

Ante, p. 96.

sarge amphioxes up to 77 » thick. Styles among the minute dermal rhabds
rare. Average orthoplagiotriaene clade-angle 104.2°. _ Mesoprotriaene-rhabd-
omes up to 25 p» thick. Anatriaene-clades thin and short, up to 175 yu long.
Large oxyasters up to 42 in diameter. True sterroids absent. Average pro-
portion of length to breadth of sterrasters 100 : 79.

GEODIA AGASSIZII. 241

Northeastern Pacific. West coast of North America; 33° 38’ 45” N.
“Albatross” Station 2942.

Geodia agassizii LeNnDENFELD.
Ante, p. 113.

Cydonium miilleri LaMBE (non Fleming), Trans. Roy. soe. Canada, 1893, 11, p. 36, plate 4, fig. 2.

The following description is also based on an examination of the type speci-
men of Cydonium miilleri Lambe 1893, in the collection of the Geological Survey
of Canada.

Massive, spherical, oval, elongate or somewhat irregular; without praeos-
cular cavities. In spirit: uniformly white to light brown, some with dark brown
patches, some dark blue.

Large amphiosxes: of adult 2.3-4.8 mm. by 60-112 y; of immature 1.8—
3.4 by 20-66 py. Large amphistrongyles: only in the immature, a little shorter
than the amphioxes, 55 » thick. Large, slender styles: rather rare, not in all
specimens; of adult 1.5-3.4 mm. by 60-110 ». Large, stout styles: rare, only
in some specimens; of adult at thickened rounded end 115-145 thick. Minute
dermal amphioxes: of adult 160-480 by 5-12 »; of immature 180-480 by 3-8 yp.
Orthoplagiotriuenes of adult: rhabdome 1.5-4.3 mm. by 65-150 p; clades 240—
560 » long, simple, more rarely with irregular branchlets; clade-angles 73-117°;
of immature: rhabdome 2-3.5 mm. by 50-100 ; clades 300-500 » long; clade-
angles 88-108°. Regular dichotriaenes: only in the smallest immature specimen ,
rhabdome 1-2.2 mm. by 50-75 w; main clades 150-300, end clades 30-130 yu
long; angles of main clades 109-112°; cladome 350-700 » broad. Mesortho-
triaenes: rare, not in all specimens; in the adult, a style-like shaft 1.8-3 mm.
by 78-164 u at the rounded end, with clades 78-300 » long, inserted near the
rounded end and concave towards it. Irregular amphiclade orthoplagiotriaene,
derivates: rare, not in all specimens, dimensions as in the orthoplagiotriaenes,
but with a branch-ray on the rhabdome, besides the ordinary clades. Mesopro-
triaenes of adult: rhabdome 2-6 mm. by 7-40 ; clades 60-250 yp, epirhabd
25-320 long; clade-epirhabd angles 22-55°; of immature: rhabdome 9-20 p.
thick; clades 60-125 , epirhabd 70-100 « long; clade-epirhabd angles 36—47°
Large anatriaenes of adult: * rhabdome 4-9 mm. by 10-50 yw; clades 40-155 yp
long, clade-angles 32-65°; of immature: rhabdome 3.3-4.7 mm. by 18-28 yp;
clades 45-110 long, clade-angles 31-52°. Irregular anatriaene-derivates:
rare, not in all specimens; dimensions as in the regular anatriaenes. Minute
dermal anaclades: rare, observed only in the smallest immature specimen;

242 GEODIA MESOTRIAENELLA.

rhabdome 290 « by 1-1.5 at the cladome and 3-5 at its thickest point below
the middle; clades 4-6 long; clade-angles 38-62°; possibly foreign.

Large oxyasters of adult: from four to sixteen rays, 0.8-3.2 p thick, distally
spined, pointed to truncate; centrum small; total diameter 9-31 «4; of immature:
from seven to fourteen rays 0.8-2.3 y thick; centrum small; total diameter 13-—
25 yw. Large oxysphaerasters of adult: in some specimens rare; from fourteen to
twenty-eight or more conical, pointed or blunt, distally spined rays, 1-2 thick;
centrum 3.5-11 #, whole aster 10-21 y, in diameter; of immature: from ten to
thirty rays, 0.9-2 » thick; centrum 2.7-7 , whole aster 8-21 y, in diameter.
Small strongylosphaerasters of adult: six to twenty rays, 0.6-1.6 « thick; centrum
1.5-6 », whole aster 3.5-11 in diameter; of immature: from ten to twenty-eight
rays, 0.5-1 » thick; centrum 2-3.5 », whole aster 5-9 «, in diameter. Sterrasters
of adult: 82-118 by 75-100 by 58-83 4; of immature: 76-110 by 70-92 by 60-—
75 ww. Slerroids: rare; similar to, but mostly somewhat smaller than, the sterr-
asters, and with thicker rays; their strongylaster-like young stages only in
the immature specimens.

Northeastern Pacific. West coast of North America; from 33° 59’ 45” N.
to the vicinity of Naha Bay, Behm Canal, 8. E. Alaska. ‘‘Albatross” Stations
2886, 2887, 2978, 3088, 3168, 4193, 4199, 4228, 4551: Queen Charlotte Island,
Houston Stewart Channel.

Geodia mesotriaenella LeNDENFELD.

Ante, p. 151.

Nearly spherical. In spirit: dirty white.

Large amphioxes: 2-2.6 mm. by 20-50 yp. Minute dermal rhabds: mostly
styles; 196-260 by 4-5 ». Orthotriaenes: rhabdome 2.1-2.4 mm. by 75-120 p;
clades concave to the rhabdome, 350-600 long; mostly simple, rarely bifureate;
clade-angles 90-96°. Mesoprotriaenes: rhabdome 2.8-3.4 mm. by 9-19 yp at
the cladome, thicker in the middle; clades 100-220 » long; clade-angles 30-
47°; epirhabd 70-165 y» long. Anatriaenes: rhabdome 3.7 mm. by 18-30 p;
clades usually simple, rarely bifureate, 87-140 » long; clade-angles 41-479.

Large oxyasters: from five to eleven conical, blunt, distally spined rays, 1.5-
2.8 « thick; centrum small; total diameter usually 17-26 yp, rarely up to 40 p,
these large asters perhaps foreign. Large oxysphaerasters: from fifteen to
twenty-three conical, pointed rays with a few spines, 2-2.5 » thick; centrum 6-
9 », whole aster 20-21 4, in diameter. Small strongylosphaerasters: from three to
twenty-five, usually from ten to seventeen rays, 0.5—-1 » thick; centrum 1.6-4.5 p,

GEODIA BREVIANA. 243

whole aster 6-11 y, in diameter. Sterrasters: 87-107, usually not over 97, by
77-92, usually not more than 86, by 58-69 yu.

Northeastern Pacific. West coast of North America; near Santa Barbara
Islands. ‘Albatross’ Station 4417.

Geodia breviana LeNDENFELD.
Ante, p. 155.
Cydonium miilleri LAMBE (non Fleming), Trans. Roy. soe. Canada, 1893, 10, p. 72, plate 4, fig. 1, plate 6,
fig. 1-la-i.

The following description is also based on an examination of the type
specimen of Cydonium miilleri Lambe, 1893, in the collection of the Geological
Survey of Canada.

Cup shaped with small cavity. In spirit: dirty white; dry: brown.

Large choanosomal amphioxes: 1.8-5 mm. by 30-88 » (Lambe, 1893, 2.77—
3.81 mm. by 80). Minute dermal amphioxes: 280-450 by 2-8.5 4 (Lambe, 1893,
288 by 13 2). Plagioorthotriaenes: rhabdome 1.8-4.1 mm. (Lambe, 1893, 2.4
mm.) by 60-130 y; clades 280-680 « (Lambe 1893, 700 #4) long; clade-angles
94-113°. Mesoprotriaenes: rhabdome 7-11 mm. (Lambe, 1893, 7.84 mm.)
by 15-32 yw; clades 65-250 « (Lambe, 1893, 95 ») long; clade-angles 20-44°.
Large anaclades: mostly triaene, rarely diaene; rhabdome 9-11 mm. (Lambe,
1893, 7.5 mm.) by 25-40 »; clades 47-115 x (Lambe, 1893, 60 ) long; clade-
angles 45-65°. Minute dermal anaclades: generally without, sometimes with
epirhabd (mesanaclades) rhabdome 350-610 by 1-4.5 at the cladome, and 5—
8.6 «at the thickest point below the middle; clades 2-12 4 long; clade-angles
42-60°; epirhabd of the mesanaclades 5-8 » long (not mentioned by Lambe,
1893).

Large thick-rayed oxyasters: from five to twelve usually simple, rarely bi-
fureate, conical, pointed or blunt, distally spined rays, 1-2.3 u thick; centrum
small; total diameter 16-26.5 4 (Lambe, 1893, apparently considers all the eu-
asterforms as oxyasters and gives 3-13 y as their total diameter). Large thin-
rayed oxyasters: rare, not always present, perhaps foreign; nine to fourteen
distally spined rays, 0.25-0.7 « thick; no centrum; total diameter 7-23 p.
Large oxysphaerasters: up to thirty rays with large spines, 1-2.7 » thick; cen-
trum 3-9 », whole aster 12-21.5 , in diameter. Small strongylosphaerasters:
from thirteen to twenty-five distally spined rays, centrum 2-5.5 , whole aster
6-12 , in diameter. Sterrasters: 84-105 by 75-98 by 55-77 p» (Lambe, 1893,
91 yp).

244 GEODIA OVIS.

Northeastern Pacific. West coast of North America; off southern Cali-

fornia. ‘‘Albatross” Station 2894: Vancouver Island, Strait of Georgia, near

Comox.

Geodia ovis LENDENFELD.

Ante, p. 161.

Cake shaped, horizontally extended. With exceedingly high spicule-fur.
In spirit: light brown.

Large choanosomal amphioxes: 4-9 mm. by 30-40 yp. Large styles and
tylostyles: of two kinds, stout and slender; the stout; mostly style; 2.6-4mm. by
85-116 ; the slender: always tylostyle; long; about 40 yu thick; tyle 60-65 p
in diameter. Minute dermal rhabds: mostly amphiox, rarely style; 270-550
by 8-13 4. Orthotriaenes (and plagiotriaenes): rhabdome 5-8 mm. by 74-100 4
at the cladome, 77-110 y a little farther down; clades concave to rhabdome
throughout or only basally, and straight distally, 310-640 » long; clade-angles
86-101°. Irregular megascleres: rare. Mesoproclades and proclades: rhabdome
6-17 mm. by 20-41 pat the cladome, near the middle from two to three times as
thick; in the normally developed mesoprotriaenes the clades 140-170 » long, the
clade-angles about 45°, the epirhabd 110 # long; in the irregular mesoproclades
(proclades) clades very unequal, the longest up to 260 » long, sometimes one or
two clades or the epirhabd suppressed. Anaclades: nearly all triaene; very
unequal in size; small and large ones distinguishable; the small: rhabdome
670 y-2.5 mm. by 2-7 w; clades 6-43 » long; clade-angles 41-65°; the large:
rhabdome up to 23 mm. long by 17-45 y; clades 70-205 » long; clade-angles
36-55°.

Large thin-rayed oxyasters: from three to ten conical, distally spined rays,
1-3.2 « thick; no centrum; total diameter 20-34.5 ». Large thick-rayed oxy-
asters: from four to nineteen conical, simple or partly bifureate rays with large
spines in their middle parts, 3-6.3 » thick; total diameter 28-45 yu; the many-
rayed appear sphaerastrose. Small thick-rayed asters: from six to fifteen trun-
cate or blunt-pointed, distally spined rays, 1-3.2 thick; centrum absent or
present and then up to 6 «in diameter; whole aster 11-24 in diameter. Sterr-
asters: 82-92 by 70-83 by 54-61 . Sterroids: rare; of similar dimensions, but
with much thicker rays.

Northeastern Pacific. West coast of North America; 34° 1/30” N. ‘“Al-
batross” Station 2975.

GEODIA MICROPORA. 245

Geodia sphaeroides (KimscHNick).

TuieELe, Abhandl. Senckenb. gesellsch., 1900, 25, p. 41, plate 2, fig. 14a-k. LenpeNnFevp, Tierreich,
1903, 19, p. 110.

Cydonium sphaeroides Kiescunick, Zool. anz., 1896, 19, p. 529.

Geodia arripiens LinDGREN, Zool. anz., 1897, 20, p. 486. Zool. jahrb. Syst., 1898, 11, p. 346, plate 18,
figs. 10, 18, plate 20, fig. 5a-i, a’, b’, e’, i’.

Spherical or ellipsoidal. Surface brown, interior yellowish or grayish.

Large amphioxes: 1.5-2.4 mm. by 30-40 px. Small dermal amphioxes: 230
by 5 p. Dichotriaenes: rhabdome 2.35-3 mm. by 60-70 p, sometimes reduced
in length and rounded; main clades 120-220, end clades 150-180 long; main
clade-angles 120°. Large anatriaenes: rhabdome 3.3-3.5 mm. by 18-20 yp;
clades 50-60 » long; divergent; cladome 80 # broad and 48 y high. Minute
anatriaenes: dermal or subcortical; rhabdome 340-360 by 2-3 y; clades 8
long; cladome 10 «broad and 6 wzhigh. Protriaenes (? mesoprotriaenes): rhabd-
ome 2.5-3 mm. by 16-20 »; clades 60-80 » long, sometimes one or two reduced;
cladome 80 y broad and 68 y high.

Oxysphaerasters: rays numerous, spined; centrum 6-15 4, whole aster 36—
50 w, in diameter. Small strongylasters (strongylosphaerasters): total diameter
8-15 »; the dermal smaller than the choanosomal. Sferrasters: sphaeroidal ;
39-88 by 72-80 p.

Western Pacific. Ternate: Coast of Cochin China; 11° 5’ N., 108° 50’ E.

Geodia micropora LeNDENFELD.

Ante, p. 170.

Lobose. In spirit: brownish white.

Large choanosomal amphioxes: mostly simple, occasionally centrotyle;
1.2-1.6 mm. by 20-28 »; tyle of centrotyles 12-30 % more than adjacent
parts of the spicule in diameter. Minute dermal amphiostrongyles: attenuated
towards both ends; 125-165 by 2-3.6 p. Orthoplagiotriaenes: rhabdome 1.1—
1.45 mm. by 28-47 »; clades blunt, concave to the rhabdome throughout or
only basally and straight distally; 175-230 »long; clade-angles 97-112°. Meso-
proclades: rhabdome about 1.7 mm. by 4-9 p at the cladome, and about 20 %
more at the thickest point near the middle; clades very variable, often one or
two suppressed, 10-30 » long; clade-angles 32-64°; epirhabd conical, 25-43 ys
long or reduced to a knob.

Large oxyasters: from six to nine conical rays, 0.6-0.7 thick, and spined
everywhere, except quite at the base; no centrum; total diameter 14-20 y.

246 GEODIA BERRYI.

Large oxysphaerasters: from sixteen to twenty-two conical, spined rays, 1-1.6 4
thick; centrum 4-6 , whole aster 14-20 y, in diameter. Small strongylosphaer-
asters: from eight to fifteen truncate, spined rays, 0.6-1.3 « thick; centrum
2-3 », whole aster 6-9.2 4, in diameter. Sterrasters: 72-82 by 65-74 by 55-62 p.

Eastern Pacific. Galapagos; Duncan Island.

Geodia berryi (Souas).

TuteLy, Abhandl. Senckenb. gesellsch., 1900, 25, p. 43.

Cydonium berryi Souuas, Rept. voy. “Challenger,” 1888, 25, p. 256.

Geodia cydonium var. cerryi ! LINDGREN, Zool. anz., 1897, 20, p. 486. Zool. jahrb. Syst., 1898, 11, p. 341,
plate 18, figs. 9, 20, plate 20, fig. 4a—k, b’, c’, f”.

Small spherical, gray or brown.

Large choanosomal amphioxes: 2.16-2.54 mm. by 24-26 yu. Minute dermal
amphioxes: 240-310 by 8-10 4. Orthoplagiotriaenes: rhabdome 2.15-3.15
mm. by 51.6-72 yw; clades 175-240 » long. Protriaenes (probably mesopro-
triaenes): rhabdome 2.54-4.5 mm. by 12.9 at cladome, and 23-28 yp at thickest
point near the middle. Large anatriaenes: rhabdome 4 mm. by 25.8 y (Sollas,
1888, error 258 #)-82 4; clades 84 » long; cladome 100-112 y broad, 65-72 pu
high. Minute dermal anatriaenes: rhabdome 480 by 2-4 yw; clades 6-8 p long;
cladome 9.5-12 » broad, 6-8 yt high.

Large choanosomal asters: according to Sollas, 1888, chiasters 12-15 y in
diameter; according to Lindgren, 1898, oxyasters with from eight to fifteen rays,
16-20 y in total diameter. Oxysphaerasters: rays numerous; total diameter
12-15 ». Small strongylasters (chiasters): from six to twenty rays; total diame-
ter 8 yw. Sterrasters: 71-80 by 65-68 p.

Formerly I was inclined ' to consider G. berryi as a synonym of G. miilleri
(cydonium), and, although there can be no doubt that these forms are very simi-
lar, the experience I have recently gained with specimens from the Pacific
has made me doubtful as to their identity, so that now, like Thiele, I think it
better to retain G. berryi as a distinct species.

The species G. miilleri (cydonium) in the wider sense given to it by me in
1894 and 1903 being thus split up, I am unable to say to what part of it the
East Australian sponge mentioned by me under this name,” and of which no
material for examination is at my disposal, should be assigned. Therefore
I cannot take this sponge into consideration.

‘R. v. Lendenfeld. Die tetractinelliden der Adria. Denk. Akad. wissensch. Wien, 1894, 61, p. 138.

Tetraxonia. ‘Tierreich, 1903, 19, p. 113.
*R. v. Lendenfeld. Die tetractinelliden der Adria. Denk. Akad. wissensch. Wien. 1894, 61, p. 146.

GEODIA AMPHISTRONGYLA. 247

Northwestern and western Pacific. Coast of China; Lingin: Coast of
Cochin China, 11° 5’ N., 108° 50’ E. Ternate.

Geodia kiikenthali Tuire.e.
Abhandl. Senckenb. gesellsch., 1900, 25, p. 43, plate 2, fig. 15. LeNpDENFELD, Tierreich, 1903, 19, p. 112.

Irregularly ellipsoidal. Whitish, in the interior yellowish.

Large amphioxes: 2.8 mm. by 50 . Small amphioxes: in the choanosome
among the large ones; 300 by 9 . Orthoplagiotriaenes: rhabdome over 3 mm.
long, about 50 » thick; clades either concave to the rhabdome throughout or
curved in the opposite direction distally, 300 4” long. Mesoprotriaenes: rhabd-
ome 3.7 mm. by 10 #; clades over 150 » long. Anatriaenes: rhabdome 20 4
thick and about as long as the mesoprotriaene-rhabdome; clades strongly re-
curved, 80 long.

Large subcortical asters: rare; rays numerous, rather stout, rough, cylin-
droconical and blunt; total diameter 30 4. Small choanosomal asters: rays
pointed or blunt; total diameter 17 4. Small dermal sphaerasters: total diame-
ter 12. Sterrasters: 70 by 55 p. Sterroids; with distant rays.

Western Pacific. Ternate.

Geodia amphistrongyla LeNDENFELD.

Ante, p. 175.

Irregular, flattened. In spirit: brown.

Large amphistrongyles: 0.5-2.3 mm. by 18-32 p. Large styles: rare;
dimensions as in the amphistrongyles. Plagioclades: mostly triaene; rhabdome
1.8-2.2 mm. by 22-32 , sometimes much reduced in length, usually rounded at
the acladomal end; clades concave to the rhabdome, 155-190 » long; clade-
angles usually 103-120°. Mesoproclades: mostly triaene, occasionally mon-
aene; rhabdome 3-5 ps thick, clades 40-60 long; clade-angles 36-41°; epirhabd
16-23 p» long. Anatriaenes: rhabdome 1.5-4 yp thick; clades 26-50 » long;
clade-angles 25-41°.

Oxyasters: from five to nine usually simple, rarely bifurcate, distally spined
rays, 0.8-2.1 » thick; total diameter 20-30 ». Oxysphaerasters: from fourteen
to eighteen conical, distally spined rays, 2-2.8 thick; centrum 6~7 4, whole aster
19-28 yw, in diameter. Small strongylosphaerasters: from seven to twelve cylin-
drical or distally slightly thickened rays, 0.8-1.8 thick; centrum spherical or
irregular, 2.2-4 » in diameter; total diameter 4.8-8 py; a few have more rays,
these perhaps foreign. Sterrasters: 100-110 by 87-94 by 72-78 p.

Southeastern Pacific. Easter Island.

GEODIA ACANTHTYLASTRA.

bo
toe
)

Geodia lophotriaena LeNpDENFELD.

Ante, p. 181.

Cushion shaped. In spirit: brownish.

Large choanosomal amphioxes: 1.2-1.8 mm. by 25-42 «. Minute dermal
rhabds: mostly blunt amphioxes, rarely styles; 110-200 by 3-6 yp. Plagio-
triaenes: rhabdome 0.6-1.2 mm. by 25-35 ys, sometimes reduced in length and
rounded terminally; clades-concave to the rhabdome throughout, 140-195
long; clade-angles 102-114°. Dichotriaenes and other lophotriaenes: the former
more frequent than the latter; rhabdome 0.8-1.2 mm. by 35-59 #; main clades
70-140 y, end clades, of which there are in the lophotriaenes from three to five,
70-80 » long; main clade-angles 105-130°; breadth of cladome 300-500 y.
Mesoprotriaenes: rhabdome about 1.3 mm. by 4-9 » at the cladome, and 7-11 yp
at the thickest point near the middle; clades 44-80 » long; clade-angles 32-419;
epirhabd 38-60 long. Large anatriaenes: rhabdome 8-13 » thick; clades 60—
85 y long; clade-angles 41-50°. Minute dermal anatriaenes: rhabdome 170-
210 by 1-3 wat the cladome, and 2-4 » at the thickest point near the middle;
clades 4-9 long; clade-angles 49-67°.

Large oxyasters: from four to eleven smooth, conical rays, 1-2.2 y thick;
centrum small; total diameter 15-41 y. Sphaerasters: from seven to twenty-
two or more, usually eylindroconical, truncate, rarely conical and pointed, usually
spined rays, 0.8-2.4 # thick; centrum 2-8 y, whole aster 7-22 y, in diameter.
Sterrasters: mostly 30-45 by 33-44 by 27-35 y; a few larger ones, up to 58
long, also observed, these perhaps foreign.

? Southwestern Pacific. Probably New Zealand.

Geodia acanthtylastra LeNpDENFELD.

Ante, p. 188.

Irregularly spherical, oval or tuberous. In spirit: brownish white.

Large amphioxes: 0.7-2.2 mm. by 14-40 . Minute dermal rhabds: for the
most part amphioxes, styles also present, but rare; 150-300 by 3-15 4. Minute
amphioxes: 41-53 by 1-1.2 yw; perhaps foreign. Plagiotriaenes: rhabdome
1.2-2.5 mm. by 40-77 p; clades concave to the rhabdome throughout, or only
basally and distally straight, 160-260 » long; clade-angles 100-116°. Irregular
plagiotriaene-derivates: of similar dimensions; either with another clade besides
the three ordinary, or with bifureate clades (some regular dichotriaenes were
observed, but these may be foreign). Mesoprotriaenes: rhabdome 2.3-3.3 mm.

GEODIA NIGRA. 249

by 13-22 yp at the cladome, and 21-31 y at the thickest point near the middle;
clades 55-130 long; clade-angles 31-53°; epirhabd 30-85 plong. Anatriaenes:
rhabdome 3-5.4 mm. by 18-28 y; clades 50-110 » long; clade-angles 38-56°.

Oxyasters: a large and a small kind can be distinguished; the large: from
six to seven rays, 3 thick; total diameter 36-38 »; the small: from five to
ten rays, 0.7-3 y thick; total diameter 22-29 #; in both kinds the rays are
spined. Large oxysphaerasters: from twelve to twenty-six conical sharp-pointed
rays with a few large spines on their distal part, 1—-2.2 thick; centrum 3.5-5 p,
whole aster 12-16.5 yw, in diameter. Transitions between these and the other
aster-forms frequent. Acanthtylasters: from four to twelve cylindroconical rays,
0.5-1.3 » thick, with a terminal verticil of stout recurved knobs or spines, which
together form a conspicuous acanthtyle; centrum small; total diameter 11-22 pu.
Small strongylosphaerasters: from fourteen to twenty-two distally spined rays,
0.3-0.8 » thick; centrum 1.5-3.4 y, whole aster 4.3-6.1 , in diameter. Large
strongylosphaerasters: rare; seventeen conical, truncate rays, 5 thick, smooth
at the sides, the convex terminal face densely covered with small spines;
centrum 13 y, whole aster 23 4, in diameter; perhaps foreign. Sterrasters: 65-76
by 55-68 by 42-64 yp.

Eastern Pacific. West coast of North America; 22° 52’ N. “Albatross”
Station 2829.

Geodia nigra LENDENFELD.

Descriptive catalogue sponges Australian museum, 1888, p. 33. Tierreich, 1903, 19, p. 116.

This description is based on an examination of part of the type specimen in
the British Museum.

Massive, lobose. In spirit: dark brown or black. Surface very uneven.
Cortex of type 2-2.5 mm. thick, composed of a sterraster-armour excavated at
very frequent intervals by large cavities, extending right through it. These
cavities are occupied by lacunose tissue containing minute styles and subtylo-
styles, which perhaps belongs to another sponge, burrowing in the Geodia.
(Lendenfeld, 1888, cortex 480 y thick).

Large choanosomal amphistrongyles: attenuated towards both ends; 1.3-
1.9, rarely over 2 mm. by 15-30 »« (Lendenfeld, 1888, error, tylostyles). Minute
dermal rhabds: mostly styles, often with annular thickening near blunt end;
subtylostyles exceedingly abundant, perhaps foreign, rarely amphistrongyles;
190-370 by 5-11 » (Lendenfeld, 1888, 100 by 8 (error 80) 4. Plagio-proclades:
rhabdome rounded or blunt pointed, 1.2-1.8 mm. by 16-33 p (Lendenfeld, 1888,
40 »); clades usually unequal, one, two, or all three reduced and terminally

250 GEODIA MEDIA.

rounded, fully developed ones conical, pointed, convex to rhabdome throughout
their length, 90-260 long, over 200 # only in monaenes; clade-angles 113-147°.

Oxyasters: from six to fifteen smooth, conical rays; total diameter 17-32 y.
Oxysphaerasters: from five to thirty and more conical rays, 1.4-4 thick, rays
usually smooth, rarely with one or two stout spines; centrum 4-12 p,, whole aster
16-31 », in diameter. Sterrasters: 55-62 (Lendenfeld, 1888, 67 2) by 50-59 by
45-54 ys

Southwestern Pacific. East coast of Australia; Broughton Island.

Geodia media BowreRBANK.

Proc. Zool. soc. London, 1873, p. 13, plate 2, figs. 24-29. LenpENFeLp, Ante, p. 194.
Synops (2) media Sous, Rept. voy. “Challenger” 1888, 25, p. 266.
Sidonops media LENDENFELD, Tierreich, 1903, 19, p. 103.

The following description is also based on an examination of part of the
type specimen in the British Museum.

Massive, irregular, with depressions, in which are conspicuous sieve-covered
canal-entrances; more rarely digitate. In spirit and dry: light brown or buff
yellow.

Large amphioxes: slender and stout; 1-1.7 mm. by 23-51 yp. Large styles:
not numerous; 0.9-1.3 mm. by 30-50 “(not mentioned by Sollas, 1888). Large
irregular rhabds: angularly bent or with one or more branch-rays, often numer-
ous; dimensions as in the regular amphioxes and styles (not mentioned by
Bowerbank, 1873, or Sollas, 1888). Minute dermal styles: attenuated towards
the rounded end; 140-265 by 2-6 p» (not mentioned by Bowerbank, 1873, or
Sollas, 1888). Slender plagiotriaenes: rhabdome 1-1.6 mm. by 17-30 p; clades
strongly concave to rhabdome throughout their whole length, 160-260 long;
clade-angles 107-114°. Stout plagiotriaenes: rhabdome 0.8-1.7 mm. by 30-80
(Sollas, 1888, 45 (error 450) #4); clades slightly concave to rhabdome or nearly
straight, 110-310 » long; clade-angles 105-120°. Mesomonaenes: orthoplagio-
or proclade; rather scarce; rhabdome 1.7-2.5 mm. by 4-10 4; clade 17-50 p
long; clade-angle 32-89°; epirhabd 40-75 » long (not mentioned by Bower-
bank, 1873, or Sollas, 1888). Amphiox-like derivates of the mesoclades: of similar
dimensions; with the clade more or less completely suppressed (not mentioned
by Bowerbank, 1873, or Sollas, 1888). Anatriaenes: scarce; rhabdome 10-15 p
thick; clades 23-35 long; clade-angles 44-52° (not mentioned by Sollas, 1888).
Mesanatriaenes: very rare, not always present; rhabdome 0.8 mm. long, termi-
nally thickened and rounded; clades similar to those of the anatriaenes; epi-
rhabd 165 y long.

GEODIA EXIGUA. 251

Large oxyasters: from four to eleven conical or cylindrical, pointed or trun-
cate, spined rays, 1-3.5 y thick; total diameter 20-45 y (Sollas, 1888, 26 1).
Large oxysphaerasters: connected by transitions with the oxyasters; from fifteen
to twenty conical rays; 0.8-1.7 » thick with spine-verticils near their ends; cen-
trum 3-4.5 4, whole aster 9-18 y, in diameter (not mentioned by Sollas, 1888).
Small sphaerasters: regular, or rarely irregular, ataxastrose; from six to eighteen,
cylindrical or cylindroconical, truncate or blunt-pointed, spined rays, 0.8-2 ys
thick; centrum 2-5, whole aster 4.6-8 y, in diameter. Sterrasters: 84-110 by
73-94 by 60-74 yp.

Eastern Pacific. Gulf of Panama: Mexico (probably Pacific coast).

Geodia magellani (Souwas).

LENDENFELD, Tierreich, 1903, 19, p. 107. Thiele, Zool. jahrb. Suppl., 1905, 6, p. 40S.
Cydonium magellani Soutas, Proc. Roy. Dublin soc., 1886, 5, p. 197. Rept. voy. “Challenger,” 1888,
25, p. 221, plate 21, figs. 1-14.

In spirit: brownish white.

Amphioxes: 3.927-5.71 mm. by 51.6-58 pt. Dichotriaenes: rhabdome
3.927-4.82 mm. by 64-90 #4; main clades 127 y, end clades 275 yp long. Ana-
triaenes: rhabdome rounded terminally, 7.14-7.5 mm. by 19-23.7 yp; clades
110-116 long; cladome 160-175 » broad and 103-119 y high.

Choanosomal oxyasters: total diameter 16 p. Subcortical oxysphaerasters:
rays spined; centrum 8 4, whole aster 21.7 y, in diameter. Small sphaerasters:
rays cylindroconical, truncate; centrum large; total diameter 12 4. Sterrasters:
spheroidal 123 by 103 p.

Southeastern Pacific. Chile; Calbuco: Patagonia; Tom Bay, 50° 8’ 30” 5.,
74° 41’ W.; Port Churruca, 52° 45’ 30” S., 73° 46’ W.

Geodia exigua THIELE.

Zoologica, 1898, 24, p. 11, plate 6, fig. 8a—h. LENDENFELD, Tierreich, 1903, 19, p. 115.

Small, cylindrical. Dry: whitish.

Amphioxes 1 mm. by 11 pp. Orthoplagiotriaenes: rhabdome 1-1.2 mm. by
15 yp; clades 100 » long. Anatriaenes: rhabdome long and slender; clades
30-40 p long.

Large sphaerasters: rays cylindroconical, blunt pointed; centrum well de-
veloped; total diameter 18 yp. Large strongylasters: rare; total diameter 12
pu. Small strongylasters: total diameter 6 yp. Sterrasters: 58 by 52 p.

Northwestern Pacific. Amami-Oshima: Liu-Kiu Islands.

952 GEODINELLA ROBUSTA.

Geodia inconspicua (BowERBANK).

LENDENFELD, Tierreich, 1903, 19, p. 116.
Pachymatisma inconspicua BOWERBANK, Proc. Zool. soc. London, 1873, p. 326, plate 31, figs. 1-6.
Cydonium inconspicuum (BOWERBANK) SOLLAS, Rept. voy. ‘‘ Challenger,” 1888, 25, p. 260.

Massive. Dry: light fawn.

Large choanosomal amphioxes: 1.9mm. by 29 p. M inute dermal amphiozes :
390 «long. Orthoplagiotriaenes: rhabdome 2 mm. by 33 ys.

Oxyasters: rays slender; “centrum small; total diameter 20 Strongylo-
sphaerasters: total diameter 6.5 y. Sterrasters: spheroidal, 64 y in diameter.

? Southern Pacific. South Sea.

GEODINELLA LeNDENFELD.

Without regular triaenes. The tetraxon megascleres are monaene or diaene
teloclades with reduced clades, and occur not only in the superficial part of the
sponge but also in the interior.

Two species are known; both occur in the Pacific Ocean.

Geodinella robusta LenpDENFELD.

Ante, p. 205.

Incrusting, cushion shaped or irregularly finger shaped. In spirit: white
or brownish white.

Large choanosomal rhabds: mostly blunt amphioxes, but also amphistron-
gyles, styles, amphityles, and tylostyles; 0.37-2.5 mm. by 40-80 yp. Plagio-
monaenes, occasionally also ortho- and pro-monaenes: rhabdome 1.1—2.1 mm. by
26-42 ; clades 30-105 » long; clade-angles-87-135°. Similar diaenes: only
in var. carolae. Similar teloclades with reduced clades.

Oxyasters and oxysphaerasters: from six to twenty simple or, rarely, bifur-
cate, distally spined rays, 0.6-4 » thick; centrum, when present, up to 8 p
in diameter; total diameter of aster 9-38 y, size in inverse proportion to ray-
number. Strongylosphaerasters: from eleven to twenty-seven distally spined
rays, 1-1.7 » thick; centrum 3.5-7 p, whole aster 7-13 p, in diameter. Sterr-
asters: 180-237 by 130-200 by 80-130 p.

Northeastern Pacific. West coast of North America; southern California,
33° 58’ N. Vancouver Island, Queen Charlotte Sound: S. E. Alaska, Behm
Canal, Naha Bay. ‘‘ Albatross” Stations 2946, 4199, 4228.

GEODINELLA CYLINDRICA. 253

Geodinella robusta var. carolae LeENDENFELD.

Ante, p. 205.

Incrusting, cushion shaped. Among the large rhabds, amphistrongyles,
amphityles, styles, and tylostyles occur besides the amphioxes. Teloclades
monaene or, more rarely, diaene, with clades up to 70 wlong. Sterrasters up to
195 » long, ellipsoidal.

Northeastern Pacific. West Coast of North America; Vancouver Island,
Queen Charlotte Sound: §. E. Alaska, Behm Canal, Naha Bay. ‘‘ Albatross”’
Stations 4199, 4228.

Geodinella robusta var. megaclada LeNpDENFELD.

Ante, p. 205.

Finger shaped. Large rhabds nearly all amphiox. Teloclades all mon-
aene, clade up to 150 # long. Sterrasters up to 217 y long, ellipsoidal or, more
rarely, three lobed.

Northeastern Pacific. West Coast of North America; 8. E. Alaska, Behm
Canal, Naha Bay. ‘“‘ Albatross” Station 4228.

Geodinella robusta var. megasterra LeNDENFELD.

Ante, p. 205.

Finger shaped. Large rhabds mostly amphiox. Teloclades monaene,
clade up to 70 «long. Sterrasters up to 237 yp long, ellipsoidal or, more rarely,
three lobed.

Northeastern Pacific. West Coast of North America; off Southern Cali-

fornia; 33° 58’ N. ‘Albatross’ Station 2946.

Geodinella cylindrica (THIELE).

LENDENFELD, Tierreich, 1903, 19, p. 117.
Geodia (? !) cylindrica, THIELE, Zoologica, 1898, 24, p. 12, plate 1, fig. 2; plate 6, fig. 9a-e.

Cylindrical. Whitish brown.

Amphioxes: one end or both ends blunt, 0.8-1 mm. by 25-30 yp. Styles:
of similar dimensions. Plagio- and pro-diaenes and monaenes: rhabdome of
similar dimensions as the amphioxes; clades short, more or less reduced.

Small sphaerasters: rays conical; total diameter 7-8 ». Sterrasters: 180
by 145 by 115 yu.

Northwestern Pacific. Japan; Enoshima.

IV. DISTRIBUTION.

The limits here assigned to the Pacifie Region extend from the South Pole
along the meridian of Cape Horn to Cape Horn and along the west coast of the
American continent to Cape Prince of Wales. From here across Bering Strait
to Cape Deshnef and along the east coast of the Eurasian continent to Cape
Buhis by Singapore. Thence across the Strait of Malacca to the north coast
of Sumatra, along the eastern coast of Sumatra across the Sunda Strait to the
north coast of Java and the group of islands east of it, and across the other
straits separating these islands, to the northeast coast of Timor. From here
across the Arafura Sea to Bathurst Island and along its north coast across
Dundas Strait to the Coburg Peninsula of northern Australia. Thence along
the north and east coasts of Australia, across Bass Strait, and along the east
coast of Tasmania to the South Cape and farther, along the meridian of this
Cape, to the South Pole.

If the Geodidae of the Pacific are compared with those of other regions it is
seen that, although several of the Pacific species are similar to species found out-
side the Pacific, not a single one of the former is really identical with any of the
latter.

With the genera, however, it is different. Of the seven genera five are
represented both in the Pacific and ultra-Pacific regions, only two, Pachy-
matisma and Geodinella, being confined to one or the other, Pachymatisma to
the ultra-Pacific and Geodinella to the Pacific region. Of the five genera
common to both regions two, Caminella and Sidonops, are about equally dis-
tributed in the two regions; two, Caminus and Isops, are represented by a
larger number of species in the ultra-Pacific region than in the Pacific; and one,
Geodia, is richer in Pacific than in ultra-Pacifie species.

The total numbers of the species of the seven genera and the absolute and
percentage numbers of their Pacifie species are tabulated below.

254

DISTRIBUTION. 255

Genera Number Number of Pacific species Bercenbace Of pie ia ea
Caminella 2 1 50
Pachymatisma 5 0 0
Caminus 4 1 25
Isops 17 4 23.5
Sidonops 20 9 45
Geodia 44 29 65.8
Geodinella 2 2 100

94 ie 3 46 . ; 49 :

Within the Pacific region the following eleven areas can be distinguished,

in no two of which the same species has been found.

Sidonops californica.
ue angulata
var. megana
“ microana
“~~ orthotriaena
“f bicolor
Geodia mesotriaena
var. pachana
““megana
microana
agassizii
mesotriaenella
breviana
ovis
acanthtylastra
Geodinella robusta

“

var. carolae
“ megaclada

“ megasterra

Sidonops reticulata
Geodia ataxastra
var. angustana
“ latana

“c

media.

Sidonops oxyastra
Geodia cooksoni
“* micropora

West coast of North America.

222157 N.
Southern California.

from 33° 18’ N. to Monterey Bay, Cal.
from 33° 38’ 45” N. to 34° 22’ N.

from 33° 50’ 45” N. to Naha Bay, Behm Canal, S. E. Alaska.

Santa Barbara Island.

Southern California and near Comox, Strait of Georgia.

34° 1’ 30” N.

220521 IN.

Southern California, Queen Charlotte Sound, and Naha Bay, Behm
Canal, S. E. Alaska.

West coast of Central America.

Mexico (? which coast).
Perico Island, Gulf of Panama.

Gulf of Panama and Mexico (? which coast).

Eastern Pacific Islands..

Duncan Island, Galapagos.
Charles Island, Galapagos.
Dunean Island, Galapagos.

256

Geodia magellani

Geodia amphistrongyla

Geodia hirsuta

Isops contorta
Geodia nux
“ Jophotriaena

Isops sollasi

Sidonops nitida

Geodia eosaster
““ erinaceus
“nigra

Geodia globostella

DISTRIBUTION.

West coast of South America.

Calbuco, Chile, Tom Bay (50° 8’ 30” S., 74° 41’ W.) and Port
Churruea (52° 45’ 30” S., 73° 46’ W.) Patagonia.

Southeastern Pacific Islands.

Easter Island.

Central Pacific Islands.

Ki Island, 5° 49’ 15” S., 132° 14” 15” W.

Western and Southwestern Pacific Islands.

Fiji Islands.
Samoa Islands.
New Zealand ?

East coast of Australia.

Port Jackson.

Port Jackson.

Port Jackson.

East coast of Australia.
Broughton Island.

North coast of Australia.

Port Darwin.

Coast of Southeastern Asia and Southeastern Asiatic Islands.

Caminella nigra
Sidonops lindgreni

= picteti
e alba
Geodia distincta
«sphaeroides
“ berryi

a kiikenthali

Caminus chinensis
Isops obscura
Geodia variospiculosa
var. typica
* clavigera

“c

‘“

micraster
reniformis
japonica
. hilgendorfi
var. typica
“ granosa

“

“c

“

exigua
Geodinella cylindrica

intermedia

Gaspar Strait, Java Sea.

Java (? probably northern side).

Bay of Amboyna.

Ternate.

Java Sea and Java (? probably northern side).

Ternate and Coast of Cochin China (11° 5’ N., 108° 50’ E.).

Lingin (China), Coast of Cochin China (11° 5’N., 108° 70’ E.) and
Ternate.

Ternate.

Coast of Northeastern Asia and adjacent Islands.

China Sea and Strait of Formosa.
Japan.
Off Honshu Island and westward of Yogashima, Japan.

Enoshima, Japan.
Japan, and Enoshima, Japan.

Japan, probably Enoshima.
Amami-Oshima, Liu-Kiu Islands.
Enoshima, Japan.

DISTRIBUTION.

2)
ou
~I

Doubtful, Southern Pacific?

Isops imperfecta South Sea.
Geodia inconspicua South Sea.

Three of the six genera of the Pacific Geodidae, Caminella, Caminus, and
Isops, have been found only in the western and northwestern Pacific, on the
coasts of eastern Australia and Asia and of the eastern Asiatic Islands; one,
Geodinella, only in the northern Pacific, on the coast of Japan, and the northern
part of the west coast of North America. The other two, Sidonops and Geodia,
are more widely distributed, and the latter (Geodia) represented in every one of
the eleven areas distinguished above.

The number of species of Geodidae found in the northern half of the Pacific
is very much greater than that of its southern half. Although this difference is
no doubt to seme extent due to the inferiority of our knowledge of the latter
compared to the former, I think that it may also, in part, be ascribed to a real
relative paucity of species in the southern half of the region.

Within the eleven areas distinguished above some species, notably the
western Geodia berryi and the eastern G. agassizii, are very widely distributed.
The latter, which I was able to study carefully, exhibits very considerable differ-
ences in the specimens from the most distant localities. This, and the fact that
the Pacific species of Geodidae differ from the ultra-Pacific species and that
none of them occurs in more than one of the eleven areas distinguished, seem to
indicate that these sponges are unable to retain their characters fairly un-
changed when dispersed over extensive areas.

V. LIST OF STATIONS.

LB é I
| fo) §& | $6 :
ie PE a nstruments
Station Locality Lat. Long. | Date Z2 g 2 g z Bottom ee
| | oe |) ase es
= as | as
= a
ogog | Of Lower Cali- | \ 55 59 00 W.109 5500 | May 1,1888 | 31 75 | 74.1 | Rocky Tangles
a fornia | |
2886 Off Oregon N.43 59 00 W.i24 56 30 | Oct. 19, 1888 50 lve 48.1 a Ship’s dredge
| =] ee Large beam
2887 = a | N.43 58 00 W.124 57 00 oe Seagess 42 59 47.1 Clay, pebbles impel
= | = =
| ae = =< Sand and broken a
2894 ____ Southern | 7 34 07 00 W.120 33 30 | Jan. 5, 1889 | 53 60 55.6 | oho Ship’s dredge
California | shells
oan " a? : Small beam
Boog shen est eoubhen | N.34 22 00 W.120 08 30 | Jan. 8, 1889 | 205 59 45.2 | Green mud pa
California |
on southern . Gray sand and | Large beam
2942 California N.33 38 45 W.118 13 45 | Feb. 5, 1889 20 59 Brokentenele eeanell
oft southern - 5 Small beam
2945 4 ; 29 30 ab. 6, 1889 59 Pebbl
2045 | California N.34 00 00 W.119 29 30) Feb. 6 30 W's Gel
is os = = =
S Lar, beam
gga | OM southern | y 33 5800 W.1193045| “ “ “ 150 | 59 56.5 | Coarse gray sand | (“78°
| California | traw
2958 | a gaa N.34 04.00 W.120 19 30| Feb. 9, 1889 | 26 58 54.9 | Gray sand - Tangles
oy |W) southern | . : Gravel and | Large beam
2975 California | N.34 01 30 W.119 2900 Feb. 12, 1889 36 60 57 brokentahele trawl
— —_ | |
| Off southern Small beam
Stal : 22 qe 6 :
2978 | California | N.33 5945 W.119 15 46 60 56.5 | Gray sand fae
: Off southern Coarse gray sand | Large beam
on 9,
2981 California N.33 18 00 W.119 2400 | Feb. 13,1889 45 58 and broken shells | trawl
lar ad ¢ We Small beam
3088 | Off Oregon | N.44 28 00 W.124 25 30 | Sept. 3, 1889 46 56 46.3 | Clay, pebbles Soil
=|
ion central Cali- |
3168 | fornia N.38 01 25 W.123 2655) Mar, 24,1890 34 52 Rocky and coral | Tangles
Off Honshu Is-| 3 =
3746 | land, Japan, | N- 87 aa km. (8.5 | way 19,1900 | 49 64 mace sande ao ee
Suno Saki | ; aan

258

LIST OF STATIONS.

LIST OF STATIONS. — (Continued.)

Station

Locality

Lat.

Depth in
fathoms

Surface tem-
perature, Fahr,

Bottom tem-
perature, Fahr,

Instruments

B
ottom “Fea

3758

Off Honshu Is-
land, Japan,
Suno Saki

§S. 55 E. 3.9 km. (2.1
miles)

May 22, 1900

Long. Date
|
|

|

73,52

Tan-
beam

8-foot
ner
trawl

Black. clay
rock

and

4193

Gulf of Georgia:
Halibut Bank;
Cape Roger
Curtis, | Bowen
Island

S. 89 E. 20 km. (10.8
miles) drift S. 1° E.

June 20, 1903

18-23

50.3

| Fine green sand

4199

4228

Queen Charlotte
Sound: —

Off Fort Rupert;
Vancouver — Is-
land, B.C,
Centre of Round
Island

S. 46 W. 11.5 km. (6.2
miles) drift S. 85° E.

June 25, 1903

68-107 |

45.9

Soft green mud
and yoleanic
sand

Vicinity of Naha
Bay: Behm
Canal; Sh Ga
Alaska, Indian
Point

N. 18 E. 1.7 km. (0.9
miles) drift N. 2° W.

July 7, 1903

41-134,
|

47.8

Gravel and
sponge spicules

4417

Off southern
California, near
Santa Barbara
Island, s. w. rock
Santa Barbara
Island

N. 8° W. 11.7 km. (6.3
miles) drift S. 73° W.

Apr. 12, 1904

Fine yellow sand
and coralline
rock

4420

Off southern
California east
of Point San
Nicolas Island

S. 77 W. 10.5 km. (5.7
miles) drift 8. 60° W.

32-33 |

Fine gray sand

4531

Monterey Bay,
Cal., Point Pinos
Light House

N. 64° E. 3.8 km. (2.1
miles)

May 28, 1904

Fine gray sand,
pebbles, and rock

4551

Monterey Bay,
Cal., Point Pinos
Light House

S. 9° E. 8.4 km. (4.5
miles) drift S. 37° E.

June 7, 1904

56

Coarse sand,
shells, and rock

4

EXPLANATION OF THE PLATES.

wy

+

wr vr e

Re

a aa a= ie

PLATE 1.

Geodinella robusta LeNDENFELD.

Figs. 1-4, 16, 1S-24.— var. megasterra LENDENFELD.
Figs. 5-12, 17. — var. carolae LENDENFELD.
Figs. 13-15. — var. megaclada LENDENFELD.

1-4.— Cladomes of monaenes of the var. megasterra; magnified 100; phot., Zeiss, apochr. 16, compens.
oc. 6:
1, 2, of plagiomonaenes with a blunt clade;
3, of an orthomonaene with a pointed clade;
4, of a plagiomonaene with a pointed clade.
5-11.— Teloclades (teloclade-cladomes) of the specimen of var. carolae from Charlotte Sound; magni-
fied 100; phot., Zeiss, apochr. 16, compens. oe. 6:
5, a plagiodiaene with short, blunt rhabdome;
6, a plagiomonaene with angularly bent rhabdome;
7, 8, cladomes of plagiodiaenes;
9, a regular plagiomonaene;
10, cladome of an orthomonaene with not quite terminal, pointed clade;
11, cladome of a plagiomonaene with short, blunt clade.
12.— The specimen of var. carolae from Naha Bay; natural size; phot., Zeiss, anastig. 167.
13-15.— Cladomes of monaenes of var. megaclada; magnified 100; phot., Zeiss, apochr. 16, com-
pens. oc. 6:
13, of an orthomonaene with not quite terminal clade;
14, of a simple plagiomonaene;
15, of a plagiomonaene with a secondary clade below the cladome proper.
16.— The specimen of var. megasterra; natural size; phot., Zeiss anastig. 480/412.
17.— The specimen of var. carolae from Charlotte Sound; natural size; phot., Zeiss, anastig. 480/412.
18-20.— Plagiomonaenes of var. megasterra; magnified 30; phot., Zeiss, planar 20.
21.— Axial section of var. megasterra; magnified 6; phot., Zeiss, planar 50:
a, cortex (sterraster-armour); b, choanosome.
22, 23.— Amphioxes of var. megasterra; magnified 30; phot., Zeiss, planar 20.
24.— The specimen of var. megasterra halved: the cut surface; magnified 6, phot., Zeiss, planar 50:
a, cortex (sterraster-armour);. b, choanosome.

SPONGES OF THE PACIFIC, I. GEODIDAE. PLATE 1.

Fig. 1—24 Geodinella robust

load 4%) ” > A al ye hes n4 ea . qo

- a dD _ ere + a 7

PLATE 2.

Geodinella robusta LeENDENFELD.

Figs. 1, 3. — var. megasterra LENDENFELD.
Figs. 2, 8. — var. megaclada LENDENFELD.
Figs. 4-7, 9-11.— var. carolae LENDENFELD.

1.— Group of sterrasters in a spicule-preparation of var. megasterra; magnified 100; phot., Zeiss, apochr.
16, compens. oc. 6:
a, sterrasters of the usual ellipsoidal form lying flat; b, a three-lobed sterraster lying flat.
2.— Group of sterrasters in a spicule-preparation of var. megaclada; magnified 100; phot., Zeiss, apochr.
16, compens. oc. 6:
a, sterrasters of the usual ellipsoidal form lying flat; b, a three-lobed sterraster lying flat;
c, a sterraster of the usual ellipsoidal form standing on one of its longer narrow sides.
3.— Group of euasters in a centrifugal spicule-preparation of var. megasterra; magnified 300; phot.,
Zeiss, apochr. 4, compens. oc. 6:
a, oxyaster; b, strongylosphaerasters.
4 — Part of a section of the choanosome of the specimen of var. carolae from Charlotte Sound; congo-
red, aniline-blue; magnified 350; phot., Zeiss, hom. imm. 2, compens. oc. 2:
a, surface view of a flagellate chamber with sparse collar cells; b, sectioned flagellate chambers.
5.— Group of sterrasters in a spicule-preparation of the specimen of var. carolae from Charlotte Sound;
magnified 100; phot., Zeiss, apochr. 16, compens. oc. 6:
a, sterrasters lying flat; c, a sterraster standing on one of its longer narrow sides.
6.— Radial section of the specimen of var. carolae from Charlotte Sound; magnified; phot., Zeiss,
planar 50:
a, cortex (sterraster-armour); b, choanosome; e, wide choanosomal canals.
7.— Part of a section of the choanosome of the specimen of var. carolae from Charlotte Sound; congo-
red, aniline-blue; magnified 300; phot., Zeiss, apochr. 4, compens. oe. 6:
a, ova; b, connective tissue forming capsules enclosing the ova.
8.— Group of strongylosphaerasters in a centrifugal spicule-preparation of var. megaclada; magnified
300; phot., Zeiss, apochr. 4 compens. oe. 6.
9.— Group of strongylosphaerasters in a centrifugal spicule-preparation of the specimen of var. carolae
from Charlotte Sound; magnified 300; phot., Zeiss, apochr. 4, compens. oc. 6.
10.— Group of euasters in a centrifugal spicule-preparation of the choanosome of the specimen of var.
carolae from Charlotte Sound; magnified 300; phot., Zeiss, apochr. 4, compens. oc. 6:
a, large oxyaster; b, small oxyaster; c, small sphaeraster.
11.— Part of a section of the choanosome of the specimer: of var. carolae from Charlotte Sound; congo-
red, azure; magnified 300; phot., Zeiss, apochr. 4, compens. 6:
a, flagellate chambers.

SPONGES OF THE PACIFIC, L GEODIDAE. PLATE 2.

Fig. 1—11 Geodinella

1, 3 G. r. var. megasterra; 2, 8 G. r. var. megac

: ul) .
nid - : 7 : yet

eee tC Op a ee ee a ee

s°* ye: a
"] a -
ui = ip
a,
-) “Sn 4

PLATE 3.

Geodinella robusta LeNDENFELD.

Fig. 1. — yar. crrolae LENDENFELD.
Figs. 2, 5, 6, 8.— var. megaclada LENDENFELD.
Figs. 3, 4, 7, 9.— var. megasterra LENDENFELD.

1— A normal sterraster of the specimen of var. carolae from Charlotte Sound; magnified 300; phot.,

Zeiss, apochr. 4, compens. oc. 6.
2.— A normal sterraster of var. megaclada; magnified 300; phot., Zeiss, apochr. 4, compens. oc. 6.
3.— A normal sterraster of var. megasterra; magnified 300; phot., Zeiss, apochr. 4, compens. oc. 6.
4.— A sterroid of var. megasterra; magnified 300; phot., Zeiss, apochr. 4, compens. oc. 6.
5, 8.— A sterroid of var. megaclada; magnified 335; phot., Zeiss, apochr. 4, compens. oc. 6:
a, spherical (in the optical section ring shaped) group of central granules;
5, the centre in focus;
8, the upper surface in focus.
6.— A sterroid of var. megaclada; magnified 300; phot., Zeiss, apochr. 4, compens. oc. 6.
7, 9.— Seectioned surface of a sterraster of var. megasterra cut in half; magnified 300;
a, spherical (in section ring-shaped) group of central granules; b, a growth-zone; ¢, umbilicus;
7, phot., Zeiss, apochr. 4, compens. oc. 6, and focused higher;
9, phot., Zeiss, hom. imm. apochr. 2, compens. oc. 2, and focused lower.

ONGES OF THE PACIFICO, I. GEODIDAE. PLATE 3.

Fig. 1—9 Geodinella robusta n. Sp.
& y

1 G. r. var. carolae; 2, 5, 6, 8 G. r. var. megaclada; 3, 4,7, 9 G. 7. var. megast

Lendenfeld photographed.

=<

eS
PLATE 4
@ =

PLATE 4.

Geodinella robusta LenpENFELD.

Figs. 1, 4-7, 13, 21, 22— var. megasterra LENDENFELD.
2, 3, 8-12, 14-20.— var. carolae LENDENFELD.
25. — var. megaclada LENDENFELD.

1.— A ray of an oxyaster of the choanosome of var. megaslerra; magnified 1800; u. v. phot., Zeiss,
q- monoehr. 1.7, q. oc. 10.
2.— A sphaeraster with blunt conie rays of var. carolae; magnified 1800; u. v. phot., Zeiss, q. monochr.
1.7, q. oe. 10.
3.— Asmall oxyaster (oxysphaeraster) of var. carolae; magnified 1800; u. v. phot., Zeiss, q. monochr.
1.7, q. oe. 10.
4.— A small oxyaster (oxysphaeraster) of the cortex of var. megasterra; magnified 1800; u. v. phot.,
Zeiss, q. monoehr. 1.7, q. oc. 10.
5.— Small oxyasters (oxysphaerasters) of the choanosome of var. megasterra; magnified 1800; u. vy.
phot., Zeiss, q. monochr. 1.7, q. oc. 10.
6, 7.— The two ends of a diactine (amphiox) spicule both ends of which are lobose, of var. megasterra;
magnified 200; phot., Zeiss, apochr. 8, compens. oc. 6.
8-12.— Ends of diactine (amphiox or amphistrongyle) spicules of var. carolae; magnified 100; phot.,
Zeiss, apochr. 16, compens. oc. 6:
8, 11, very blunt (strongyle) ends;
9, 12, more tapering ends;
10, an irregular end with a clade-like process.
13.— Part of the surface of a normal sterraster of var. mejasterra; magnified 1800; u. v. phot., Zeiss,
q- monochr. 1.7, q. oc. 10:
a, the umbilicus.
14, 15.— Surface of a sterroid of the specimen of var. carolae from Charlotte Sound; magnified 1800;
u. v. phot., Zeiss, q. monochr. 1.7, q. oe. 10:
14, focused on the summits of the uppermost rays; 15, focused 1.5 p» lower.
16.— A short amphistrongyle of the specimen of var. carolae from Charlotte Sound; magnified 100;
phot., Zeiss, apochr. 16, compens. oc. 6.
17.— The rounded end of a subtylostyle of the specimen of var. carolae from Charlotte Sound; magni-
fied 100; phot., Zeiss, apochr. 16, compens. oc. 6.
15—20.— Strongylosphaerasters of the cortex of the specimen of var. carolae from Charlotte Sound;
magnified 1800; u. v. phot., Zeiss, q. monochr. 1.7, q. oe. 10.
21.— Small oxyaster (oxysphaeraster) of the choanosome of var. megasterra; magnified 1200; u. v.
phot., Zeiss, q. monochr. 2.5, q. compens. oe. 10.
22.— Oxyaster of the choanosome of var. megasterra; magnified 1200; u. v. phot., Zeiss, q. monochr.
2.5, q. oe. 10.
—25.— Parts of megascleres of var. megaclada; magnified 200; phot., Zeiss, apochr. 8, compens. oc. 6:
23, cladome of an orthomonaene with not quite terminal clade;
24, end of a blunt amphiox;
25, central part of a mesomonaene.

23

SPONGES OF THE PACIFIC, 1. GEODIDAE. PLATE 4

Fig. 1-25 Geodinella robusta n. sp.
1, 4-7, 13, 21, 22 G. r. var. megasterra; 2,3, 8-12, 14-20 G.r.var. carolae; 23-25 G.r. var. megaclada.

a Pwo! & i oe

PLATE 5.

} ‘a

PLATE 5.

Sidonops californica LENDENFELD.
Figures 1-37.

1-4.— Cladomes of normal anatriaenes; magnified 200; phot., Zeiss, apochr. 8, compens. oc. 6.
5.— Cladome of an anatriaene with reduced clades; magnified 200; phot., Zeiss, apochr. 8, compens. oc. 6.
6.— The larger of the two specimens; magnified 1.5; phot., Zeiss, anastig. 480 / 412.
7-9.— Cladomes of mesoplagioclades with more or less reduced clades; magnified 200; phot., Zeiss,
apochr. 8, compens. oc. 6:
7, a monaene;
8, a triaene;
9, a diaene.
10.— Branched end of an irregular, perhaps anatriaene-derivate megasclere; magnified 200; phot., Zeiss,
apochr. 8, compens. oc. 6. ;
11, 12.— Large amphioxes; magnified 30; phot., Zeiss, planar 20.
13, 14.— Plagiotriaenes; magnified 30; phot., Zeiss, planar 20.
15, 16.— Cladomes of plagiotriaenes; magnified 75; phot., Zeiss, apochr. 16, compens. oc. 6.
17-19.—Plagiotriaenes; magnified 30; phot., Zeiss, planar 20.
20.— Group of sterrasters and one sterroid from a spicule-preparation; magnified 200; phot., Zeiss,
apochr. 8, compens. oc. 6.
21, 22.— Group of euasters from a centrifugal spicule-preparation; magnified 300; u. v. phot., Zeiss,
q- monoehr. 6, q. oc. 7:
21, focused higher;
22, focused lower;
c, larger oxyaster.
23-26.— Groups of euasters from centrifugal spicule-preparations; magnified 300; phot., Zeiss, hom.
imm. apochr. 2:
a, small strongylosphaerasters; ec, large oxyasters.
27.— Part of a section vertical to the surface; magnified 20; phot., Zeiss, planar 20:
a, sterraster-armour; b, subcortical cavities; ¢, subcortical plagiotriaenes; d, sterrasters in
the choanosome; e, large amphioxes in the choanosome.
28, 29.— Rays of large oxyasters; magnified 1800; u. v. phot., Zeiss, q. monochr. 1.7, q. oc. 10.
30, 31.— Groups of euasters from centrifugal spicule-preparations; magnified 300:
30, phot., Zeiss, hom. imm. apochr. 2;
31, phot., Zeiss, apochr. 4, oc. 6;
a, smal] strongylosphaerasters; b, small oxysphaerasters; c, large oxyaster.
32-35.— Small strongylosphaerasters; magnified 1800; u. v. phot., Zeiss, q. monochr. 1.7, q. oc. 10.
36, 37.— Parts of the surface of two sterrasters; magnified 1800; u. v. phot., Zeiss, q. monochr. 1.7, q.
oc. 10.

SPONGES OF THE PACIFIC, 1. GEODIDAE. PLATE 5.

Fig. 1—87 Sidonops californica n. sp.

Lendenfeld photographed.

PLATE 6.

Sidonops oxyastra LENDENFELD.
Figures 1-23.

1, 2.— Parts of sections showing strands of spindle-cells traversing the choanosome; haematoxylin,
aniline-blue:

1, magnified 400; phot., Zeiss, apochr. 4, compens. oc. 6;
2, magnified 200; phot., Zeiss, apochr. 8, compens. oc. 6.

3.— Part of section through the choanosome showing canal-end branches and flagellate chambers;
haematoxylin, aniline-blue; magnified 400; phot. Zeiss, apochr. 4, compens. oc. 6.

4.— A lobe of the sponges showing a group of uniporal efferents; magnified 3; phot., Zeiss, anastig. 167.

5.— View of the largest specimen, attached to a flat stone; reduced 1 :0.63; phot., Zeiss, anastig. 480 /
412

6-13.— Plagiotriaenes; magnified 43; phot., Zeiss, achr. aa, compens. oc. 6:

6, with all the clades shortened and blunt;
7, with one reduced, blunt clade;
8-13, with pointed clades;

7, 8, 10, with unequal clades;

6, 9, 11-18, with equal clades.

14.— Two large choanosomal amphioxes from a spicule-preparation; magnified 43; phot. Zeiss, achr.
aa, compens. oc. 6.

15-18.— Cladomes of minute dermal anaclades; magnified 300:

15, 17, 18, phot., Zeiss, apochr. 8, compens. oc. 12;
19 phot., Zeiss, apochr. 4, compens. oc. 6;

15, a regular mesanatriaene;

16, an irregular mesanatriaene;

17, 18, more or less irregular anatriaenes.

19, 20.— Parts of a radial section through a region of the cortex bearing afferent pores:

19, magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6;
20, magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6;

a, monaxonid sponge attached to this part of the surface of the Sidonops; b, minute dermal
anaclades of the Sidonops; e, dermal membrane of the Sidonops oecupied by masses of
small oxysphaerasters (oxyasters).

21.— Part of a radial section through a region of the cortex bearing efferent pores, showing the dermal
layer occupied by dense masses of minute dermal rhabds; magnified 200; phot. Zeiss, apochr. 8,
compens. oc. 6.

22.— A minute dermal rhabd; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.

23.— Part of a radial section through a region of the curtex bearing afferent pores, showing numerous
minute protruding dermal anaclades; magnified 50; phot., Zeiss, achr. aa, compens. oc. 6.

SPONGES OF THE PAOIFIO, I. GEODIDAE. PLATE 6.

Fig. 1—23 Sidonops oxyastra n. sp.

al

>-

PLATE 7.

a

PLATE 7.

Sidonops oxyastra LENDENFELD.
Figures 1-20.

1, 2.— Two radial sections through a lobe of the sponge; magnified 10; phot. Zeiss, planar 50:
1, a thin section stained with haematoxylin and aniline-blue;
2, a thick unstained section;
a, sterraster-armour; b, monaxonid sponge attached to the Sidonops; ec, afferent cortical
canals; d, wide efferent canals; e, efferent pores.
3-5.— Groups of asters from a centrifugal spicule-preparation; magnified 300; phot. Zeiss, apochr. 4,
compens. oc. 6:
a, large oxyasters; b, small oxysphaerasters and oxyasters; c, large oxysphaeraster.
6.— Part of a radial section through a region of the cortex bearing afferent pores; magnified 20; phot.
Zeiss, planar 20:
a, sterraster-armour; b, monaxonid sponge attached to the Sidonops.
7, 8.— Large oxyasters from centrifugal spicule-preparations; magnified 900; u. v. phot. Zeiss, q. mon-
ochr. 1.7, q. oc. 5.
9, 10.— Group of small oxysphaerasters (oxyasters) from a centrifugal spicule-preparation; magnified
1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
9, focused higher; 10, focused lower.
11, 12.— Small oxysphaeraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
11, focused higher; 12, focused lower.
13-15.— Large oxyasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
16-18.— Small oxysphaerasters (oxyasters); magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc.
10.
19, 20.— Large oxysphaeraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
19, focused higher; 20, focused lower.

ONGES OF THE PACIFIO, 1. GEODIDAE. PLATE 7.

fig. 1—20 Sidonops oxyastra n. sp.

denfeld photovraphed.

a eter) = ‘
o :
iS oe
Siar
wat
—
‘
PLATE 8.
a
& -
Ty =
.
a J 7
is :
7 0 ae
: Cuchde hee 3 id

PLATE 8.

Sidonops oxyastra LENDENFELD.

Figures 1-15.

1-3.— Sterrasters; magnified 300:
1, phot. Zeiss, apochr. 4, compens. oc. 6;
2, 3 phot. Zeiss, hom. imm. apochr. 2.
4.— Group of spicules from a spicule-preparation; magnified 30; phot. Zeiss, planar 20:
a, large amphioxes;_b, lateral views of plagiotriaenes; c, apical view of a plagiotriaene-cladome.
5.— Group of large amphioxes from a spicule-preparation; magnified 30; phot. Zeiss, planar 20.
6-S.— Part of the lateral surface of a thick-rayed sterraster; magnified 1800; u. v. phot. Zeiss, q.
monochr. 1.7, q. oc. 10:
6, focused high; 7, focused lower: 8, focused still lower.
9.— Part of the lateral surface of a sterraster not quite fully developed; magnified 1800; u. v. phot.
Zeiss, q. monochr. 1.7, q. oc. 10.
10, 11.— Part of the lateral surface of a thin-rayed sterraster; magnified 1800; u. v. phot. Zeiss, q.
monochr. 1.7, q. oc. 10:
10, focused higher; 11, focused lower.
12.— Sterraster; magnified 1000; phot. Zeiss, hom. imm. apochr. 2, compens. oc. 6; the centrum of the
spicule in focus:
a, rosette of central granules; b, umbilicus.
13, 14— Views of parts of the surface (superficial paratangential sections) with transmitted light;
magnified 20; phot. Zeiss, planar 20:
13, of a region bearing afferent pores;
14, of a region bearing efferent pores.
15.— Part of a region of the surface with afferent pores; magnified 100; phot. Zeiss, apochr. 16, compens.
oc. 6.

PONGES OF THE PACIFIC, I. GEODIDAR. PLATE 8.

RP Hs
£4 }

Fig. 1—15 Sidonops oxyastra n. sp.

oo

a

e.
PLATE 9.
- ;

im

PLATE 9.

Sidonops bicolor LENDENFELD.
Figures 1-19.

1-6.— Cladomes of plagiotriaenes; magnified 50; phot. Zeiss, apochr. 16, compens. oe. 4:
1, 3-5, of a specimen from Station 2958;
2, 6, of a specimen from Station 4531.
7, 8.— Long slender amphioxes; magnified 20; phot. Zeiss, planar 20:
7, of a specimen from Station 4551;
8, of a specimen from Station 2781.
9-11.— Stout amphioxes; magnified 20; phot. Zeiss, planar 20:
9, a large one of a specimen from Station 4531;
10, a medium-sized one of a specimen from Station 2781;
11, a small one of a specimen from Station 2781.
12-14.— Plagiotriaenes; magnified 20; phot. Zeiss, planar 20:
12, a plagiotriaene with blunt rhabdome of a specimen from Station 2781;
13, a plagiotriaene with pointed rhabdome of a specimen from Station 2958;
14, a plagiotriaene with pointed rhabdome of a specimen from Station 2781.
15, 16.— Parts of radial sections through the cortex and adjacent parts, showing the chones of a speci-
men from Station 4420; magnified 30; phot. Zeiss planar 20: ;
a, surface of the sponge.
17.— Part of a radial section through a specimen from Station 3168; magnified 7.5; phot. Zeiss, planar
50:
a, surface of the sponge.
18.— Part of a section through the choanosome of a specimen from Station 4531; haematoxylin;
magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6.
19.— Part of a section through the choanosome of a specimen from Station 3168; haematoxylin; mag-
nified 30; phot. Zeiss; planar 20.

S OF THE PACIFIC, I. GEODIDAK. PLATE 9,

= — en, — = EZ a a .
' NN = ~ -

|
—_ ——— /
\
/
; d
a
\

Fig. 1—19 Sidonops bicolor n. sp.
1, 8—5, 13 from station 2958; 2,6, 9, 18 from station 4531; 7 from station 4551; 8, 10—12, 14 from station 2781;
15, 16 from station 4420; 17, 19 from station 3168,

Rok OS ee ee ne a

PLATE 10.

PLATE 10.

Sidonops bicolor LENDENFELD.

Figures 1-15.

1—4.— Groups of sterrasters from spicule-preparations; magnified 200; phot. Zeiss, apochr. 8, compens.
oc. 6:
1, of a specimen from Station 2958;
2, of a specimen from Station 3168;
3, of a specimen from Station 2781;
4, of a specimen from Station 4420.
5.— Sterraster of a specimen from Station 4551; magnified 300; u. v. phot. Zeiss, q. monochr. 6, q. oc. 7.
6-12.— Groups of euasters from centrifugal spicule-preparations; magnified 300; u. v. phot. Zeiss, q.
monoehr. 6, q. oc. 7:
6, of a specimen from Station 3168;
7, of a specimen from Station 4420;
8, 11, of a specimen from Station 2781; . <2
9, 10, 12, of a specimen from Station 4551;
a, oxyasters; b, strongylosphaerasters.
13.— Group of euasters from a centrifugal spicule-preparation of a specimen Fromm Station 4551; mag-
nified 300; u. v. phot. Zeiss, q. monochr. 1.7: ;
a, oxyasters; b, strongylosphaerasters. ;
14, 15.— Pores in the dermal membrane of a specimen from Station 3168; magnified 75; phot. Zeiss,
apochr. 16, compens. oc. 4:
14, a uniporal efferent opening;
15, a cribriporal afferent opening.

ONGES OF THE PACIFIC, I. GEODIDAE. PLATE 10

Fig. 1—15 Sidonops bicolor n. sp.
from station 2958; 2,6, 14, 15 from station 3168; 3,8, 11 from station 2781; 4,7 from station 4420; 5, 9, 10, 12, 13 from station 4551.

=
. \
.
i J
v
*
é
: «
: ‘
’
‘ in
2 Ay F wy
: &
Hg . 2
“ -
fi :

PLATE 11.
*
=

PLATE 11.

Sidonops bicolor LENDENFELD.

Figures 1-17.

1, 2.— A larger and a smaller strongylosphaeraster of a specimen from Station 4551; magnified 1800;
u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10: ci
1, focused higher; 2, focused lower.
3-5.— Strongylosphaerasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
3, 5, of a specimen from Station 2781;
4, of a specimen from Station 3168.
6-8.— Groups of euasters from a centrifugal spicule-preparation; magnified 1800; u. v. phot. Zeiss, q.
monochr. 1.7, q. oc. 10:
6, a strongylosphaeraster (a) and a medium-sized oxyaster (b) of a specimen from Station 4551;
7, a strongylosphaeraster (a), and an oxysphaeraster (b) of a specimen from Station 2781;
8, a strongylosphaeraster (a) and a large oxyaster (b) of a specimen from Station 2781.
9.— An oxyaster of a specimen from Station 2781; magnified 900; u. v. phot. Zeiss, q. monochr. 1.7, q.
oe. 5. _
10.— Group of strongylosphaerasters of a specimen from Station 2781; magnified 900; u. v. phot.
Zeiss, q. monoehr. 1.7, q. oc. 5. ; ‘
11-14.— Parts of the surface of sterrasters; magnified 1800; u. v. phot, Zeiss, q. monochr. 1.7, q. oc. 10:
11, 12, of the umbilical side of sterrasters of a specimen from Station 2958;
13, of the side opposite the umbilicus of a specimen from Station 2781;
14, of the side opposite the umbilicus of a specimen from Station 2958.
15-17.— ‘Three specimens; natural size; phot. Zeiss, anastig. 480 / 412:
15, 16, two specimens from Station 4420;
17, a specimen from Station 4551.

SPONGES OF THE PACIFIC, 1. GEODIDAE.

Tre
ee

ion 4420.

2 j7vo

L

958:

11, 12, 14 fron

3168 ;

On

4 from stati

1, 2, 6, 17 from station 4551; 3, 5, 7—10, 13 from station 2781;

Lendenfeld photographed.

- 4

PLATE 12.

ae

» J.

PLATE 12.

Sidonops angulata LENDENFELD.

Figs. 1-4, 16, 19. — var. megana LENDENFELD (lobose form).
Figs. 5-8, 17, 20. — var. megana LENDENFELD (massive form).
Figs. 9, 10. — yar. orthotriaena LENDENFELD.

Figs. 11-15, 18, 21, 22.— var. microana LENDENFELD. _

1-14.— Cladomes of anaclades; magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6:
1-4, of the lobose specimen of var. megana;
5-8, of the massive specimen of var. megana;
9, 10, of var. orthotriaena;
11-14, of var. microana;
1, 5, of anatriaenes with long equal clades;
2, 4, 12-14, of anatriaenes with short equal clades;
3, 6, 8, 9, of anatriaenes with medium-sized equal clades;
7, of an anatriaene with medium-sized unequal clades;
10, 11, of anadiaenes with short clades.
15.— The acladomal end of the anatriaene of var. microana, the cladome of which is represented in Fig.
14; magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6.
16, 17.— Groups of spicules from spicule-preparations; magnified 10; phot. Zeiss, planar 50:
16, of the lobose specimen of var. megana;
17, of the massive specimen of var. megana;
a, slender, dermal, simply curved amphioxes; b, slender, dermal, angularly bent amphiox;
c, stout, choanosomal amphioxes; d, plagiotriaenes.
18-20.— Views of three of the specimens; phot. Zeiss, anastig. 480 / 412:
18, var. microana; magnified 1.1;
19, the lobose specimen of var. megana; reduced 1 : 0.86;
20, the massive specimen of var. megana; reduced 1 : 0.9.
21, 22.— Groups of plagiotriaenes from a spicule-preparation of var. microana; magnified 10; phot.
Zeiss, planar, 50.

PONGES OF THE PACIFIC, I. GEODIDAE. PLATE 12.

a

d

b

16

20 29

Fig. 1—22 Sidonops angulata n
I—4, 16, 19 lobose Specimen of S. a. var. megana; 5—8, 17, 20 massive specimen of S. a. var. megana; 9, 10 S. a. var
11— 15, 18, 21, 22 S. a. var. microana.

denfeld photographed.

mr! Ye

; oe
: -
. 7
i: - im aa
-— :
ic
oH. Ser a

PLATE 13.

Sidonops angulata LENDENFELD.

Figs. 1-8. — var. megana LENDENFELD (lobose form).
Figs. 9-12, 22-25.— var. megana LeNDENFELD (massive form).
Figs. 13-17, 21. — var. microana LENDENFELD.

Figs. 18-20. — var. orthotriaena LENDENFELD.

1, 2.— Stout, choanosomal amphioxes of the lobose specimen of var. megana; magnified 40; phot.
Zeiss, planar 20.
3.— A plagiotriaene of the lobose specimen of var. megana; magnified 40; phot. Zeiss, planar 20.
4.— Apical view of the cladome of an irregular plagiotriaene of the lobose specimen of var. megana;
magnified 40; phot. Zeiss, achr. aa, compens. oc. 6.
5-16.— Cladomes of plagiotriaenes seen from the side; magnified 40; phot. Zeiss, achr. aa, compens.
oc. 6:
5-8, from the lobose specimen of var. megana;
9-12, from the massive specimen of megana;
13-16, from var. microana;
5, with long, somewhat unequal, terminally strongly recurved clades;
6, 8, with long, somewhat unequal, terminally slightly recurved clades;
7, 13, 16, with unequal, slightly recurved clades;
9, with short, equal, nearly straight clades;
10, with short, irregularly curved, somewhat unequal clades;
11, 15, with medium-sized, equal, slightly recurved clades;
12, with unequal, nearly straight clades;
14, with medium-sized, slightly unequal, rather strongly recurved clades.
17.— Group of spicules from a spicule-preparation of var. microana; magnified 40; phot. Zeiss, planar

20:
a, an angularly bent amphiox; b, a nearly straight amphiox; ec, the cladomal half of an
anatriaene.
18.— Cladome of an orthotriaene of var. orthotriaena; magnified 40; phot. Zeiss, achr. aa, compens.
oc. 6.

19.— An orthotriaene with angularly bent rhabdome of var. orthotriaena; magnified 40; phot. Zeiss,
achr. aa. compens. oc. 6.
20.— Two rhabds of var. ortholriaena; magnified 40; phot. Zeiss, planar 20:
a, a club-shaped style; b, an amphiox.
21.— Part of a radial section of var. microana; magnified 10; phot. Zeiss, planar 50:
a, cortex (sterraster-armour); b, spicule-fur.
22-24.— Surface views (parts of superficial paratangential sections) of the massive specimen of var.
megana; magnified 100; phot. Zeiss, apochr. 16, compens. oe. 6:
22, an efferent uniporal opening;
23, 24, afferent cribriporal openings (pore-sieves).
25.— Part of a radial section through the superficial part of the massive specimen of var. megana; mag-
nified 40; phot. Zeiss, achr. aa, compens. oc. 6:
a, sterraster-armour; b, a chone; e¢, a subcortical cavity.

NGS OF THE PACIFIC, I. GEODIDAE. PLATE 13

1—8 lobose specimen of S. a. var. megana; 9—12, 22—25 massive specimen of S. a. var. megana; 13—17, 21 S. a. var. microana;

18—20 S. a. var. orthotriaena

nfeld photographed,

Phototype by Charles Bellmann Prague

a ee ew bo ge a ce

Pay ee)

PLATE 14.

PLATE 14.

Sidonops angulata LENDENFELD.

Figs. 1-4, 18, 19. — var. megana LENDENFELD (lobose form).
Figs. 5, 6, 16, 17, 20-22.— var. megana LENDENFELD (massive form).
Figs. 7-9. — var. microana LENDENFELD.

Figs. 10-15, 23-80. — var. orthotriaena LENDENFELD.

1-15.— Microscleres from centrifugal spicule-preparations; magnified 300:
1, 5, 7, 8, u. v. phot. Zeiss, q. monochr. 6, q. oc. 7;
24, 6, 9-15, u. v. phot. Zeiss, q. monochr. 1.7;
14, from the lobose specimen of var. megana;
5, 6, from the massive specimen of var. megana;
7-9, from var. microana;
10-15, from var. orthotriaena:

a, small oxysphaerasters with large centrum; b, large oxyasters without centrum; c,
young sterrasters; d, strongylosphaerasters; e, intermediate oxyasters (oxysphaer-
asters) with small centrum.

16.— A small oxysphaeraster of the massive specimen of var. megana; magnified 1800; u. v. phot. Zeiss,
q- monoehr. 1.7, q. oc. 10.
17.— Two strongylosphaerasters of the massive specimen of var. megana; magnified 1800; u. v. phot. —
Zeiss, q. monochr. 1.7, q. oc. 10.
18, 19.—— A strongylosphaeraster with few, irregularly distributed, fully developed rays, of the lobose —
specimen of var. megana; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
18, focused higher; 19, focused. lower.
20.— Radial section through the dermal membrane of the massive specimen of var. megana; haematoxy- —
lin; magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6:
a, surface of the sponge; b, conspicuous, granular, subdermal cells; ¢, asters protruding
beyond the surface.
21.— Radial section through the choanosome of the massive specimen of var. megana; azure; magnified
200; phot. Zeiss, apochr. 8, compens. oc. 6:
a, flagellate chambers; c, asters.
22.— Radial section through the cortex of the massive specimen of var. megana; azure; magnified 200;
phot. Zeiss, apochr. 8, compens. oc. 6:
a, fibrous inner cortical layer; b, subcortical cavity; ec, asters, protruding into the subcorti-
cal cavity. . .
23.— Small oxysphaeraster of var. orthotriaena; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q.
oc. 10. b
24.— Large triactine oxyaster with rudiment of a fourth ray of var. orthotriaena; magnified 1800; u.
v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
25-30.— Strongylosphaerasters of var. orthotriaena; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, —
q- oe. 10: 4
25, part of a regular strongylosphaeraster with cylindrical rays;
26, a strongylosphaeraster with somewhat irregularly distributed, cylindrical rays;
27, 28, two views of a regular strongylosphaeraster with somewhat conical rays;
27, focused higher; 28, focused lower;
29, 30, two views of a strongylosphaeraster with only one fully developed ray;
29, focused higher; 30, focused lower. ;

INGES OF THE PACIFIC, I. GEODIDAR.
PLATE 14.

29 30

Fig. 1—380 Sidonops angulata n. sp.
* 79S. a. var. microana; 10—15,

—4, 18, 19 lobose specimen of S. a. var. megana; 5,6, 16, 17, 20—22 massive specimen of S. a, var. megana;
23—30 S. a. var. orthotriaena.

idenfeld photographed.

- Phototype by Charles Bellmann, Prague.

PLATE 15. ¢

PLATE 15.

Sidonops angulata LeNDENFELD.

Figs. 1, 2, 4, 9. — var. megana LENDENFELD (massive form).
Figs. 3, 7, 8, 11— var. megana LENDENFELD (lobose form).
Figs. 5, 6, 12. — var. orthotriaena LENDENFELD.

Fig. 10. — var. microana LENDENFELD.

1, 2.— Two views of the umbilicus and the adjacent parts of the surface of a sterraster of the m
specimen of var. megana; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
1, focused higher: 2, focused lower. 4
3.— The umbilicus and the adjacent parts of the surface of a sterraster of the lobose specimen of var
megana; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
4.— Part of a section through the choanosome of the massive specimen of var. megana;-haematoxy
magnified 300; phot. Zeiss, apochr. 4, compens. oe. 6: 4
a, masses of small cells; b, lumen of a canal.
5, 6.— Part of the surface opposite to the umbilicus of a sterraster of var. orthotriaena; magnified
u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10: bs
5, focused higher; 6, focused lower.
7.— Part of a paratangential section of the cortex of the lobose specimen of var. megana; magni
phot. Zeiss, apochr. 8, compens. oc. 6:
a, blunt cones protruding into a radial cortical canal; b, lumen of the radial cortical can:
c, sterrasters. .
8.— Part of a paratangential section of the lobose specimen of var. megana, transverse through a ch
magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6:
a, position of the (closed) chonal canal.
9-12.— Groups of sterrasters from spicule-preparations: magnified 200; phot. Zeiss, apochr. 8, con
oc. 6:
9, of the massive specimen of var. megana;
10, of var. microana;
11, of the lobose specimen of var. megana;
12, of var. orthotriaena.

NGES OF THE PAOIFIC, I. GEODIDAE. PLATE 15.

Fig. 1—12 Sidonops angulata n. sp.

2, 4,9 massive specimen of S. a. var. megana; 3, 7, 8, 11 lobose specimen of S. a. var. megana; 5, 6, 12, S. a.

var. orthotriaena; 10 S, a, var. microana.

idenfeld photographed.

wate he 6 wet to eile Praone-

—

_

|

i

Re

PLATE 16. : :
Pep

PLATE 16.

Geodia media BoweRBANK.
Figs. 1-21.

1-14.— Megascleres; magnified 30; phot. Zeiss, planar 20:
1-5, 7n, 0, 8b, 9y, 10x, 13, 14, from the digitate specimen;
6, 7m, p, 8z, 9e, w, 10d, 11, 12, from massive specimens;
1-6, 11, groups of megascleres from spicule-preparations;
7, mesoclades with reduced clades;
8, regular stout amphioxes;
9, angularly bent amphioxes;
10, branched amphioxes;
12, 14, plagiotriaenes;
13, regular, angularly bent, and branched styles,

a, (Figs. 3-5) large, slender, regular amphioxes; b, (Figs. 1-65, 8) large, stout, regular
amphioxes; ¢, (Fig. 9) large, stout amphiox, strongly angularly bent at one —
point; d, (Figs. 5, 10) large amphioxes with one simple branch near one of the
ends, enclosing a small angle with the axis of the spicule; e, (Fig. 5) large am-
phiox with a bifid branch near one end; f, (Figs. 11, 13) large, regular, simple
styles; g, (Fig. 13) large, angularly bent styles; h, (Figs. 11, 13) large styles with
one simple branch; i, (Fig. 13) large style with a bunch of simple branches near
the pointed end; j, (Fig. 2) large angularly bent amphiox with one simple —
branch; k, (Fig. 13) large angularly bent style with one simple branch; 1, (Fig. 6)
anatriaene, probably foreign to the sponge; m, (Fig. 7) mesoplagiomonaene with
nearly straight rhabdome; n, (Fig. 7) mesopromonaene; 0, (Fig. 7) mesoplagio-
orthomonaene; p, (Fig. 7) mesoplagiomonaene with strongly curved rhabdome;
q, (Figs. 1, 12, 14) slender regular plagiotriaenes; r, (Fig. 5) slender plagiotriaene
with unequal clades; s, (Figs. 1-6, 11, 12) stout regular plagiotriaenes; t, (Fig. 12) _
stout plagiotriaene with one bifid clade; u, (Fig. 3) stout plagiodiaene with one
bifid clade; v, (Fig. 1) stout mesoplagiotriaene with long epirhabd (pointing
downwards in the figure); w, (Fig. 9) large, stout amphiox angularly bent in
two places; x, (Fig. 10) large stout amphiox with one simple branch near the
blunt end, enclosing a large angle (of nearly 90°) with the axis of the spicule; y,
(Figs. 1, 9) large, stout amphioxes slightly angularly bent at one point; z,
(Figs. 6, 8, 11) very large, stout, regular amphioxes.

15.— The umbilical part of the surface of a sterraster of a massive specimen; magnified 1800; u. v.
phot. Zeiss, q. monochr. 1.7, q. oc. 10.
16.— The largest of the massive specimens; natural size; phot. Zeiss, anastig. 480 / 412.
17.— The digitate specimen; natural size; phot. Zeiss, anastig. 480 / 412.
18, 19.— Groups of sterrasters from spicule-preparations; magnified 200; phot. Zeiss, apochr. 8, com-
pens. oc. 6:
18, of a massive specimen;
19, of the digitate specimen. :
20, 21.— A sterroid of the digitate specimen; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
20, the upper surface in focus; 21, the equatorial profile in focus.

—_—

ONGES OF THE PACIFIC, I. GEODIDAE. PLATE |
*LATE 16

Fig. 1—21 Geodia media Bwbk.
d—5, 7 2, 0, 8 6,97, 10 x, 13, 14, 17, 19—21 digitate form; 6,7 m, p, 8 2, 9 ¢, w, 10 d, 11, 12, 15, 16,

nfeld photographed,

Photoivpe by Charles Bellmann, Prague

33.— The rounded end of a style (subtylostyle) of var. micraster; magnified 75; phot. Zeiss, apo
16, compens. oc. 6.
34-38.— Cladomes of orthoplagiotriaenes and dichotriaenes of var. intermedia; magnified 40; pho
Zeiss, apochr. 16:
34-36, cladomes of regular adult orthoplagiotriaenes ;
37, cladome of a young orthoplagiotriaene;
38, a, cladome of a rather irregular adult orthoplagiotriaene;
b, cladome of a dichotriaene. ;
39, 40.— Two aspects of the specimen of var. intermedia; phot. Zeiss, anastig. 480 / 412:
39, natural size;
40, magnified 1.14.
41.— The specimen of var. micraster; magnified 1.07; phot. Zeiss, anastig. 480 / 412.
42.— A choanosomal amphiox of var. micraster; magnified 20; phot. Zeiss, planar 20.
43.— A choanosomal style of var. micraster; magnified 20; phot. Zeiss, planar 20.
44-47.— Orthoplagiotriaene cladomes of var. micraster; magnified 40; phot. Zeiss, apochr. 16.
44, 45, with unequal but otherwise regular clades;
46, with one shortened and truncate and one abruptly bent clade;
47, with rather equal regular clades.
48,— Cladome of a dichotriaene of var. micraster; magnified 40; phot. Zeiss, apochr. 16.
49,— Cladome of a fairly regular orthoplagiotriaene of var. intermedia seen from above, magnified
phot. Zeiss, apochr. 16.
50.— Cladome of a dichotriaene of var. micraster; magnified 40; phot. Zeiss, apochr. 16.

SPONGES OF THE PACIFIC, I. GEODIDAE. PLATE 17

> <2 > goes

oe:
+t are 3
Hee o

Fig. 1—22 Geodia media Bwbh.
1—10, 13, 14, 16, 17, 20—22 massive form; 11, 12, 15, 18, 19 digitate form.
Fig. 23—50 Geodia variospiculosa Thiele
23—26, 34—40, 49 G. v. var. intermedia; 27—33, 41—48, 50 G. v. var. micraster
ndenfeld photographed.

PLATE 18.

Geodia variospiculosa Tuie.e.

Figs. 1-7, 9, 11, 12, 21, 23-26.— var. micraster LENDENFELD.
Figs. 8, 10, 13-20, 22, 27. — var. intermedia LenDENFELD.

1.— Large hexactine oxyaster of var. micraster; magnified 300; u. v. phot. Zeiss, q. monoehr. 6, q. oc. 7.
2.— Group of asters from a centrifugal spicule-preparation of var. macraster; magnified 300; u. v. phot.
Zeiss, q. monochr. 6, q. oc. 7:
a, large triactine oxyaster; b, smaller oxyasters; c, small strongylosphaerasters.
3.— Large hexactine oxyaster of var. micraster; magnified 300; u. v. phot. Zeiss, q. monochr. 6; q-
Clee
4.— Large triactine oxyaster of var. micraster; magnified 300; u. v. phot. Zeiss, q. monochr. 6, q. oc. 7.
5.— Group of spicules from a centrifugal spicule-preparation of var. maicraster; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
¢, small strongylosphaerasters; d, large monactine oxyaster; e, parts of dermal styles.
6.— A large monactine oxyaster of var. micraster; magnified 300: phot. Zeiss, hom. imm. apochr. 2.
7.— Group of asters from a centrifugal spicule-preparation of var. micraster; magnified 300; phot. Zeiss
apochr. 4, compens. oc. 6:
b, smaller oxyasters; c, small strongylosphaerasters.
8.— Part of a radial section through the subcortical layer of var. intermedia; magnified 100; phot.
Zeiss, apochr. 16, compens. oc. 6:
a, cortical sterraster-armour; b, bundle of radial megascleres; c, a subcortical group of small
dermal styles; d, cladome of an anatriaene.
9.— View of part of the surface of var. micraster, showing a pore-sieve; magnified 75; phot. Zeiss, apochr.
16, compens. oc. 6.
10.— Group of spicules from a centrifugal spicule-preparation of var. intermedia; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
a, large triactine oxyaster; b, small oxyaster; c, small strongylosphaerasters; f, oxysphaeraster.
11.— Group of small strongylosphaerasters from a centrifugal spicule-preparation of var. micraster;
magnified 300; phot. Zeiss, apochr. 4; compens. oc. 6.
12.— A small oxyaster of var. micraster; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
13.— An oxysphaeraster of var. intermedia; magnified 300; u. v. phot. Zeiss, q. monochr. 6, q. oc. 7.
14.— Group of asters from a centrifugal spicule-preparation of var. intermedia; magnified 300; u. vy.
phot. Zeiss, q. monochr. 6, quo: 75
b, smaller oxyaster; c, small strongylosphaerasters.
15-20.— Large oxyasters of var. intermedia; magnified 300; u. v. phot. Zeiss, q. monochr. 6, q. oc. 7:
15, 17, 19, large pentactine oxyasters;
16, a large triactine oxyaster;
18, 20, large tetractine oxyasters.
21.— Radial section through the superficial part of var. micraster; magnified 10; phot. Zeiss, planar
50:
a, sterraster-armour; b, a chone; ¢, subcortical cavities; d, choanosome.
22.— Part of a radial section of the proximal portion of the choanosome of var. intermedia; magnified
200; phot. Zeiss, apochr. 8, compens. oc. 6:
a, large oxyasters; b, smaller oxyasters.
23.— Part of a radial section of the proximal portion of the cortex of var. micraster, passing through a
(closed) chonal canal; magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6:
a, oxysphaerasters surrounding the (closed) chonal canal.
24.— Part of a radial section through the distal portion of the cortex and the spicule-fur of var. micraster;
magnified 200; phot. Zeiss, apochr. 8, compens. oe. 6:
a, minute dermal anaclades: b, rhabdome of a large protruding anatriaene; c, surface of the
sponge.

’

25.— Part of a radial section through the distal part of the choanosome of var. micraster; Tinewiied 200;
phot. Zeiss, apochr. 8, compens. oc. 6:
a, medium-sized oxyaster; b, smaller oxyasters.
26.— Radial section through the superficial part of var. micraster; magnified 2% phot. eis
20:
a, sterraster-armour; b, a chone; ¢, small dermal styles and anaclades piven )
surface; d, choanosome; e, small dermal styles and anaclades in the subcorti
27.— Part of a radial section of the proximal portion of the choanosome of var. intermedia;
200; phot. Zeiss, apochr. 8, compens. oc. 6:
a, large oxyasters; b, smaller oxyasters.

PLATE 18.

SPONGES OF THE PACIFIC, 1. GEODIDAR.

Fig. 1—27 Geodia variospiculosa Thiele.

9, 11, 12, 21, 23—26 G, v.

micraster; 8, 10, 13—20,

var.

a5

1

enfeld photographed.

a ea

| E
\ =
: a
Ay

PLATE 19.

Geodia variospiculosa THIELE.

Figs. 1-8, 12-18, 21, 23, 25-30, 32.— var. micraster LENDENFELD.
Figs. 9-11, 19, 20, 22, 24, 31. — var. intermedia LENDENFELD.

1, 2.— An umbilical part of the surface of a normal sterraster of var. micraster; magnified 1800; u. v.
phot. Zeiss, q- monochr. 1.7, q. oc. 10:
1, focused higher; 2, focused lower.
3.— Cladome of a minute dermal anatriaene of var. micraster; magnified 750; phot. Zeiss, hom. imm.
apochr. 2, compens. oc. 6.
4, 5.— Small dermal styles of var. micraster; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
6.— Minute dermal anamonaene of var. micraster; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
7, 8.— Minute dermal anatriaenes of var. micraster; magnified 300; phot. Zeiss, apochr. 4, compens.
oc. 6.
9, 10.— Minute dermal anatriaenes of var. intermedia; magnified 300; u. v. phot. Zeiss, q. monochr. 6,
q. oc. 7.
11.— An umbilical part of the surface of a normal sterraster of var. intermedia; magnified 1800; u. v.
phot. Zeiss, q. monochr. 1.7, q. oc. 10.
12, 13.— Group of asters from a centrifugal spicule-preparation of var. micraster; magnified 1800;
u. v. phot. Zeiss, q. monochr. 1.7 q. oc. 10:
12, focused higher; 13, focused lower;
a, oxysphaeraster; b, small strongylosphaerasters.
14.— Cladome of a minute dermal anatriaene of var. micraster; magnified 1800; u. v. phot. Zeiss, q.
monochr. 1.7, q. oc. 10.
15, 16.—Sterroid with stout, smooth, spineless rays of var. micraster; magnified 300; phot. Zeiss,
apochr. 4, compens. oc. 6:
15, focused higher (the upper surface in focus); 16, focused lower (the contour in focus).
17, 18.— Sterroid with very stout, terminally densely spined rays of var. micraster; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
17, focused higher (the upper surface in focus); 18, focused lower (the contour in focus).
19-24.— Small strongylosphaerasters and groups of such from centrifugal spicule-preparations; magni-
fied 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
19, 20, 22, 24, of var. intermedia;
21, 23, of var. micraster.
27.— Rays of oxyasters of var. micraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
8.— A monactine aster of var. micraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
29, 30.— Rays of oxyasters of var. micraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
31.— Group of sterrasters from a spicule-preparation of var. intermedia; magnified 200; phot. Zeiss,
apochr. 8, compens. oc. 6.
32.— Group of sterrasters from a spicule-preparation of var. micraster; magnified 200; phot. Zeiss,
apochr. 8, compens. oc. 6.

2 OF THE PACIFIC, I. GEODIDAE. PLATE 19,

=}

Se Ee ee ca) WY

2

Fig. 1—32 Geodia vartospiculosa Thiele,
1—8, 12—18, 21, 23, 25—30, 32 G. v. var. micraster; 9—11, 19, 20, 22, 24, 31 G. v. var. intermedia,

ndenfeld photographed.

rz

| : =
-

PLATE 20.

Geodia amphistrongyla LeNDENFELD.
Figures 1-41.

les; magnified 30; phot. Zeiss, planar 20.

1-3.— Choanosomal amphistrongy
blique epirhabd; magnified 200; phot. Zeiss, apochr. 8, com-

4.— Cladome of a mesanatriaene with o
pens. oc. 6.
5, 6.— Cladomes of anatriaenes; magnified 200; phot. Zeiss, apochr. 8, compens. oe. 6:
5, with smaller clade-rhabdome angle;
6, with larger clade-rhabdome angle.
7, 8.— Cladomes of mesoproclades, magnified 200; phot. Zeiss, apochr. 8, compens. oe. 6:
7, of a mesopromonaene,

8, of a mesoprotriaene.
9— Part of radial section through the distal portion of the choanosome, showing a cluster of young

sterrasters in situ; magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6.
10.— Cladome of an anatriaene; magnified 300; phot. Zeiss, hom. imm. apochr. 2.
11.— Cladome of an anatriaene; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
. phot. Zeiss, q. monochr. 6, q. oc. 10.

12.— Large choanosomal oxyaster; magnified 450; u. v.
magnified 750; phot. Zeiss, hom. imm. apochr.

13, 14.— Parts (rays) of large choanosomal oxyasters;
2, compens. oc. 6.
15, 16.— Parts (rays) of large choanosomal oxyasters; magnified 1800; u. v. phot. Zeiss, q. monochr.

1.7, q. oc. 10:
17.— One end of a cylindrical amphistrongyle with thickenings; magnified 200; phot. Zeiss, apochr.
8, compens. oc. 6.
18-21.— Large choanosomal rhabds and parts of such; magnified 100; phot. Zeiss, apochr. 100, com-
pens. oc. 6.:
18, a short anisoactine, somewhat style-like amphistrongyle;
19, a larger, nearly isoactine, cylindrical amphistrongyle;
20, the rounded end of a large style (the pointed end of this spicule is represented in Fig. 21);
21, the pointed end of a large style (the rounded end of this spicule is represented in Fig. 20).
92-95.— Cladomes of plagiotriaenes and derivates of such: magnified 100; phot. Zeiss, apochr. 16,
compens. oc. 6:
22, cladome of a regular plagiotriaene with short clades;
23, cladome of a plagiomonaene with normal rhabdome and simple clade;
24, an orthodichomonaene with shortened rhabdome;
25, cladome of a regular plagiotriaene with long clades.
26-30.— Groups of microscleres from a centrifugal spicule-preparation ; magnified 300:
26-28, 30, phot. Zeiss, apochr. 4, compens. oc. 6;
29, phot. Zeiss, hom. imm. apoehr. 2;
a, large choanosomal oxyasters; b, large oxysphaerasters; ¢, small strongylosphaerasters; d,
young sterrasters.
31.— View of the sponge; magnified 1.1; phot. Zeiss, anastig. 480 / 412.
32.— Group of sterrasters from a spicule-preparation; magnified 200; phot. Zeiss, apochr. 8, compens.
oc. 6.
33.— Radial section through the superficial part of the sponge; magnified 20; phot. Zeiss, planar
20:
a, cortex; b, a chone; ¢, subcortical cavities; d, choanosome.
34-36.— Small strongylosphaerasters, two single ones and a group, from a centrifugal spicule-prepa-
ration; magnified 1800; u. v. phot. Zeiss, monoehr. 1.7, q. oc. 10.
37, 38.— The umbilicus of a sterraster and the adjacent parts; magnified 1800; u. v. phot. Zeiss, q. —
monochr. 1.7, q. oc. 10:
37, focused lower; 38, focused higher.
39.— Radial section through the superficial part of the sponge; magnified 10; phot. Zeiss, planar 50:
a, cortex; ¢, subcortical cavities; d, choanosome.
40, 41. Part of the surface of the side of a sterraster; magnified 1800; u. v. phot. Zeiss, q. monochr.
1:7, q: oc: 10: i
40, focused lower; 41, focused higher.

PLATE 20.
NGES OF THE PACIFIC, I. GEODIDAE.

Fig. 1—41 Geodia amphistrongyla n. sp.
Fig. 1 Geodia z)

enfeld photographed,

+

fen, ae

a

i : : | s >

PLATE 21.

Geodia mesotriaena LENDENFELD.

Fig. 1. — var. pachana LENDENFELD.
Figs. 2-6.— var. megana LENDENFELD.

1.— Dry specimen of var. pachana seen from above; reduced 1 : 0.67; phot. Zeiss, anastig. 480 /
412.
2.— Radial section of a spirit specimen of var. megana, reduced 1: 0.67; phot. Zeiss, anastig. 480 /
412:
a, hirsute part, where the projecting spicules have not been broken off; b, praeoscular cavity.
3_6.— Teloclades of var. megana; magnified 10; phot. Zeiss, planar 50:
3, 5, orthotriaenes fully developed;
4, young orthotriaene;
6, mesoprotriaene.

SPONGES OF THE PACIFIC, 1. GEODIDAE.

Fig. 1—6 Geodia mesotriaena n. sp.

1 G. m. var. pachana; 2—6 G. m. var.

endenfeld photographed.

PLATE

|

rt,

ie ee ee ee ee ee ee eee —s Le

. AT er a! a en Toa
oes ao ah : ES

——-- a

j e ni . ie ) a

ae ao w

PLATE 22. E

rye

a>. =

PLATE 22.

Geodia mesotriaena var. megana LENDENFELD.

,

Figures 1-10.

1-6.— A series of paratangential sections through a chone and the canals leading to it from the
pores (a chonal system); magnified 30; phot. Zeiss, planar 20: 7
1, first section, the central part in the level of the centre of the concave pore-sieve;
2, second section, the central part half way down the ectochrotal layer overlying the s
armour;
3, 4, third and fourth section, in the distal and proximal part of the sterraster-armour res
5, fifth section, at the proximal limit of the sterraster-armour;
6, sixth section, through the chone, below the sterraster-armour;
a, dermal pores (in Fig. 1); b, oblique superficial canals leading from the pores to th
canal (Figs. 1, 2); ¢, chonal canal (Figs. 3-6).
7.— Slightly oblique paratangential section, the lower part a, a little higher than the upper |
magnified 10; phot. Zeiss, planar 50:
c, chones.
$.— The central part of the first of the paratangential sections represented in Fig. 1; magni
phot. Zeiss, apochr. 8, compens. oe. 6.
9.— Part of a radial section through the choanosome, showing large granular cells; azure; 0
750; phot. Zeiss, hom. imm. apochr. 2, compens. oc. 6:
a, spindle-shaped granular cells, pointed at both ends; b, a granular cell, poin'
one end.
10.— Part of a paratangential section, a transverse section through a chone; magnified |
Zeiss, apochr. 8, compens. oc. 6:
a, chonal canal; b, dense mass of small asters; ¢, circular fibres.

99

PLATE

SPONGES OF THE PACIFIC, I. GEODIDAR.

Geodia mesotriaena n.

j—1l9)

Fig.

Lendenfeld photographea.

ee ah) os

PLATE 23.

Geodia mesotriaena LeENDENFELD.

Figs. 1, 2. — yar. microana LENDENFELD.
Figs. 3, 5, 6, 8, 9.— var. pachana LENDENFELD.
Figs. 4, 7, 10-25. ;— var. megana LENDENFELD.

1-12.— Cladomes of anatriaenes; magnified 109; phot. Zeiss, apochr. 16, compens. oe. 6:
1, 2, of fully developed anatriaenes of var. microana;
3, 5, 6, 8, 9, of fully developed anatriaenes of var. pachana;
4, 7, of a young anatriaene of var. megana;
10-12, of fully developed anatriaenes of var. megana.
13-19.— Cladomes of teloclades and mesoclades of var. megana; magnified 100; phot. Zeiss, apochr
16, compens. oe. 6:
13, of an irregular mesoprotria
epirhabd;
14, of a mesoprotriaene with one clade abruptly bent, and one strongyle;
15, of an irregular protriaene with one clade nearly vertical;
16, of a regular orthotriaene;
17, of an irregular mesanatriaene with strongly deflected epirhabd;
18, of a young regular mesoprotriaene;
19, of a regular mesoprotriaene fully developed. ;
20-23.— Cladomes of orthotriaenes of var. megana; magnified 50; phot. Zeiss, apochr. 16, com
oc. 4:
20, of an orthotriaene fully developed with clades not very different in size;
21, of a young quite regular orthotriaene;
22, of an orthotriaene fully developed with one clade much shorter than the others;
23, of an orthotriaene fully developed with one clade abruptly bent and shorter than the o
24, 25.— Radial section through a part of the cortex and the adjacent choanosome of var. megana:
24, magnified 10; phot. Zeiss, planar 50;
25, (a part of 24) magnified 30; phot., Zeiss, planar 20;
a, outer surface; b, sterraster-armour; ce, superficial canals; d, chone; e, radial main
somal canal; f, tissue free from flagellate chambers surrounding the radial main
somal canal; g, tissue containing flagellate chambers; h, radial spieule bundles; i,
more or less radial, dermal rhabds protruding beyond the surface; k, chonal canal.

ene with branches forming a second verticil of clades on the

PLATE 28.

Fig. 1—25 Geodia mesotriaena n. sp.
1, 2 G. m. var. microana; 38, 5, 6, 8, 9 G. m. var. pachana; 4, 7, 10—25 G. m. var. megana,

PLATE 24.

Geodia mesotriaena LeENDENFELD.

Figs. 1, 4, 8, 14, 15, 17, 18, 20, 22-32.— var. megana LENDENFELD.
Figs. 2, 6, 7, 10-13, 16, 19, 21. — var. Microana LENDENFELD.
Figs. 3, 5, 9. — var. pachana LENDENFELD.

1.— Group of small spicules in a centrifugal spicule-preparation of var. megana; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
a, parts of small dermal rhabds; b. medium oxyasters.
2.— Group of small spicules in a centrifugal spicule-preparation of var. microana; magnified 100; phot.
Zeiss, apochr. 16, compens. oc. 6:
a, small dermal rhabds; b, euasters.
3.— Group of small spicules in a centrifugal spicule-preparation of var. pachana; magnified 100; phot.
Zeiss, apochr. 16, compens. oc. 6:
a, small dermal rhabds; b, euasters.
4.— Group of small spicules in a centrifugal spicule-preparation of var. megana; magnified 100; phot.
Zeiss, apochr. 16, compens. oc. 6:
a, small dermal rhabds; b, euasters.
5.— Group of small spicules in a centrifugal spicule-preparation of var. pachana; magnified 300. phot.
Zeiss, apochr. 4, compens. oc. 6:
a, parts of small dermal rhabds; b, oxysphaerasters; ec, small strongylosphaerasters.
§.— Group of small spicules in a centrifugal spicule-preparation of var. microana; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
a, part of a small dermal rhabd; b, large oxyaster; c, small strongylosphaerasters; d, medium
oxyaster.
7.— Group of euasters in a centrifugal spicule-preparation of var. microana; magnified 300; phot. Zeiss,
apochr. 4, compens. oc. 6:
a, large oxyaster; b, large strongylosphaeraster; c, oxysphaeraster; d, small strongylosphaer-
asters.
8.— Group of euasters in a centrifugal spicule-preparation of var. megana; magnified 300; phot. Zeiss,
apochr. 4, compens. oc. 6:
a, sterroid; b, oxysphaeraster.
9.— Group of euasters in a centrifugal spicule-preparation of var. pachana; magnified 300; phot. Zeiss,
apochr. 4, compens. oc. 6:
a, large strongylosphaeraster; b, small oxyaster.
10, 11.— Groups of euasters in a centrifugal spicule-preparations of var. microana; magnified 600; u. v.
phot. Zeiss, q. monochr. 2.5, q. oc. 5:
a, large medium oxyaster with blunt conic rays; b, small strongylosphaeraster.
12, 13.— Group of euasters in a centrifugal spicule-preparation of var. microana; magnified 600; u. v.
phot. Zeiss, q. monochr. 2.5, q. oe. 5:
12, focused higher; 13, focused lower;
a, large medium oxyaster with stout blunt rays; b, small strongylosphaerasters.
14.— Group of euasters in a centrifugal spicule-preparation of var. megana; magnified 600; u. v. phot.
Zeiss, q. monochr. 2.5, q. oe. 5:
a, medium oxyasters; b, small strongylosphaeraster.
15.— Group of large medium oxyasters in a centrifugal spicule-preparation of var. megana; magnified
200; phot. Zeiss, apochr. 8, compens. oc. 6.
16.— Small strongylosphaeraster with numerous rays of var. microana, magnified 600; u. v. phot.
Zeiss, q. monochr. 2.5, q. oe. 5.
17, 18.— Small strongylosphaerasters with a medium number of rays of var. megana; magnified 900;
u. v. phot. Zeiss, q. monochr. 2.5, q. oe. 7.
19.— Medium oxyaster of var. microana; magnified 600; u. v. phot. Zeiss, q. monochr. 2.5, q. oc. 5.
20.— Group of small strongylosphaerasters in a centrifugal spicule-preparation of var. megana; magni-
fied 600; u. v. phot, Zeiss, q. monochr. 2.5, q. oc. 5.

21.— Group of euasters in a centrifugal spicule-preparation of var. microana; magnified 600; u. v.
phot. Zeiss, q. monochr. 2.5, q. oc. 5: >
a, large strongylosphaeraster ; b, small strongylosphaerasters.
2, 23.— Large oxyasters of var. megana, magnified 1200; u. v. phot., Zeiss, q. monoehr. 2.5, q-
4, 25.— Group of euasters in a centrifugal spicule-preparation of var. megana; magnified 900; 1
phot. Zeiss, q. monochr. 2.5, q. 0c. Ue s
24, focused higher; 25, focused lower;
a, large oxyaster with slender rays; b, medium oxyasters; ¢, small strongylosphaerast
26-31.— Tips of rays of oxyasters of var. megana: magnified 2650; u. v. phot., Zeiss, q- monoe
q. oc. 14: 7
26, a nearly smooth ray;
27, a slightly spined ray;
28-31, strongly spined rays.
39.— Two small strongylosphaerasters of var. megana; magnified 2650; u. v phot., Zeiss, q-

1.7, q. oc. 14.

2
2

a

SPONGES OF THE PACIFICO, I. GEODIDAE.

PLATE 24.

et.
»: Ee

27 28° 29

Ko)
eS)
o
ie)
_
eS)
nN

Fig. 1—32 Geodia mesotriaena n. sp.
1, 4, 8, 14, 15, 17, 18, 20, 22—832 G. m. var. megana; 2,6, 7, 10—13, 16, 19, 21 G. m. var. microana; 3,5, 9 G. m. var. pachana.

— photographed. .

vinta 7
\

i
| -
AD)

a 5

PLATE 25.

PLATE 25.

Geodia mesotriaena var. megana LENDENFELD.

Figures 1-11.

1.— Radial section through the outer, ectochrotal layer of the cortex; magnified 100; phot.,
apochr. 16, compens. oc. 6:
a, surface of the sponge; b, distal, freely projecting ends of the small dermal rhabds: ¢
oblique superficial canals; d, radial tufts of small dermal amphioxes; e, sterraster-arm«
2, 3.— Sterraster; magnified 300; u. v. phot., Zeiss, q. monochr. 1.7:
2, focused higher; 3, focused lower.
4.— Sterraster; magnified 300; phot., Zeiss, apochr. 4, compens. oc. 6.
5.— Group of sterrasters in a spicule-preparation; magnified 150; phot., Zeiss, apochr. 8, compens.
6,7.— The umbilicus and the adjacent parts of the surface of two sterrasters; magnified 1800
phot., Zeiss, q. monochr. 1.7, q. oc. 10.
8.— Surface of a sterraster opposite the umbilicus; magnified 1800; u. v. phot. Zeiss, q. monoch
q. oe. 10.
9-11.— The umbilicus of a sterraster and the adjacent parts of the surface; magnified 1200; u. ne
Zeiss, q. monochr. 2.5, q. oe. 10:
9, focused near the bottom of the umbilical pit; 10, focused halfway up the umbilical
11, focused on the surface round the umbilical pit. :

PLATE 25,

GES OF THE PACIFIC, I. GEODIDAR,

8
S
8
~
8
B
i)

ta
&
x
8
ae)
S
=
»
S
S
S
8
8
8
~~
S
™
~
~
e

é
:

PLATE 26.

Geodia agassizii LenDENFELD (adult).
Figures 1-21.

1-12.— Orthoplagiotriaenes and orthoplagiotriaene-derivates; magnified 20; phot. Zeiss, planar
20:
1, a mesorthotriaene orthoplagiotriaene-derivate with simple, terminally abruptly bent clades,
of a specimen from Station 2978;
2, an amphielade orthoplagiotriaene-derivate with three simple, very unequal clades and one opistho-
clade a considerable distance from the cladomal end of the rhabdome, of a specimen from
Station 3168;
3, an orthoplagiotriaene with oblique cladome and simple clades, of a specimen from Station 2886;
4, a regular orthoplagiotriaene with short, simple clades, of a specimen from Station 2978;
5, 6, 8, 12, orthoplagiotriaenes some or all the clades of which are terminally branched, of a speci-
men from Station 4199;
7, a regular orthoplagiotriaene with long simple clades, of a specimen from Station 2886;
9, 10, 11, orthoplagiotriaenes some or all the clades of which are terminally branched, of a speci-
men from Station 3168.
13.— Surface-view of a detached piece of the cortex of a specimen from Station 4199; magnified 7.5;
phot. Zeiss, planar 50.
14, 15.— An efferent cortical canal in a detached piece of an efferent area of the cortex of a specimen
from Station 4199; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
14, seen from within, the chonal sphincter, a, is in focus; an indistinct image of the dermal sieve, b, is
seen behind; ;
15, seen from without, the dermal sieve, b, is in focus; an indistinet image of the chonal sphincter, a,
is seen behind.
16—21.— Six specimens of the sponge; 16, 18-21, natural size, 17 reduced 1:0.67; phot. Zeiss, anastig.
480 / 412:
16, 19-21, from Station 2886;
17, from Station 4193;
18, from Station 2887.

SPONGES OF THE PACIFIC, I. GEODIDAE. PLATE 26.

Fig. 1-21 Geodia agassizit n. sp. (adult). ra
1,4 from station 2978; 2,9-11 from station 3168; 3, 7,16, 19-21 from station 2886; 5, 6,8, 12-15 fri
17 from station 4193 ; 18 from station 2887.

station 4199;

PLATE 27.

’

Teepe oe a el ee ee Pe

‘mn de

PLATE 27.

Geodia agassizii LeNDENFELD (adult).

Figures 1-19.

1.— Radial section through the region bearing efferent pores of a large specimen from Station 4193;
magnified 7.5; phot. Zeiss, planar 50:
b, sterraster-armour of the cortex; d, large afferent canals; e, large efferent canal.
2.— Radial section through the region bearing afferent pores of a medium-sized specimen from Station
3168; magnified 7.5; phot. Zeiss, planar 50: 7
a, ectochrotal outer layer of the cortex, free from sterrasters, with small, radial dermal
amphioxes; b, sterraster-armour of the cortex; c, subcortical cavities; d, large afferen
canal.
3-13.— Groups of spicules from centrifugal spicule-preparations; magnified 300; phot. Zeiss, apo
4, compens. oe. 6:
3, 4, 9, of specimens from Station 2978;
5, of a specimen from Station 2887;
6, 11, of a specimen from Station 3168;
7, of a specimen from Station 4551;
8, of a specimen from Station 4193;
10, of a specimen from Station 2886;
12, of a specimen from Station 4199;
13, of a specimen from Station 3088; : 7
a, small dermal amphioxes: b, large choanosomal oxyasters; ¢, large cortical oxysphaerasters;
d, small strongylosphaerasters. .
14.— Group of asters from a centrifugal spicule-preparation of a specimen from Station 4193; magnified
500; phot. Zeiss, hom. imm. apochr. 2, compens. oc. 6:
b, large oxyaster; e¢, large oxysphaeraster; d, small strongylosphaeraster.
15-19.— Groups of sterrasters from spicule-preparations; magnified 200; phot. Zeiss, apochr. 8, com-
pens. oc. 6: .
15, of a specimen from Station 2978;
16, of a specimen from Station 4551;
17, of a specimen from Station 4193;
18, of a specimen from Station 3088;
19, of a specimen from Station 4199.

PLATE 27.

at ‘Sea =p.

4

~~ a gy ape tf)

| Fig. 1-19 Geodia agassizit n. sp. (adult).
, 8, 14, 17 from station 4193; 2,6, 11 from station 3168; 3, 4,9, 15 from station 2978; 5 from station 2887; 7,16 from station 4551;
10 from station 2886; 12,19 Srom station 4199; 13,18 from station 3088.

:

,

-

: A
th Z

PLATE 28.

Geodia agassizii LENDENFELD (adult).
Figures 1-28.

1-7.— Cladomes of mesoprotriaenes; magnified 75, phot. Zeiss, apochr. 16, compens. oc. 6:
1, with long and stout curved clades, one of which is terminally abruptly bent, and a rather long
epirhabd, of a specimen from Station 4199;
2, with rather long and stout curved clades and long epirhabd, of a specimen from Station 4199;
3, with stout and short curved clades and short epirhabd, of a specimen from Station 4199;
4, with short, nearly straight clades and medium epirhabd, of a specimen from Station 4551;
5, with slender, slightly curved clades and rather long epirhabd, of a specimen from Station 2978;
6, with long and slender, curved clades and short epirhabd, of a specimen from Station 2978;
7, with rather stout, nearly straight clades and very long epirhabd, of a specimen from Station 4199.
8-11.— The branched ends of quite irregular telo- and mesoclade-like spicules; magnified 75; phot.
Zeiss, apochr. 16, compens. oc. 6:
8, the branched end of a long rhabd with three conie branch-rays (clades) lying nearly in the same
plane as the rhabd (rhabdome), of a specimen from Station 4199;
9, the branched end of a rhabd with two branch-rays lying in a straight line and together appearing
as a small style attached obliquely to the end of the rhabd, of a specimen from Station 2887;
10, the branched end of an amphiox-like spicule with two very oblique, backwardly directed spines
inserted a little below the end, of a specimen from Station 2887; ;
11, the branched end of an amphiox-like spicule, with a straight branch-ray arising nearly vertically
a little distance below one end, of a specimen from Station 2887. E
12-14.— Parts of stout, club-shaped styles; magnified 75; phot. Zeiss, apochr. 16, compens. oc. 6;
of specimens from Station 4193:
12, 13, the thick blunt ends of two such styles;
14, the thin pointed end of the style, the other end of which is represented in Fig. 13.
15.— An end of a regular large choanosomal amphiox, of a specimen from Station 3168; magnified 75;
phot. Zeiss, apochr. 16, compens. oc. 6. t
16, 17.— Groups of spicules from spicule-preparations; magnified 20; phot. Zeiss, planar 20: j
16, of a specimen from Station 2978;
17, of a specimen from Station 2886;
a, large choanosomal amphioxes; b, a club-shaped style; ¢c, an anatriaene; d, a mesopro-~
triaene. .
18—28.— Cladomes of anaclades; magnified 75; phot. Zeiss, apochr. 16, compens. oc. 6:
18, of an anatriaene with short, somewhat unequal clades, of a specimen from Station 2887;
19, of an anatriaene with short, rather stout, somewhat unequal clades, of a specimen from Station
4193;
20, 21, of anatriaenes with long and slender, equal clades, of a specimen from Station 2978;
22, of an anatriaene with rather long and fairly slender, equal clades, of a specimen from Station 2887;
23, of an anatriaene with short, stout, equal clades, of a specimen from Station 4199;
24, of an anatriaene with long and stout, equal clades, of a specimen from Station 4551;
25, of an anatriaene with long and stout, equal clades, of a specimen from Station 4193;
26, of an anatriaene with small, nearly straight, equal clades, of a specimen from Station 4199;
27, of an anatriaene with slender, very unequal clades, of a specimen from Station 3168;
28, of an anadiaene, of a specimen from Station 3168.

|
)
|
!

OF THE PACIFIC, 1. GEODIDAR. PLATE 28,

Fig. 1-28 Geodia agassizit n. sp. (adult).
iss 23, 26 from station 4199; 4, 24 from station 4551; 6, 16, 20, 21; from station 2978; 9-11, 18,22 from station 2887;
12-14, 19, 25 from station 4193; 15, 27, 28 from station 3168; 17 from station 2880.

ee ee 8

a.

,
PLATE 29.
S.

PLATE 29.

Geodia agassizii LENDENFELD (adult).

Figures 1-17.

1-17.— Cladomes of orthoplagiotriaenes and orthoplagiotriaene-derivates; magnified 75; phot. Zeiss,

1,

2

3,

9,
10,
11,
12,

13,

14,
15,
16,

17,

apochr. 16, compens. oc. 6:
of an orthoplagiotriaene with medium rhabdome and simple, somewhat unequal clades, of a
specimen from Station 2886;

, of an orthoplagiotriaene with slender cladome and simple, equal clades, of a specimen from

Station 4199;
of an orthoplagiotriaene with thick rhabdome and short, rather equal, simple clades, of a speci-
men from Station 2978;

, of an orthoplagiotriaene with slender rhabdome, and irregularly extending, simple clades, of a

specimen from Station 3088;

of an orthoplagiotriaene with thick rhabdome, and simple unequal clades, of a specimen from
Station 2978; .

of an orthoplagiotriaene, with slender rhabdome, and clades, partly abruptly recurved and
partly branched terminally, of a specimen from Station 4199;

of a mesorthotriaene with thick rhabdome, and clades terminally abruptly bent, of a specimen
from Station 2978;

of an orthoplagiotriaene with thick rhabdome, and simple, equal clades, of a specimen from
Station 4551;

of an orthoplagiotriaene with slender rhabdome, and unequal clades, one of which is branched, of a
specimen from Station 4199;

of an orthotriaene with slender rhabdome and unequal clades, one of which is branched, of a
specimen from Station 4193;

of a plagiotriaene with medium rhabdome and simple, equal clades, of a specimen from Station
2887;

of an orthoplagiotriaene with slender rhabdome and unequal clades, one of which is branched, of
a specimen from Station 4199;

of an amphiclade orthotriaene-derivate with slender rhabdome with a verticil of three unequal,
terminally branched clades and one simple and knob-shaped opisthoclade, of a specimen from
Station 4199;

of an orthoplagiotriaene with medium rhabdome and unequal clades, one of which is branched,
of a specimen from Station 3088; ,

of an orthoplagiotriaene with medium rhabdome and very unequal, simple clades, of a speci-
men from Station 3168;

of an orthoplagiotriaene with medium rhabdome and unequal, branched clades, of a specimen
from Station 3168;

of an orthoplagiotriaene with slender rhabdome and unequal clades, one of which is branched,
of a specimen from Station 4199. :

PLATE 29.

Fig. 1-17 Geodia agassizit n. sp. (adult).
tf Srom station 4199; 3, 5,7 Srom station 2978; 4, 14 from station 3088; 8 from station 4551;

Il from station 2887; 15,16 from station 3168.

——a <r

PLATE 30. :

=
A
i? ‘
7 *
at a mis = sf ~s

PLATE 30.

Geodia agassizii LeNDENFELD (adult).

Figures 1-17.
1, 2.— Group of asters from a centrifugal spicule-preparation, of a specimen from Station 2886; |
fied 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
1, focused higher; 2, focused lower;
a, small strongylosphaerasters; b, large oxyaster.
3.— Large oxysphaeraster of a specimen from Station 2978; magnified 1800; u. v. phot. Zeiss,
ochr. 1.7, q. oc. 10.
4.— Large oxyaster of a specimen from Station 2978; magnified 1800; u. v. phot. Zeiss, q. 0
ec eOC> LO.
5.— Part Ge a large oxyaster of a specimen from Station 4193; magnified 1800; u. v. pho’
monoehr. 1.7., q. oc. 10.
6-9.— Small strongylosphaerasters: magnified 1800; u. vy. phot. Zeiss, q. monoehr. 1.7, q. oc
6, 8, of a specimen from Station 4193;
7, 9, of a specimen from Station 3088.
10.— Group of asters from a centrifugal spicule-preparation of a specimen from Station 3168;
1800; u. v. phot. Zeiss, q. monoehr. 1.7, q. oc. 10:
a, small strongylosphaerasters; b, large oxyasters.
11-17.— Parts of sterrasters in different stages of development (growth) of a specimen from
3088; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
11, 12, part of the side opposite the umbilicus of a quite young sterraster with slender,
sharp-pointed rays;
11, focused lower; 12, focused higher;
13, part of the umbilical side of a sterraster somewhat older than the one represented 1
and 12, with stouter, but still simple, pointed rays;
14, part of the umbilical side of a sterraster older than the one represented in Fig. 13, t
which are still simple but already stout and blunt; 7
15, 16, part of the side opposite the umbilicus of two sterrasters still older, in which the
cils are beginning to appear on the summits of the rays; be
17, part of the side opposite the umbilicus of a young sterraster with rays already d
extended and crowned by verticils of slender spines. ;

PLATE 30.

°
-_ ae Fig. I-17 Geodia agassizit n. sp. (adult). ;
om station 2886; 3,4 from station 2978; 5, 6,8 from station 4193; 7,9, 11-17 from station 3088; 10 from station 3168.

PLATE 31.

PLATE 31.

Geodia agassizii LeNpDENFELD (adult).

Figures 1-10.

1-10.— Parts of fully developed sterrasters; magnified 1800; u. v. phot. Zeiss, q. monochr. ale

oc. 10:

the umbilical side of a normal sterraster of a specimen from Station 3088;

1, focused higher; .

2, focused lower;

3, 4, the umbilical side of a sterroid with large terminal extensions of the rays, of a spec
from Station 4193;
3, focused higher:
4, focused lower;

5, the umbilical side of a sterraster of a specimen from Station 4193, focused just above the bot
of the umbilical pit to show the roughness of its sides;

6, 7, the umbilical side of a normal sterraster of a specimen from Station 4193;
6, focused higher;
7, focused lower;

8-10, the umbilical side of a sterroid of a specimen from Station 3088, with few and tanael r
spines on the ends of the rays, focused in three levels about 2 » apart; }
8, focused high; 9, focused intermediate; 10, focused low.

1,2;

ONGES OF THE PACIFIC, I. GEODIDAR. PLATE 31

i

Fig. 1-10 Geodia agassizii n. sp. (adult).
1, 2, 8-10 from station 3088; 3-7 from station 4193;

PLATE 32.

PLATE 32.

Geodia agassizii LeNDENFELD (young) from Station 4425.
Figures 1-46.

1.— Part of a radial section through the choanosome; haematoxylin; magnified 200; phot. Zeiss,
apochr. 8, compens. oc. 6:
a, flagellate chambers.
2, 3.— Small sphaerasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
2, with conical, pointed rays;
3, with cylindrical, truncate rays.
4.— Group of euasters from a centrifugal spicule-preparation; magnified 300; u. v. phot. Zeiss, q. mon-
ochr. 1.7.
5.— Part of a radial section through the choanosome; haematoxylin; magnified 10; phot. Zeiss, planar
50:
a, an efferent canal with constrictions.
6, 7— Large oxysphaerasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
8.— Part of a radial section; magnified 10; phot. Zeiss, planar 50:
a, sterraster-armour. >
9, 10.— Large choanosomal amphioxes; magnified 30; phot. Zeiss, planar 20.
11.— Part of a radial section; magnified 10; phot. Zeiss, planar 50: é
a, sterraster-armour; b, protruding parts of spicules forming the spicule-fur.
12.— Part of a radial section; magnified 10; phot. Zeiss, planar 50:
a, sterraster-armour; b, rhabdome of an orthotriaene with rounded and thickened aclad
end.
13-39.— Sterroids and sterrasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
13, 15, 17, 19, 21, 23, 25, 27, and 29-39, the highest part of the upper surface in focus;
14, 16, 18, 20, 22, 24, 26, 28, the spicules represented above them in Figs. 13, 15, 17, 19, 21, 23, 25,
27, focused lower.
40—42.— Cladomes of mesoproclades; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
40, 41, of regular slender ones;
42, of a stout one with one rudimentary clade. (This spicule may be foreign).
43-46.— Cladomes of anaclades; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
43, 46, of regular anatriaenes with well-developed, pointed clades;
44, of an anadiaene with well-developed pointed clades;
45, of an irregular anatriaene with small clades, one of which is shortened and rounded at the end.

PLATE 32.

NGES OF THE PACIFIC, I. GEODIDAE.

Fig. 1-46 Geodia agassizii n. sp. (young specimen from station 4228).

PLATE 33.

Geodia agassizii LeNDENFELD, (young) from Station 4228.

Figures 1-14.

1-8.— Four stages of development of the sterraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7,
q. oc. 10:
1, 3, 5, 7, the highest part of the upper surface in focus;
2, 4, 6, 8, the sterrasters (parts of sterrasters) represented to the left of them in Figs. 1, 3, 5, 7,
focused lower;
1, 2, young sterraster 17 » in diameter;
3, 4, young sterraster 23 » in diameter;
5, 6, young sterraster 50 » in diameter;
7, 8, young sterraster 70 » in diameter.
9-14.— Parts of the surface of adult sterrasters and sterroids; magnified 1800; u. v. phot. oe 015
monochr. 1.7, q. oc. 10: *

“es of which are rough, apparently covered with secondary spinelets;
10, umbilical part of the surface of a sterraster with extremely thick se the spines of which are
numerous and smooth; 1
11, part of the lateral surface of a sterroid;
2, 13, part of the surface, some distance from the umbilicus, of a sterraster with thin protruding
standing very close together;
12, focused lower; 13, focused higher.

PLATE 33.

Fig. 1-14 Geodia agassizii n. sp. (young specimen from station 4228).

3

PLATE 34.

Geodia agassizii LenpeNreLp (young) from Station 4228.

Figures 1-17:

1-16.— Orthoplagiotriaenes and orthoplagiotriaene-derivates; magnified 30; phot. Zeiss, planar 20:
1-7, orthoplagiotriaenes with the rhabdome pointed at the acladomal end;
8, 10-12, 14, orthoplagiotriaenes with the rhabdome slightly shortened and rounded and more
or less thickened at the acladomal end;
9, orthoplagiotriaene with the rhabdome slightly shortened and simply rounded at the acladomal
end;
13, amphiclade orthoplagiotriaene-derivate with the rhabdome shortened, rounded, thickened, and
provided with a spine-like opisthoclade at the acladomal end;
15, orthoplagiotriaene-derivate with one clade bifureate and the rhabdome much shortened, and
rounded and thickened at the acladomal end;
16, mesoclade orthoplagiotriaene-derivate.
17.— Transverse section of the specimen; magnified 1.6; phot. Zeiss, planar 100.

Geodia mesotriaenella LeNDENFELD.
Figures 18-26.

18.— Group of asters from a centrifugal spicule-preparation; magnified 300; u. v. phot. Zeiss, q. monochr.
LESF/E
a, large oxyasters; b, small strongylosphaerasters.
19.— View of the sponge; natural size; phot. Zeiss, anastig. 480 / 412.
20.— Portion of the surface bearing afferent pores, a superficial paratangential section viewed with trans-
mitted light; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
21, 22.— Large oxyaster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
21, focused higher; 22, focused lower.
23.— A sterraster; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
24.— A large oxyaster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
25.— Portion of the surface bearing efferent pores, a superficial paratangential section viewed with
transmitted light; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
26.— Group of small strongylosphaerasters from a centrifugal spicule-preparation; magnified 1800;
u. v. phot. Zeiss, q. monoehr. 1.7, q. oc. 10:
a, with longer rays attenuated towards the end; b, with shorter cylindrical rays.

ON GES OF THE PACIFIC, I. GEODIDAE.

PLATE 34.

eet = =o 8 |p See ee

Fig. 1-17 Geodia agassizii n. sp. (young specimen from station 4228).
Fig. 18-26 Geodia mesotriacnella n. sp.

Le 7 . z =
PLATE 35.
) ¢
> te
J
=
Loe

PLATE 35.

Geodia breviana LENDENFELD.
Figures 1-27.

1—4.— Large choanosomal amphioxes; magnified 30; phot. Zeiss, planar 20.
5-7.— Cladomes of anatriaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
8-13.— Small strongylosphaerasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
8, 9, group of two strongylosphaerasters from a centrifugal spicule-preparation;
8, focused higher; 9, focused lower;
10, group of two strongylosphaerasters from a centrifugal spicule-preparation;
11, 12, a strongylosphaeraster; 3
11, focused higher; 12, focused lower;
13, a strongylosphaeraster.
14.— Cladome of a mesoprotriaene; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
15-17.— Ortho- and plagiotriaenes; magnified 30; phot. Zeiss, planar 20.
1S8-22.— Groups of asters from centrifugal spictile-preparations; magnified 300.
18, 20-22, u. v. phot. Zeiss, q. monochr. 1.7;
19, phot. Zeiss, apochr. 4, compens. oc. 6;
a, large oxyasters; b, small strongylosphaerasters; ec, large oxysphaeraster.
23.— A sterraster; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.

24-27.— Large oxyasters and oxysphaerasters; magnified 1800; u. v. phot. Zeiss, q. monoehr. 1.7, q.
oc. 10: i
24, 27, two large oxyasters;
25, 26, a large oxysphaeraster;
25, focused higher; 26, focused lower.

Geodia mesotriaenella LenpDENFELD.
Figures 28-35.

28—30.— Mesoprotriaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oe. 6.
31.— Group of megascleres from a spicule-preparation; magnified 30; phot. Zeiss, planar 20:
a, orthotriaenes; b, mesoprotriaenes; c, large choanosomal amphioxes.

32-35. Cladomes of anatriaenes; magnified 100; phot. Zeiss. apochr. 16, compens. oc. 6.

GES OF THE PACIFIC, I. GEODIDAE. Z | PLATE 35.

Fig. 1-27 Geodia breviana 1. sp.
Fig. 28-35 Geodia mesatriaenella n. sp.

ee Sl

} ‘on
= 7
~:
PLATE 36 :

PLATE 36.

Geodia breviana LeNDENFELD.
Figures 1-12.

1.— Two minute dermal anaclades in siiw protruding from the surface; magnified 300; phot. Zeiss,
apochr. 8, compens. oc. 6.
2-9.— Minute dermal anaclades:
2, 4, 6, 8, entire dermal anaclades; magnified 150; phot. Zeiss, apochr. 8, compens. oc. 6;
3, 5, 7, 9, the cladomes of the dermal anaclades represented to the left of them in Figs. 2, 4, 6, and
8; magnified 400; phot. Zeiss, apochr. 4, compens. oc. 6;
2-5, triaenes:
6, 7, a diaene;
8, 9, a monaene.
10-12.— Minute dermal amphioxes; magnified 150; phot. Zeiss, apochr. 8, compens. oc. 6:
10, a slightly curved one;
11, 12, two angularly bent ones.

Geodia micropora LENDENFELD.
Figures 13-36.

13-17.— Cladomes of mesoproclades; magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6:
13, 15, of mesoprotriaenes;
14, of an irregular mesoclade;
16, of a mesoprodiaene; 4
17, of a mesopromonaene.
18, 19.— Group of asters from a centrifugal spicule-preparation; magnified 1800; u. v. phot. Zeiss, q.
monochr. 1.7, q. oc. 10:
18, focused higher; 19, focused lower;
a, small strongylosphaerasters; b, large oxysphaeraster.
20.— Two small strongylosphaerasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7. q. oc. 10. —
21-23.— Sterrasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
21, a young sterraster;
22, 23, full-grown sterrasters. ;
24-27.— Spicules and groups of such from centrifugal spicule-preparations; magnified 300; phot. Zeiss,
apochr. 4, compens. oc. 6: - q
a, dermal rhabds; b, small strongylosphaerasters; ec, large oxysphaerasters; d, large oxyasters.
28.— Group of small strongylosphaerasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
29.— Part of the surface of the sponge; magnified 3; phot. Zeiss, anastig. 167.
30, 31— Umbilical part of the surface of a full-grown terraster; magnified 1800; u. v. phot. Zeiss, q.
monochr.1.7, q. oe. 10:
30, focused higher; 31, focused lower.

32.— View of the sponge; reduced 1 : 095; phot. Zeiss, anastig. 480 / 412 mm. q
33.— Group of asters from a centrifugal spicule-preparation; magnified 1800; u. v. phot. Zeiss, q. mo
ochr. 1.7, q. oc. 10: ,

a, small strongylosphaeraster; b, large oxysphaeraster. P
34.— Group of asters from a centrifugal spicule-preparation; magnified 1800; u. v. phot. Zeiss,
monochr. 1.7, q. oc. 10:
a, small strongylosphaerasters; b, large oxyaster.
35, 36.— Umbilical part of the surface of a not quite fully developed sterraster; magnified 1800; u.
phot. Zeiss, q. monochr. 1.7, q. oc. 10:
35, focused higher; 36, focused lower.

F THE PACIFIC, I. GEODIDAE. PLATE 36.

(eS Se SS SS SS eS

De ye

Ee

it

vee» ge 1s
WA Se

Fig. 1-12 Geodia breviana x. sp.
Fig. 13-36 Geodia micropora n. sp.

AY

PLATE 37.

Geodia micropora LENDENFELD.
Figures 1-14.

1.— Portion of a radial section through the superficial part of the sponge: magnified 20; phot. Zeiss,
planar 20:
a, cortex.
2 — Portion of a radial section through the superficial part of the sponge; magnified 7.5; phot. Zeiss,
planar 50:
a, cortex; b, large efferent canal-stem.
3 — Portion of a radial section through the superficial part of the sponge; magnified 100; phot. Zeiss,
apochr. 16, compens. oc. 6:

a, dermal membrane; b, tufts of small dermal rhabds in position in the dermal layer; c¢,
sterraster-armour; d, subcortical cavity; e, small dermal rhabds still situated subcorti-
cally; f, large subcortical triaenes.

1-7.— Orthoplagiotriaenes; magnified 50; phot. Zeiss, apochr. 16, compens. oc. 4.
8, 9.— Surface views of thin superficial, paratangential sections in transmitted light; magnified 10;
phot. Zeiss, planar 50:
8, part of an afferent area;
9, part of an efferent area.
10, 11.— Chelotrops; magnified 50; phot. Zeiss, apochr. 16, compens. oc. 4.
12.— A pore-sieve from an afferent area; magnified 50; phot. Zeiss, apochr. 16, compens. oc. 4.
13.— A group of pore-sieves from an efferent area; magnified 50; phot. Zeiss, apochr. 16, compens. oc. 4.
14.— Part of a radial section through the choanosome; congo-red; magnified 100; phot. Zeiss, apochr.
16, compens. oc. 6.

Geodia japonica (Souuas).
Figures 15-30.

15-17.— Cladomes of orthotriaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
15, side-view of an orthotriaene-cladome with terminally irregular clades;
16, side-view of an orthotriaene with regular clades;
17, an orthotriaene-cladome (with the rhabdome broken off) seen from below.
18-21.— Large choanosomal amphioxes; magnified 20; phot. Zeiss, planar 20.
22.— Group of megascleres from a spicule-preparation ; magnified 20; phot. Zeiss, planar 20:
a, ordinary, large choanosomal amphioxes; b, orthotriaene with regular clades; ce, ortho-
triaene with an irregular clade; d, smaller, slender curved amphiox; e, large anatriaene.
23-28.— Orthotriaenes; magnified 20; phot. Zeiss, planar 20:
23-27, with fairly straight rhabdome;
28, with curved rhabdome.
29, 30.— Large anatriaenes; magnified 20; phot. Zeiss, planar 20.

‘

PLATE 37.

GES OF THE PACIFIC, I. GEODIDAE.

22

vopora

odia mic

Fig. 1—14 Ge

fig.

denfeld photographed.

a 4 ¥ " y » om!
. ’ . e f
> : * up
ri fi 7 ‘ = ans

‘ eo - - ts i 1
Bs ,« ms
a “f - 7

i 6
\ 4 ;
é

‘2 \ .

PLATE 38.

Geodia japonica (Souuas).
Figures 1-29.

1-7.— Cladomes of orthotriaenes; magnified 50; phot. Zeiss, apochr. 16, compens. oe. 2.
8.— Side-view of the specimen; reduced 1 : 0.58; phot. Zeiss, anastig. 480 / 412.
9-17.— Cladomes of mesoproclades; magnified 100; phot. Zeiss, apochr. 16, Cos oc. 6:
9, 14, of regular mesoprotriaenes;
10, 12, of regular mesoprodiaenes;
11, of a regular mesoprodiaene with a rudiment of a third clade;
13, of a mesoclade with all three clades reduced;
15, 16, of mesopromonaenes with rudiments of the two other clades;
17, of a mesoprodiaene with one shortened, truncate clade.
1S-29.— Cladomes of large anatriaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
18, 19, of young, large anatriaenes with not fully developed clades;
20-22, of full-grown, large anatriaenes, the distal parts of the clades of which are straight;
23-29, of full-grown, large anatriaenes, the distal parts of the clades of which are more o

curved outwards (sigmaclade).

PLATE 38.

5 OF THE PACIFIC, I. GEODIDAE.

Fig. 1—29 Geodia japanica (Sollas).

Id photographed.

ee aot oe ee

PLATE 39.
“

PLATE 39.

Geodia japonica (Souwas).
Figures 1-41.

1-9.— Small dermal rhabds (amphioxes); magnified 300; phot. Zeiss, apochr. 4, compens.. oc. 6.
10-12.— Groups of small strongylosphaerasters and a single one from a centrifugal spicule-preparation;
magnified 1800; u. v. phot. Zeiss, q. monocehr. 1.7, q. oe. 10.
13.— Group of spicules from a centrifugal spicule-preparation ; magnified 300; phot. Zeiss, apochr. 4,
compens. oc. 6:
a, large oxyasters; b, small strongylosphaerasters; c¢, a minute dermal anaelade.
14-17.— Minute dermal anaclades; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
14, 15, with well-developed cladomes;
16, 17, with reduced cladomes.
18-24.— Large oxyasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
25, 26.— Parts of large oxyasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
27.— Group of asters from a centrifugal spicule-preparation ; magnified 540; u. v. phot. Zeiss, q. mo!
ochr. 6, q. oc. 10: ?
a, large oxyaster; b, small strongylosphaerasters.
28-32.— Sterrasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
28, 29, en face views with the umbilicus near the centre of the upper side;
30-32, profile views with the umbilicus at or near the margin.
33.— A large oxysphaeraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
34, 35.— The umbilical part of the surface of a sterraster; magnified 1800; u. v. phot. Zeiss, q. monoch
1.7, q. oc. 10:
34, focused higher; 35, focused lower.
36, 37.— Two strongylosphaerasters from a centrifugal spicule-preparation ; magnified 1800; u.
phot. Zeiss, q. monochr. 1.7, q. oc. 10:
36, focused lower; 37, focused higher.
38, 39.— Cladomes of minute dermal anaclades; magnified 1000; phot. Zeiss, hom. imm. apochr.
compens. oc. 6.
40, 41.— Large oxysphaeraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oe. 10:
40, focused higher; 41, focused lower.

PLATE 59.

BS OF THE PACIFIC, I. GEODIDAE.

ve

°

ce
ae

ort. es
oo vt

PoLy .

Ae,
/ * fen

K]

oS

, .

Taal a!
& oN
e Woe

<t

(COENEN KE te iy Ce

ee

gry
Oa we a

12

Fig. 1—41 Geodta japonica (Soélas).

Ly
PLATE 40.
>
it 7 ¥ i! 4
Ay oa - i P

PLATE 40.

Geodia ovis LENDENFELD.
Figures 1-30.

1-4.— Orthotriaenes; magnified 10; phot. Zeiss, planar 50.
5.— View of a thick radial slice of the sponge; magnified 1.5; phot. Zeiss, anastig. 480 / 412:
a, spicule-fur; b, cortex; c¢, choanosome.
6-13.— Large amphioxes; magnified 10; phot. Zeiss, planar 50.
14-16.— Sterrasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
14, a side-view;
15, 16, front-views of the umbilical side. ]
17, 18.— The umbilical part of the surface of a sterraster: magnified 1800; u. v. phot. Zeiss, q- monochr.
1.7, q. oc. 10: :
17, focused lower; 18, focused higher.
19-23.— Cladomes of orthotriaenes; magnified 50; phot. Zeiss, apochr. 16, compens. oc. 2.
°24— Part of the surface of a sterraster opposite the umbilicus; magnified 1800; u. v. phot. Zeiss, q.
monochr. 1.7, q. oe. 10.
25.— Part of an afferent pore-sieve; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
26.— Tip of an abnormal spicule with numerous branch-rays; magnified 100; phot. Zeiss, apochr. 16
compens. oc. 6.
27.— Tip of a normal, large amphiox; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
28.— The cut face of the halved specimen; reduced 1 : 0.76; phot. Zeiss, anastig. 480 / 412.
29.— Sterroid; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
30.— The tip of an orthotriaene-clade with abnormally branched axial thread; magnified 100; phot.
Zeiss, apochr. 16, compens. oc. 6.

40),

PLATE

INGES OF THE PACIFIC, I. GEODIDAE.

Fig. 1—80 Geodia

nfeld photographed.

a
=
:
ca

\\

PLATE 41.

Geodia ovis LENDENFELD.

Figures 1-20.

1-20.— Asters and parts of such; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, compens.
1, 2, small strongylosphaerasters;
3, thin-rayed oxyaster;
4, strongylosphaeraster;
5-8, medium strongylasters;
9, group of asters from a centrifugal spicule-preparation;
a, small, thick-rayed aster; b, oxyaster with thin, more cylindrical rays;
10, 11, group of asters from a centrifugal spicule-preparation;
10, focused higher; 11, focused lower; z
a, small thick-rayed aster; b, large thick-rayed oxyaster with one ray bifureate;
12, oxyaster with rays intermediate in thickness; :
13, 14, large strongylaster; : >
13, focused higher; 14, focused lower;
15, large oxyaster with slender rays;
16, large oxyaster with thick rays;
17, part of a large oxyaster with thick rays; ;
18, a ray of a large thick-rayed oxyaster; .)
19, a ray (the left below) of the large thin-rayed oxyaster represented in Fig. 15, focused |
20, a ray of a large thick-rayed oxyaster.

Fig. 1—20 Geodia ovis n. sp.

Phototype by Charles Bellmann, Prague.

%

mere,

12

- _*%

PLATE 41.

con FF « -
_ 7
= *
a

PLATE 42.

_
Te

PLATE 42.

Geodia ovis LENDENFELD.
Figures 1-40.

1.— Radial section through the superficial part of the sponge; magnified 30; phot. Zeiss, planar 20:
a, dermal membrane; b, sterraster-armour; ¢, choanosome; d, a chone.
2.— Radial section through the superficial part of the basal part of the sponge and the spicule-fur;
magnified 10; phot. Zeiss, planar 50:
a, surface of the sponge; b, ¢, d, freely protruding spicules of the spicule-fur: b, large
amphioxes; ¢, small anaclades; d, orthotriaenes.
3.— Group of spicules from a centrifugal spicule-preparation; magnified 200; phot. Zeiss, apochr. 8,
compens. oc. 6:
a, minute dermal amphioxes; b, minute dermal anaclade; ec, asters.
4-7.— Minute dermal amphioxes; magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6.
8.— Radial section through the superficial part of the sponge; magnified 20; phot. Zeiss, planar 20:
a, dermal membrane; b, sterraster-armour; ¢c, choanosome.
9.— Radial section through the dermal layer of the cortex; magnified 100; phot. Zeiss, apochr. 16;
compens. oc. 6:
a, dermal membrane; b, sterraster-armour; ¢, tufts of minute dermal rhabds.
10, 11.— Cladomes of minute dermal anaclades; magnified 1800; u. v. phot. Zeiss, q- monochr. 1.7,
q. oc. 10.
12-17.— Asters and groups of such from centrifugal spicule-preparations; magnified 300; phot. Zeiss,
apochr. 4, compens. oc. 6:
a, large thick-rayed oxyasters; b, large thin-rayed oxyasters; ¢, small strongylasters.
18, 19.— Groups of oxyasters with medium rays from a centrifugal spicule-preparation; magnified 330;
phot. Zeiss, apochr. 4, compens. oc. 6.
20-22.— Groups of asters from a centrifugal spicule-preparation; magnified 540; u. v. phot. Zeiss, q.
monocehr. 6, q. oc. 10: P
a, large thick-rayed oxyasters; b, large thin-rayed oxyasters; ¢, small strongylasters.
23, 24.— Groups of spicules from a centrifugal spicule-preparation; magnified 330: phot. Zeiss, apochr.
4, compens. oc. 6:
a, minute dermal rhabd; b, minute dermal anaclades; e¢, asters.
25.— Small anatriaene; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
26-40.— Cladomes of anatriaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.

|
|

PLATE 42.

NGES OF THE PACIFIC, I. GEODIDAE.

i

32

31

29

Bo) 27 (28

Fig. 1—40 Geodia ovis n. sp.

denfeld photographed.

a a” : J
PLATE 43.

PLATE 43.

Geodia ovis LENDENFELD.
Figures 1-8.

1-8.— Cladomes of mesoproclades and teloclades; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
1, of an irregular mesoprodiaene ;

2, of an irregular plagiotriaene with clades convex towards the rhabdome;

4, of quite regular mesoprotriaenes;

of a quite regular mesopromonaene;

of an irregular mesoprotriaene, with one elongated clade;

, of a prodiaene;

8, of an irregular mesoprotriaene with one clade reduced toa rounded knob, and another much elon-
gated.

Geodia ataxastra LeNDENFELD.

Figs. 9-25, 28-38.— var. angustana LENDENFELD.
Figs. 26, 27. — var. latana LENDENFELD.

9-14.— Cladomes of orthotriaenes (plagiotriaenes) of var. angustana; magnified 50; phot. Zeiss, apochr.
16, compens. oc. 2:
9, of a young one;
10, of an adult, somewhat irregular one;
11-14, of adult regular ones.
15-22.— Orthotriaenes of var. angustana; magnified 20; phot. Zeiss, planar 20.
23, 24.— Groups of megascleres from spicule-preparations of var. angustana; magnified 20; phot. Zeiss,
planar 20:
a, orthotriaene; b, mesoprotriaene; c, anatriaenes; d, large amphioxes.
25.— Part of a radial section of var. angustana; magnified 10; phot. Zeiss, planar 50:
a, cortex; b, choanosome.
26.— An afferent pore-sieve of var. latana; magnified 100; phot. Zeiss, apochr. 16, compens. oe. 6.
27.— Group of megascleres from a spicule-preparation of var. latana; magnified 20; phot. Zeiss, pla
20: ;
a, orthotriaenes; d, large amphioxes.
28.— A group of efferent pore-sieves of var. angustana; magnified 30; phot. Zeiss, planar 20.
29-32.— Large oxysphaeraster of var. angustana; magnified 1800; u. v. phot. Zeiss, q. monoehr. 1.7,
q. oe. 10:
29, focused higher; 30, focused lower; 31, focused still lower; 32, focused lowest.
33, 34.— Groups of small strongylosphaerasters from a centrifugal spicule-preparation of var. angustana
magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
35.— Part of a large slender-rayed oxyaster of var. angustana; magnified 1800; u. v. phot. Zeiss, q.
monoehr. 1.7, q. oc. 10.
36.— Part of a large thick-rayed oxyaster of var. angustana; magnified 1800; u. v phot. Zeiss, q.
monochr. 1.7, q. oc. 10.
37.— Group of spicules from a centrifugal spicule-preparation of var. angustana; magnified 1800; u. v.
phot. Zeiss, q. monochr. 1.7, q. oc. 10:
a, part of a large thick-rayed oxyaster; b, small strongylosphaeraster.
38.— Part of a large thin-rayed aster of var. angustana; magnified 1800; u. v. phot. Zeiss, q. mono
1.7, q. oc. 10.

-
PONGES OF THE PACIFIC, I. GEODIDAE. sein

Fig. 1—8 Geodia ovis n. sp-;
Fig. 9—38 Geodia ataxastra n. Sp.

9—25, 28—38 G. a. var. angustana; 26, 27 G. a, var. latana

dene photographed.

Phatntvne by Charles Bellmann, Prague.

+
:
co

PLATE 44.

Geodia ataxastra LmeNDENFELD.

Figs. 1-12, 14-49.— var. angustana LENDENFELD.
Fig. 13. — var. latana LENDENFELD.
1-14.— Cladomes of mesoproclades; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
1—12, 14, of var. angustana;
13, of var. latana;
1, 2, of mesoprotetraenes;
3-11, of fairly regular mesoprotriaenes;
12, of a mesoprodiaene;
13, 14, of irregular mesoprotriaenes with one or more clades reduced in length and terminally
rounded.
15-22.— Cladomes of large anatriaenes of var. angustana; magnified 100; phot. Zeiss, apochr. 16, com-
pens. oe. 6.
23.— Cladome of an anatriaene-derivate with clades reduced to small knobs of var. angustana; magni-
fied 100; phot. Zeiss, apochr. 16, compens. oc. 6.
24.— Part of a section through the choanosome of var. angustana; haematoxylin; magnified 200;
phot. Zeiss, apochr. 8, compens. oc. 6.
25.— View of a cluster of specimens of var. angustana growing together on a stone; natural size; phot.
Zeiss, anastig. 480 / 412.
26.— Radial section through the superficial part of var. angustana; magnified 30; phot. Zeiss, planar
20:
a, sterraster-armour; b, radial cortical canal; ce, choanosome.
27.— Thick radial section through the superficial part of var. angustana; magnified 20; phot. Zeiss,
planar 20:
a, surface of the sponge; b, protruding mesoproclades.
28.— Group of spicules from a centrifugal spicule-preparation of var. angustana; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
b, small strongylosphaerasters; ¢, large oxyaster; d, minute dermal anaclades.
29. 30.— Groups of asters from a centrifugal spicule-preparation of var. angustana; magnified 540; u. v.
phot. Zeiss, q. monochr. 6, q. oc. 10:
b, small strongylosphaerasters; ec, large thick-rayed oxyasters; e, large thin-rayed oxyaster.
31.— Group of spicules from a centrifugal spicule-preparation of var. angustana; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
a, minute dermal amphioxes; b, small strongylosphaerasters.
32.— A minute dermal amphiox of var. angustana; magnified 300; phot. Zeiss, apochr. 4, compens.
oc. 6.
33.— Group of spicules from a centrifugal spicule-preparation of var. angustana; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
a, minute dermal amphioxes; b, small strongylosphaerasters; ¢c, large thick-rayed oxyasters.
35.— Groups of asters from a centrifugal spicule-preparation of var. angustana; magnified 540;
u. v. phot. Zeiss, q. monochr. 6, q. oc. 10:

34

,

b, small strongylosphaerasters; e, large thick-rayed oxyasters; e, large thin-rayed oxyasters..

36-38.— Sterrasters of var. angustana; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
36, side-view.
37, 38, views of the umbilical face.
39.— Large oxyaster with medium rays; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
40.— Group of spicules from a centrifugal spicule-preparation of var. angustana; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:

a, minute dermal rhabds; b, small strongylosphaerasters; c, large thick-rayed oxyasters; d, —

minute dermal anatriaene.

41, 42.— Cladomes of anatriaenes of var. angustana; magnified 300; phot. Zeiss, apochr. 4, compens.
oc. 6.

43-45.— The umbilicus of a sterraster and the adjacent parts of its surface of var. angustana; magnified
1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oe. 10:

43, focused high; 44, focused intermediate; 45, focused low.

{6-49.— Minute dermal anaclades of var. angustana; magnified 300; phot. Zeiss, apochr. 4, compens.

oe. 6.

————

eee ae

ES OF THE PACIFIO, I. GEODIDAR.
PLATE 44.

Fig. 1—49 Geodia ataxastra n. sp.
1—12, 14—49 GC. a. var. angustana, 13 ©. a. var. Jatana.

Phototype by Charles Bellmann, Prague.

$
|

e

PLATE 45.

Geodia acanthylastra LeENDENFELD.
Figures 1-39. >

~ 1-7.— Cladomes of mesoprotriaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
8-15.— Cladomes of anatriaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.
16.— The cut face of a halved specimen; magnified 3; phot. Zeiss, anastig. 167.
17-19.— Groups of spicules from spicule-preparations; magnified 20; phot. Zeiss, planar 20:
a, large amphioxes; b, plagiotriaenes; c, anatriaenes.
2.— Dermal amphioxes:
20-21, magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6;
22, magnified 300; phot. Zeiss, apochr, 4, compens. oc. 6.
23-25.— Groups of asters from centrifugal spicule-preparations; magnified 300; phot. Zeiss, apochr. 4,
compens. oc. 6:
a, large acanthtylasters; b, small strongylosphaerasters.
26, 27.— Groups of asters from a centrifugal spicule-preparation; magnified 540; u. v. phot. Zeiss, q.
monoehr. 6, q. oc. 10:
a, large acanthtylasters; b, small strongylosphaerasters. 4
28.— Part of a section through the ee aniline-blue; magnified 200; phot. Zeiss, apochr. 8,
compens. oc. 6:
a, young sterraster; b, flagellate chambers.
29.— View of the sponge; magnified 1.5; phot. Zeiss, anastig. 167.
30-38.— Plagiotriaenes, dichotriaenes, and cladomes of these spicules; magnified 50; phot. Zeiss, apochr.
16, compens. oc. 2: ¥
30, 35, cladomes of somewhat irregular plagiotriaenes with one lengthened clade;
31, 32, regular plagiotriaenes;
33, a plagiotriaene with one bifurcate clade;
34, cladome of a regular plagiotriaene;
36,37, dichotriaenes (possibly foreign) ; i
38, a plagiotriaene with a shortened, terminally rounded rhabdome.
39.— Radial section through the superficial part of the sponge, magnified 100; phot. Zeiss, apochr. 16,
compens. oc. 6:
a, dermal membrane; b, sterraster-armour; ec, tufts of dermal rhabds; d, dermal cavities.

20-2

RES OF THE PACIFIC, I. GEODIDAE.

Fig. 1—39

nfeld photographed.

74. 3-- Ratlesann Pracie.

ae ee ee ‘2 ee a ee on

>. al
PLATE 46.
eT : ~
ww a £

PLATE 46.

Geodia acanthylastra LENDENFELD.

Figures 1-21.

1, 2.— Group of small strongylosphaerasters; magnified 1800, u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
1, focused higher; 2, focused lower. 1
3~-7.— Small strongylosphaerasters and groups of such from centrifugal spicule-preparations;
1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
8, 9.— Sterrasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
10, 11.— Large oxysphaerasters; magnified 1800; u. v. phot. Zeiss, q. monoehr. 1.7, q. oc. 10:
10, focused lower; 11, focused higher.
12, 13.— Large oxysphaeraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
12, focused lower; 13, focused higher.
14-16.— The umbilicus of a sterraster and the adjacent parts of its surface; magnified 1800; u. v. ph
Zeiss, q. monochr. 1.7, q. oc. 10:
14, focused high; 15, focused intermediate; 16, focused low.
17-19.— The umbilicus of a sterraster and the adjacent parts of its surface; inagnised 1800; u. v.
Zeiss, q. monochr. 1.7, q. oc. 10:
17, focused high; 18, focused intermediate; 19, focused oo,
20.— Radial section through the superficial part of the sponge; magnified 30; phot. Zeiss, planar 20:
a, protruding minute dermal rhabds; b, dermal membrane; e¢, sterraster-armour; d, pr
truding mesoproclades; e, choanosome.
21.— A pore-sieve; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.

INGES OF THE PACTFIO, I. GEODIDAE. PLATE 46.

Fig. 1—21 Geodia acanthtylastra n. sp.

enfeld photographed.

:

PLATE 47.

Geodia acanthtylastra LeNDENFELD.

Figures 1-8.

1, 2.— Large acanthtylaster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
1, focused higher; 2, focused lower.
3.— Group of asters from a centrifugal spicule-preparation; magnified 1800; u. v. phot. Zeiss, q.
monochr. 1.7, q. oc. 10:
a, small strongylosphaeraster; b, large acanthtylasters.
4-6.— Acanthtylasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
7.— Group of asters from a centrifugal spicule-preparation: magnified 1800; u. v. phot. Zeiss, q. mor
ochr. 1.7, q. oc. 10:
a, small strongylosphaeraster; b, large acanthtylaster.
§.— Large acanthtylaster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q oc. 10.

Geodia lophotriaena LenpENFELD.
Figures 9-36.

9-21.— Side views of lophotriaenes and plagiotriaenes; magnified 50; phot. Zeiss, apochr. 16, com
oc. 2:
9-12, 14, 19, lophotriaenes with more or less irregular cladomes and long simple, pointed rhabd
13, 15, 16, 18, dichotriaenes with regular cladomes and long pointed rhabdomes; a
17, plagiotriaene;
20, lophotriaene with bifureate rhabdome;
21, plagiotriaene with shortened and terminally rounded rhabdome.
22.— The sponge seen from above; magnified 3.5; phot. Zeiss, anastig. 167. ;
23.— Radial section through the superficial part of the sponge; magnified 100; phot. Zeiss. apo chr
compens. oc. 6: ;
a, dermal membrane; b, sterraster-armour; ec, choanosome.
24.— Radial section through the superficial part of the sponge; magnified 10: phot. Zeiss, planar 50 :
a, cortex; b, choanosome.
25-33.— Cladomes of dicho- and other lophotriaenes seen from below (the rhabdome is directed
cally upwards); magnified 50; phot. Zeiss, apochr. 16, compens. oc. 2:
25-29, cladomes of dichotriaenes;
30, 33, cladomes of lophotriaenes with one trifureate and two bifurcate clades;
31, cladomes of a lophotriaene with one bifureate, one trifureate and one quaterfurcate cla
32, eladome of a lophotriaene with one quaterfureate and two trifureate clades.
34.— Mesoprotriaene; magnified 50; phot. Zeiss, apochr. 16, compens. oe. 2.
35, 36.— Cladomes of anatriaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6.

SPONGES OF THE PACIFIC, L GEODIDAE. PLATE 47.

Fig. 1—8 Geodia acant.

o
Fig. 9—36 Geodia lop/

i Lendenfeld photographed.
Phatatvne hy Charles Bellmann, Prague-

:
=

PLATE 48.

Geodia lophotriaena LeNDENFELD.
Figures 1-34.

1, 2.— Large oxyaster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
1, focused higher; 2, focused lower.
3-7.— Large amphioxes; magnified 50; phot. Zeiss, apochr. 16, compens. oc. 2.
8, 9.— Minute dermal amphioxes; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
10.— Group of spicules from a centrifugal spicule-preparation; magnified 300; phot. Zeiss, apochr. 4,
compens. oc. 6: ig
a, minute dermal amphiox; b, sphaerasters.
11.— Large oxyaster; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
12.— Group of asters from a spicule-preparation; magnified 300; phot. Zeiss, apochr. 4, compens. oc. oe
a, sphaeraster; b, large oxyasters.
13-15.— Strengylosphaerasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
16—-18.— Large oxyaster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
16, focused high; 17, focused intermediate; 18, focused low.
19.— Oxysphaeraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
20.— Strongylosphaeraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
21.— A ray of a large oxyaster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
22, 23.— Strongylosphaerasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
24-26.— Strongylosphaeraster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
24, focused high; 25, focused intermediate; 26, focused low. 2
27-29.— Sterrasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
30, 31— The umbilicus and adjacent parts of the surface of a sterraster; magnified 1800; u. v. ph
Zeiss, q. monochr. 1.7, q. oc. 10:
30, focused lower; 31, focused higher.

ochr. 6, q. oc. 10:
a, sphaerasters; b, large oxyasters.
33, 34.— Part of the surface of a sterraster opposite the umbilicus; magnified 1800; u. v. phot.
q. monochr. 1.7, q. oc. 10.
33, focused lower; 34, focused higher.

SPONGES OF THE PACIFIC, I. GEODIDAE.

PLATE 48.

Ht. Sp.

Geodia lophotriaena

Fig. 1—34

Lendenfeld photographed,

Memoirs of the Museum of Comparative Zodlogy
AT HARVARD COLLEGE.

Wot xile INov 2:

REPORTS ON THE SCIENTIFIC RESULTS OF THE EXPEDITION TO THE
EASTERN TROPICAL PACIFIC, IN CHARGE OF ALEXANDER AGASSIZ,
BY THE U. S. FISH COMMISSION STEAMER “ALBATROSS,” FROM
OCTOBER, 1904, TO MARCH, 1905, LIEUT. COMMANDER L. M. GARRETT,
U.S. N., COMMANDING, AND OF OTHER EXPEDITIONS OF THE “ALBA-—
TROSS,” 1888-1904.

XXII.

THE SPONGES.
2. THE ERYLIDAE.

By ROBERT VON LENDENFELD.

WITH EIGHT PLATES.

[Published by permission of George M. Bowers, U. S. Commissioner of Fish and Fisheries.]

CAMBRIDGE, U. S. A.:

Printed for the Museum.
SEPTEMBER, 1910.

rs tae

+7}

.
a)

By Oe

CONTENTS.

t PORTS on the Scientific Results of the Expedition to the Eastern Tropical
ic, in Charge of ALEXANDER AGassiz, by the U. S. Fish Commission Steamer
ross,” from October, 1904, to March, 1905, Lieut. Commander L. M. Garrert,
N., Commanding, and of other Expeditions of the “ Albatross,” 1888-1904. XXII.
Sponers. 2. Tue Eryiipar. By Roperr von LENDENFELD. 63 pp. 8
September, 1910.

E. oxyaster (Plate 3, foe 29-35; Plate is one 1-43) Z
KE. sollasii (Plate 1, figs. 1-48; Plate 2, figs. 1-26; Plate 3, figs. 1- 28)
__ E. rotundus 5 : q b
var. megarhabda (Plate 5, Hey 18-23, 32; Plate 6, Ge “, 18, 24, 33-35;
Plate 7, figs. 22-30, 57-73; Plate 8, fig. 13)
var. typica (Plate 5, figs. 1-4, 11-17, 30; Plate 6, figs. 15-17, 25, 27, 30-32:
_ Plate 7, figs. 16-21, 46-56)
var. cidaris (Plate 5, figs. 5, 26-28, 31; Plate ", ‘eh 1-10, 42-45, 5, 76, 79;
Plate 8, fig. 14)
BE. ealiculatus (Plate 5, figs. 6-10, 24, 25, 29; Plate 6, Son 1- 13, 19-23, 26, 28, 29:
Plate 7, figs. 11-15, 31-41, 74, 77, 78, 80; Plate 8, figs. 1-12, 15-20)

‘ERAL SYSTEMATIC ACCOUNT OF THE GENER: fa SPECIES, AND VARIETIES
FROM THE PACIFIC OCEAN : : : : é
* Eryius

Summary of the species found in the Pacific Ocean
. placenta
. monticularis
. decumbens
- oxyaster
. nobilis
. sollasii
. rotundus
var. megarhabda
var. typica
var. cidaris
E. caliculatus

Wom M omc oles]

Pace
267

267
268
268
272
290

309

309

310

310

316
316
316
317
317
317
318
318
318
319
320
320
320
321

322
323

I. DESCRIPTION OF THE SPECIES COLLECTED BY THE
“ALBATROSS.”

Erylidae.

Tetraxonia with rhabd and teloclade megascleres, and a superficial armour
composed of aspidasters*’ and microrhabds. Euasters are always present in the
choanosome.

Sollas * divided the family Geodidae (Geodiadae) of Gray * into the two
subfamilies Erylina and Geodina, and placed the geodid genera Erylus, Carninus,
“and Pachymatisma in the former. Later authors, I‘ among others, have not
retained these subfamilies and have placed the three genera named, together
with the typical genus Geodia and its allies,in the family Geodidae. The
genera Caminus and Pachymatisma are not represented in the collections of the
“Albatross,”’ so that I have not been able to give any new data concerning their
systematic position during this work. The genus Erylus on the other hand is
well represented, and the examination of the ‘‘ Albatross” material has shown
that, as I have already stated in the first part of this monograph (ante p. 17),
Erylus differs very considerably from Geodia the typical genus of the Geodidae.
I therefore now not only revert to Sollas’s (loc. cit.) original proposition of divid-
ing the family Geodidae into two subfamilies but propose to go even farther and
to place Erylus in a separate family: the Erylidae.

The question whether other genera (Pachymatisma, Caminus) should also
be placed in this new family I shall not, for the reason given above, discuss here,
and I leave them, for the present at least, in the Geodidae; the description of

'The aspidasters are those spicules of the Erylidae which were previously termed sterrasters.
They are distinguished from the sterrasters of the true Geodidae by passing, during their development,
through a stage with perfectly smooth surface which does not occur in the development of the sterr-
asters of Geodia and its allies, and also by their flattened, disc-like shape.

2W.J.Sollas. Tetractinellida. Rept. voy. ‘‘Challenger,”’ 1888, 25, p. exlvii.

3 J. E. Gray. Notes on the arrangement of sponges... . Proc. Zool. soc. London, 1867, p. 492.

4R. v. Lendenfeld. Tetraxonia. Tierreich, 1903, 19, p. 84.

267

268 ERYLUS OXYASTER.

the Pacific species, not in the ‘‘ Albatross” collection, will be found in the first

part of this monograph.
The family Erylidae thus comprises the single genus Erylus.

ERYLUS Gray.

With uniporal afferents and uniporal efferents or larger oscules. Without
ana- or protriaenes.

There are in the ‘‘ Albatross” collection twenty-two specimens which belong
to four species, one of which is divided into three varieties. All the species and
varieties are new.

Erylus oxyaster, sp. nov.
Plate 3, figs. 29-35; Plate 4, figs. 1-43.

I establish this species for a specimen obtained in the Galapagos Islands.
Its asters are oxyasters and to this the name refers.

The single, somewhat fragmentary specimen (Plate 4, fig. 24) is 30 mm.
in maximum diameter and consists of two rounded lobose parts, one of which
is broad cushion shaped, the other slender, digitate. The surface is smooth
and bears numerous small afferent pores. These are quite uniformly dis-
tributed and 0.7-1 mm. apart. On the summit of the broader of the two lobes
an irregularly circular oscule, 1.8 mm. in diameter, is situated.

The colour of the surface of the sponge (in spirit) is brown. A small part
of it, which was probably sheltered from the light, is much lighter than the rest.
The interior is light greenish yellow.

The superficial part of the body is differentiated to form a cortex 450-650 pz
thick. This is composed of two layers, an outer layer, 75-120 y thick, occupied
by microrhabds, and an inner layer 360-560 thick, occupied by aspidasters.

Canal-system. Many of the afferent pores appear to be quite closed. The
open ones (Plate 4, fig. 25) are circular and surrounded by fine sphincter-mem-
branes in which numerous more or less radially disposed microrhabds are
situated. These pores are 30-60 «4 wide. They lead into radial canals which
traverse the cortex and open out into subcortical cavities the radial diameters
of which are usually greater than the paratangential.

The skeleton consists of regular rhabd megascleres, irregular derivates of
these, microrhabds, teloclades, aspidasters, and oxyasters. The rhabd mega-
scleres are for the most part amphioxes, but a few styles have also been
observed, These rhabds and their irregular derivates form bundles which

ERYLUS OXYASTER. 269

extend radially from the base of the sponge to the cortex and abut vertically on
the latter. The microrhabds occupy in dense masses the outer layer of the
cortex. The superficial ones are situated paratangentially, the deeper ones
mostly obliquely or radially. The teloclades are mostly regular dichotriaenes,
but some irregular dichotriaene-derivates and simple plagiotriaenes have also
been observed. The cladomes of these teloclades extend paratangentially; just
below the cortex their rhabdomes are directed radially inward. The oxyasters
forma series from small many-rayed to large few-rayed ones. As, however, the
asters of medium diameter and ray-number are not nearly so numerous as the
large few-rayed and small many-rayed ones, this series does not appear uniform
and large few-rayed and small many-rayed oxyasters can readily be distin-
guished. The large few-rayed oxyasters are quite uniformly scattered through-
out the choanosome; the small many-rayed ones on the other hand, although
also present in all parts of the choanosome, are much more numerous in the
subcortical region, particularly in the roofs of the subcortical cavities and the
walls of the cortical canals, than elsewhere. The aspidasters, which occupy
the inner layer of the cortex, are rather irregularly arranged. They exhibit
hardly a trace of a paratangential orientation.

The regular amphiozes (Plate 4, figs. 6-9) are isoactine, gradually attenuated
towards the ends, and usually rather sharply pointed (Plate 4, figs. 6-8), more
rarely blunt (Plate 4, fig. 9). They are straight (Plate 4, fig. 6) or slightly and
uniformly curved (Plate 4, figs. 7, 9), exceptionally abruptly bent in the middle
(Plate 4, fig. 8). The amphioxes are 1.8-2.9 mm. long and 60-85 y thick.

The rare styles (Plate 4, figs. 10, 11) are slightly curved, simply rounded
off at one end and gradually attenuated towards the other, which is usually
very blunt. They are 1.9-2.3 mm. long and 60-105 y thick.

The irregular derivates of the rhabd megascleres (Plate 4, figs. 12-19) have
similar dimensions to the regular rhabds. They appear as more or less curved
amphioxes, either strongly angularly bent near one end, or provided with one
or more branches. The angle in the angularly bent forms (Plate 4, figs. 12-13)
is 15-95°. The branched forms bear one (Plate 4, fig. 15) or, more frequently,
two (Plate 4, figs. 16, 19) or three branches (Plate 4, figs. 17, 18), which arise
either from the same part (Plate 4, figs. 16, 19) or from different parts of the
shaft (Plate 4, figs. 17, 18). The branches are always very much shorter than
the shaft, rarely over 400 » long, straight, conical, and terminally either pointed
(Plate 4, figs. 16, 17, 18 the upper right one, 19) or rounded (Plate 4, figs. 15, 18
the upper left and the lower one). The angle at which they arise is very vari-

270 ERYLUS OXYASTER.

able. Sometimes two similar branches lie opposite each other in a straight
line (Plate 4, fig. 17 the two lower ones, 19). The axial threads of the shaft and
the branches are either joined in a regular manner, or slight irregularities occur
at their junction. The most remarkable of these were observed in the two spic-
ules represented on Plate 4, figs. 15 and 18. The axial thread of the single
branch of the former and that of the lowest one of the latter do not reach down
to the axial thread of the shaft, which passes the junction unaltered, but termi-
nate with a bulbous thickening at a distance of about 3 « from it.

The microrhabds (Plate 3, figs. 29-31, 32a, 35a; Plate 4, figs. 28-33a) are
more or less curved, centrotyle amphistrongyles, and generally isoactine. The
curvature is either uniform or one or both ends are also abruptly bent in the
direction of the curvature. The isoactine microrhabds are usually 31-47 y long,
but occasionally very much larger ones, up to 93 sin length, are observed (Plate
4, fig. 31). The ordinary microrhabds are, near the centre (tyle), 3.5-4.5
thick; the tyle measures 4.5-5.5 y, usually about 1 » more than the adjacent
parts of the spicule, in diameter. Towards the rounded ends the actines taper
gradually to about 2 ~. The rare giant microrhabds are thicker in proportion
to their greater length. The centre of the spherical tyle usually lies in the axis
of the spicule; sometimes, however, it is eccentric and then the tyle bulges much
more on one side than on the others. In some microrhabds one actine is reduced
in length; these appear as anisoactines. In a few one of the actines is completely
suppressed; these appear as blunt tylostyles. The anisoactine microrhabds are
shorter and also somewhat thicker than the ordinary isoactine ones.

The rare plagiotriaenes (Plate 4, fig. 20) have a straight, conical rhabdome,
about 0.9 mm. long and, at the cladomal end, 75-90 y thick. The clades are
nearly straight, about 0.7 mm. long, and enclose angles of 109-112° with the
rhabdome. The breadth of the cladome is 1.3-1.4 mm.

The regular dichotriaenes (Plate 4, figs. 1-5, 21-23) have a fairly straight,
conical rhabdome, 0.6-1.6 mm. long and, at the cladomal end, 70-105 y thick.
The main clades are straight, 250-400 long and enclose angles of 109—120°
with the rhabdome. The end clades are conical, pointed, and straight or, more
rarely, slightly curved inwards (Plate 4, fig. 5, below), and 50-450 » long. The
cladome is 0.9-1.5 mm. broad.

In the rare irregular dichotriaene-derivates either the clades are reduced in
number or the rhabdome reduced in length, or both. Forms with two and
with only one clade (dichodiaenes and dichomonaenes) have been observed.
The reduced rhabdomes are cylindrical, rounded at the acladomal end, 200-

ERYLUS OXYASTER. Dall

600 » long, and about as thick as the rhabdome of the regular dichotriaenes.
In such rhabdomes the axial thread terminates some distance from the aclad-
omal end, in the centre of its hemispherical surface. In the dichotriaene-
derivates with reduced clade number, the central parts of the axial threads of
both the rhabdome and the clades usually exhibit considerable irregularities.

The oxyasters (Plate 3, figs. 32e, d, 33d, 34c, 35b, d; Plate 4, figs. 26d, 27b,
28c, 29f, 30d, f, 32-34e, 38-40) usually have a slight central thickening. This
is most clearly discernible in the monactine (Plate 4, fig. 27b) and diactine forms
(Plate 3, fig. 35b; Plate 4, fig. 28c). The rays are from one to twenty or more
in number, concentric, regularly distributed, and usually equal in size. They
are perfectly smooth (Plate 4, figs. 38-40), conical, and pointed. Very rarely
one or two rays are reduced in length, much shorter than the others, and termi-
nally rounded (Plate 3, fig. 34c). The properly developed rays of the small
- oxyasters (Plate 4, figs. 39-40) are uniformly attenuated towards the pointed
end, those of the large ones (Plate 4, fig. 38) attenuated more rapidly in their
distal than in their proximal part. The rays are 6-55 long and, at the base,
0.8-4.5 » thick. The whole aster measures 10-90 y in diameter. The ray-
number is, as the following table shows, in inverse proportion to the size of the
rays and, apart of course from the monactines, of the whole aster.

, : | 14-2)
Number of rays 1 2-3 45 | 6-9 | 10-13
or more

eee e micter> of 57 48-90 45-75 32-58 | 1523 | 10-17
aster, #2 |

Length of rays
measured from 55 29-50° | «=: 27-40 19-38. | 9-14 6-10
eentrum, 4 |

Se oe enees cf 4.5 3-4.5 3-45 17455 DAE 0.8-1.5
Tays, /t | |

As mentioned above there is a kind of gap in the series of these oxyasters,
produced by the scarcity of forms 20-35 in diameter with from nine to eleven
rays.

Most of the aspidasters (Plate 4, figs. 35-87, 41-43) are stout oval dises.
Very rarely roundish (Plate 4, fig. 41) or irregular aspidasters have been observed.
The ordinary oval aspidasters are 208-243 y long, 125-150 y broad and in the
middle 30-40 « thick. Towards the margin they thin out gradually. The
average proportion of length to breadth to (central) thickness is 100 : 63.3 : 19.7.

272 ERYLUS SOLLASIL.

The umbilical pit is more or less circular in outline, 30-50 » broad, and
about 15 « deep. Its walls are usually quite smooth. The rest of the surface
of the adult aspidaster is covered with protruding rays, often somewhat irregu-
larly distributed, and 1.4-1.7 y» thick. These rays bear terminal verticils of
usually six to eight lateral spines (Plate 4, figs. 36, 37). The youngest aspid-
asters observed were about 55 # long and appeared as oval, radially striated
dises with deeply serrated margins. In a more advanced stage these spicules
are smooth dises with slightly undulating margins. On the faces and the
margin of such, small protuberances then make their appearance and these
grow out to form the protruding rays of the adult aspidasters above described.
The centre of the aspidaster is occupied by a rosette-shaped granule about 1.7 4
in diameter. Viewed in profile the adult aspidasters show a distinct stratifica-
tion. The limits of the layers are smooth and nearly parallel to the two faces.
Viewed en face they show fine straight striae radiating from the central granule.

This sponge was trawled in the Galapagos Islands, Station 2809, on April
4, 1888; 0° 50’S., 89° 36’ W.; depth 82 m. (45 f.); it grew on a bottom of gray
sand; the bottom temperature was 23.4° (74.1° F.).

The only known species which appears to be allied to the sponge described
above is #. polyaster Lendenfeld from the Agulhas bank, South Africa. From
this it differs, apart from minor peculiarities, by the aspidasters, which are,
absolutely and relatively more than three times as thick in ZF. polyaster

as in 2. oxyaster, a difference, of course, quite sufficient for specific distinction.

Erylus sollasii, sp. nov.

Plate 1, figs. 148; Plate 2, figs. 1-26; Plate 3, figs. 1-28.

I establish this species for seven specimens obtained at five stations among
the Hawaiian Islands. Among the known species of Erylus the one named after
me by Sollas appears to be its nearest ally. I therefore return my distin-
guished friend’s compliment by naming this new species after him.

The two specimens from Station 3847 are both smal]; one is partly light,
partly dark in colour; in the other the whole of the surface is dark. Both the
specimens from Station 3848 are large and whitish. The specimens from Sta-
tions 3849 and 4055 are large and dark. The specimen from Station 4062 is
middle sized and light coloured.

The two specimens from Station 3848 are in every way identical; all the
others differ to some extent from these and from each other. We have to deal

therefore with six different forms. As is shown below, these six forms fall into

ERYLUS SOLLASII. 273

three groups, which I consider as three distinct races. These I denominate I,
IJ, and III. Race I contains four forms; the races II and III one each. The
four forms of race I are designated A, B, C, and D. To race I, form A, belongs
the partly light and partly dark specimen from South Molokai (Station 3847);
to race I, form B, the small, entirely dark specimen from South Molokai (Sta-
tion 3847); to race I, form C, the two large whitish specimens from South
Molokai (Station 3848); to race I, form D, the middle-sized, light-coloured
specimen from northeast Hawaii (Station 4062); to race II the large dark
specimen from South Molokai (Station 3849); and to race III the large dark
specimen from northeast Hawaii (Station 4055).

Shape and size. The smallest specimen is the entirely dark one from
Station 3847 (race I, form B). It is massive, lobose, covered with small but
relatively high gyriform ridges, and measures 18 by 9mm. It has one oscule,
about 1 mm. wide, which lies on the summit of a slight elevation. The partly
light, partly dark specimen also from Station 3847 (race I, form A) is an irregu-
lar mass, with still higher gyriform ridges and measures 24 by 17 by 13 mm.
It has two oscules, the larger 1.8 mm. in diameter, and numerous small pores,
which are scattered over its surface. The specimen from Station 4062 (race I,
form D) is meandric, has small, but relatively quite high, gyriform ridges and
measures 36 by 27 by 18 mm. Here and there small apertures are observed on
its surface. Of the two specimens from Station 3848 (race I, form C) one (Plate
1, fig. 27) is meandric, while the other appears as a mass with lobose, somewhat
digitate processes 7-12 mm. thick. The former measures 60 by 35 by 20 mm.,
the latter is only 51 mm. long. A few oscules, up to 1.5 mm. wide, are situated
on the elevations, and groups of small pores are scattered over the other parts
of the surface. The specimen from Station 3849 (race IT) (Plate 1, fig. 28)
appears as an aggregate of vertical digitate parts about 15 mm. thick, which
are joined for the greater part of their length. It is 73 mm. high, 64 mm. long,
and 42 mm. broad. From the surface ridges protrude which are about 1 mm.
high and 2.5 mm. apart. These ridges are somewhat curved and most of them
extend longitudinally, parallel to the vertical axis of the sponge. On or near the
summit of each digitate protuberance an oscule is situated. The largest of these
oscules is oval and measures 3.2 by 2 mm.; the other oscules are 1—-2.5 mm. wide
and more or less circular. On the sides of the digitate parts groups of small
pores are met with. At one place there is a group of six much larger apertures
about 1 mm. wide; here the sponge seems to have been injured some time before
its capture. The specimen from Station 4055 (race III) (Plate 3, fig. 27) is an

274 ERYLUS SOLLASII.

irregular lobose mass with small gyriform ridges on parts of its surface and
measures 78 by 62 by 40 mm. On or near the summits of some of the lobes
oscules are observed. The largest are oval and measure 4 by 2mm. The other
parts of the surface are occupied by groups of small pores.

In regard to their colour the (spirit) specimens differ to a considerable
extent. The two specimens from Station 3848 (race I, form C) are for the most
part yellowish white, only some of the protruding parts having a purplish brown
tinge. The specimen from Station 4062 (race I, form D) is dirty light brownish
gray. The lower, basal portion of the partly dark, partly light-coloured speci-
men from Station 3847 (race I, form A) is light gray, the upper portion dark
chestnut-brown. In the entirely dark specimen from the same Station (race I,
form B), the surface is dark chestnut-brown throughout. The specimen from
Station 3849 (race IT) is dirty brownish purple, some parts of its surface being
considerably darker than others. The specimen from’ Station 4055 (race IIT)
is rather dark purple-brown.

The superficial part of the body is differentiated to form a cortex (Plate 1,
fig. la). This is 100-250 » thick and composed of an outer, middle, and inner
layer. The outer, dermal layer is generally quite insignificant and on parts of
the surface of most of the specimens absent altogether, probably rubbed off.
In those parts of the cortex of race II where it is most highly developed, it
attains a thickness of 80 4. This layer is rich in microrhabds, but contains no
fibres and no aspidasters. The middle layer is from 83 4 (in parts of the cortex
of race IT) to 210 «(in parts of the cortex of race IIT) thick and oceupied by
dense masses of aspidasters. The inner layer is often insignificant and hardly
to be made out in the sections. In race I, form C, it attains in places a thick-
ness of 20-30 4. It is composed of paratangential fibres and contains a few
granular cells and groups of granules.

Granular cells, oval to spherical in shape and 12-18 » long, are abundant
in most of the specimens. These cells are most numerous just below the cor-
tex and here often quite densely packed (Plate 1, fig. 3). They also occur
scattered in the interior of the choanosome. These cells are filled with granules
of fairly equal size. In the granular cells of pale parts of the sponge, particu-
larly in the region overgrown with symbionts, the granules are colourless. In the
subcortical and cortical granular cells of the dark and exposed parts the granules
are brown. Both the colourless and the brown granules stain deeply with
aniline-blue. As mentioned above a few granular cells and groups of granules
also occur in the inner layer of the cortex. These groups of granules are

ERYLUS SOLLASII. 275

massive, flattened, or drawn out so as to appear like strings of beads. Some of
the granules forming these groups are similar to the granules in the granular
cells, others are larger. They seem to be remnants of disintegrated granular cells.

Canal-system. The afferent apertures, which, as mentioned above, generally
form groups on the surface, are uniporal. They are always circular but they
differ very considerably in size, their diameter varying between 70 and 520 p.
The largest pores were observed in race I, form C, and in race III. From each
pore a canal leads vertically downwards. This canal penetrates the cortex and
opens out into one of the subcortical cavities which underlie the poral areas of
the cortex. Its proximal opening into the subcortical cavity is surrounded by a
chonal sphincter usually more or less contracted, but only rarely completely
closed. The canals leading down from small pores are distended, those leading
down from large pores constricted below the entrance. I am inclined to ascribe
these differences and also, to a great extent, the differences in the width of the
pores themselves, to differences in the degree of contraction. The afferent canals,
which arise from the subcortical cavities and lead down into the choanosome,
are not particularly wide and divide into numerous narrow branch-canals which
supply the flagellate chambers. The latter (Plate 1, fig. 4) are spherical or
somewhat compressed in the direction of their axis, and measure about 20 y
in diameter. The efferent canals join to form lacunose cavities, which some-
times attain a very considerable width (Plate 1, fig. 1). They open out on the
surface with the oscules described above.

Skeleton. Loose strands of rhabd megascleres traverse the choanosome.
Their distal portions extend more or less radially and they terminate just below
the cortex. Triaenes with radial, centripetally directed rhabdomes are quite
abundant in the subcortical layer. In some of the sections a few spicules of
this kind have also been found in the interior of the choanosome. I do not
consider that as their natural position, however, and believe that these triaenes
were brought there from the subcortical layer in cutting the section. No mega-
scleres protrude beyond the surface, and this is entirely destitute of a spicule-
fur. Microrhabds occupy in large numbers the outer layer of the cortex and are
found in smaller numbers also in the choanosome, chiefly in its distal parts. In
the outer, cortical layer of the cortex these microrhabds are not regularly ar-
ranged; some aresituated paratangentially, others obliquely, and others radially.
The middle layer of the cortex is occupied by dense masses of aspidasters,
most of which are arranged paratangentially, with their broad faces parallel
to the outer surface. Only around the pores some of them are differently sit-

276 ERYLUS SOLLASII.

uated; these turn one broad side towards the pore-canal and the edge towards
the outer surface of the sponge. Young and adult aspidasters also occur scat-
tered in the choanosome. In some specimens aspidasters are rather numerous
in the choanosome. In all parts of the choanosome acanthtylasters are met with.
These asters, particularly the small many-rayed ones, are more numerous in the
subcortical region than in the interior. In.the spicule-preparations of race I,
form C, large smooth-rayed oxyasters up to 56 in diameter, and in those of
race II] small smooth-rayed oxysphaerasters (Plate 3, fig. 26b) have been ob-
served. I consider these asters, which were not found in situ in the sections, as
foreign spicules. About the foreign nature of the small oxysphaerasters in the
spicule-preparations of race III there can indeed be no doubt, as a Donatia-like
sponge-crust, containing such oxysphaerasters in large numbers, covers parts
of its surface.

The rhabd megascleres (Plate 1, figs. 29-35, 42-48; Plate 3, figs. 19-22)
are for the most part blunt amphioxes (Plate 1, figs. 29-31, 33, 44-46; Plate
3, figs. 19, 20, 22). Besides these also sharp-pointed amphioxes (Plate 1, figs.
34, 35, 42, 48, 47, 48), amphistrongyles (Plate 1, fig. 32), and styles (Plate 3,
fig. 21) occur. Generally these rhabds are rather uniformly curved (Plate 1,
figs. 29-31, 33-35, 42, 48, 46-48; Plate 3, figs. 19-22), rarely straight (Plate 1,
figs. 32, 45), or abruptly bent near one end (Plate 1, fig. 44; Plate 3, fig. 18).
The styles and particularly the amphistrongyles are curved much less than the
amphioxes. The rhabds are 425-980 » long and 8-24 » thick. The longest are
found among the amphioxes, the thickest among the amphistrongyles and
styles. ‘The small specimens, race I, forms A and B, have smaller rhabds than
the larger ones. Among the latter race I, form D, and race III have larger
rhabds than race I, form C, and race II. Besides these simple rhabds, spicules
similar in shape and size, possessing however a short branch-ray, are met with,
chiefly in race III. In these mesomonaene-like rhabd-derivates the branch-ray
(clade) is pointed or blunt, up to 50 long, and situated near one end, in styles
thus branched near the pointed end. The branch-ray is either turned upwards
proclade-, or downward anaclade-fashion. (See table, p. 277.)

Most of the adult microrhabds (Plate 1, figs. 37-41a, 39b, 41b; Plate 2,
figs. 16-18; Plate 3, figs. 13-15, 26a) are quite stout, slightly and uniformly
curved, centrotyle, isoactine amphioxes. Most of them are blunt, some sharp
pointed. Sharp-pointed microrhabds are particularly frequent in race III.
The blunt amphiox microrhabds are often somewhat constricted just below their
ends, so that the ends themselves appear as terminal knobs (Plate 2, fig. 18).

ERYLUS SOLLASII. 77

DIMENSIONS OF RHABD MEGASCLERES.

Race

I II TT | all forms

general aver-
A B Gc | D limits | ~ age of the
| | largest three

of all the spie-
ules measured,

wn
D
i=)
bo |

|
425— | 530— | 470— | 530- | 450— | 720- | 425-
980 | 980

ft

Length

average of the 630

710 | 747 | 953 | 743 | 913 | 76
three longest, / y Siigeatt 853 | 7 | 766

of all the spie-
ules measured, | 8-13 | 8-14 | 10-24) 10-22] 9-20 | 12-22

Thickness B | | aa |

|
average of the 12 eel
: 2 3 22 2 9 2
three thickest, } ieee : ;

Besides these greatly preponderating isoactine forms, some anisoactine ones,
with one actine shortened and rounded at the end (Plate 1, figs. 39b, 41b), are
met with. In some of the microrhabds of raceI, form A, and race II this
reduction has gone so far that one actine is absent altogether. Such spicules
appear as styles (tylostyles). Also branched microrhabds, composed of more
than two actines, have occasionally been met with. These are most frequent
in race I, form B. Most of them are tetractine and appear to have been pro-
duced by an early concrescence of two simple microrhabds lying crosswise.
Two opposite rays of such spicules usually form a microrhabd of similar dimen-
sions to the ordinary ones. The microrhabd represented by the two other
rays is usually considerably shorter. Rarely the two microrhabds presuma-
bly composing these spicules are equal in length. Their axis generally encloses
small angles, 30° or less; rarely these angles are greater; forms with axis crossing
at right angles are exceedingly rare. Sometimes the one microrhabd is attached
to the other by its end; such spicules appear as triactines. The tyle is usually
asimple spherical thickening. In the isoactine forms it occupies the centre, in
the anisoactine forms it lies nearer to one end than to the other, and in the
microrhabds with one actine quite suppressed, it is situated terminally. Ina
few of the microrhabds of race I, form D, the tyle is irregular and appears as a

cluster of rounded protuberances.

278 ERYLUS SOLLASII.

No correlation between the size of the sponge and the dimensions of its
microrhabds is discernible. The thickness of these spicules is about the
same in all the forms; their length however varies, those of race III being con-
siderably shorter than those of the races I and IT, although the specimens of some
of these (race I, A and B) are very much smaller. The microrhabds are 30-78 fu
long and 2.5-5 «thick. The tyle is 0.3-1.5 » thicker than the adjacent parts
of the spicule and measures 3.5-6.5 in diameter.

DIMENSIONS OF MICRORHABDS.

Race
1 Il Ill all forms
| Form
| general average
A B Cc D limits of the largest
three
of all the spie-
ules measured, | 46-76 | 39-56 | 30-57 | 38-78 | 39-61 | 30-44 | 380-78
Length B
average of the = . x
65 54 56 62 98 44 56.5
three longest, 1
of all the spic- so
: ules measured, | 2.5-5| 2-4 | 3.5-5| 3.5-5) “y4 | 34.5) 2.5-5
Thickness ie |
average of the| , _
Z 4 8 S 8 4.1
three thickest, | —““ = : =e 43 -
of all the spic-
: aries : 3.5— | 3.5- | 3.5-
ules measured, | 4-6.5| 3-5 | 3.8-6] 4-6 x
- 0.5 5.5 6.5
Diameter of | 4
tyle —
average of the B 4 = ete
5.8 4.8 5.7 5.7 5.2 5.4 5.4
three largest,

The triaenes (Plate 1, figs. 5-26; Plate 3, figs. 1-6, 12, 23, 24) are ortho-
or, more frequently, plagio-triaenes with simple or branched clades. The
rhabdome is always straight. Usually it is simple and conical. Its acladomal
end is sharp pointed (Plate 1, figs. 17, 19, 20; Plate 3, figs. 12, 23, 24) or blunt
(Plate 1, figs. 15, 16, 24, 26). Sometimes slight knob-like protuberances are
observed near the acladomal end. Rarely the rhabdome is reduced in length,
cylindrical, and terminally rounded and slightly thickened (Plate 1, fig. 18;
Plate 3, fig. 1). The properly developed, conical rhabdomes are 210-520 p, the

ERYLUS SOLLASII. 279

reduced, cylindrical ones 140-220 y» long. Their thickness at the acladomal
end is 8-22. Cylindrical rhabdomes are always thick; all the slender triaene-
rhabdomes observed were conical. The triaenes of the small specimens (race I,
forms A and B) have somewhat shorter and very much thinner rhabdomes
than those of the larger ones. The averages of three thickest of the former
being 10-13, those of the latter 18-21 ». Among the large specimens the one of
race III has far larger triaene-rhabdomes than those of the races I and II.

The cladomes of these triaenes are very polymorphic. Triaenes with
simple clades occur in all the specimens. In race I, form A, no other triaenes
were observed. In race I, forms B, C, and in race II a few triaenes with one,
more rarely two or three branched (bifurcate) clades occur besides the ones
with simple clades. The ramification of the triaene-clades is still greater in
race I, form D, and in race III: in these the triaenes with branched clades are
more numerous than the ones with simple clades.

In the triaenes with three simple clades (Plate 1, figs. 7, 11, 12, 15-20,
24-26) the clades are usually conical and blunt pointed, rarely reduced in
length, cylindrical, and rounded at the end (Plate 1, figs. 16,18). Such a reduc-
tion of the clades is usually associated with a reduction or other abnormity
of the rhabdome. The simple triaene-clades are slightly and uniformly curved,
concave to the rhabdome (Plate 1, figs. 17, 19), or nearly straight (Plate 1, figs.
22, 25), or, more rarely, abruptly bent down at the end (Plate 1, figs. 18, 20,
24). Their chords are 120-300 yp long.

As stated above, one, two, or all three clades of the triaenes may be
branched. This branching is most frequently a simple and regular bifurcation,
the two branches (end clades) being simple, and fairly equal in length and angu-
lar position (Plate 1, figs. 6, 9, the lower ones in fig. 10; Plate 3, the lower left
ones in figs. 2, 3, 5,6). Irregularities due to a difference in the length or the
position of the two branches or to secondary ramifications of the branches are
frequently met with. The difference in the length of two end clades forming
a pair is caused by the reduction of one of them. This reduction sometimes
becomes so great as to lead to a complete suppression of one of the end clades,
in which case a single end elade arises from the, in such spicules usually some-
what thickened end of the main clade (Plate 1, fig. 8). The differences of posi-
tion are frequently so great that one end clade appears as a continuation of the
main clade (Plate 1, the upper one in fig. 5; Plate 3, the upper one in fig. 3).
Some of these spicule-rays might indeed be considered as simple clades from
which a branch-ray arises laterally. A secondary ramification of the end

280 ERYLUS SOLLASII.

clades has been observed only in race I, form D, and in race III. It affects
either both end clades of a pair in a similar manner (Plate 3, the left ones in fig. 4)
or one of them only (Plate 1, the upper one in fig. 10; Plate 3, the upper in fig. 2
and the right ones in figs. 2-5). The proportion of the length of the main clade
to the length of the end clades is, as a comparison of figs. 6 and 9 on Plate 1
shows, subject to very considerable variation. The main clades are 70-270, the
end clades 10-160 « long. The cladomes are 160-550 » broad. The triaenes
of the small specimens of race I, form A, have the narrowest cladomes (average
of the three largest 373 /), those of the large specimen of race III, the broadest
(average of the three largest 530 4). The angle enclosed by the axis of the
rhabdome and the chords of the simple clades and the stems (main clades) of
the branched ones, is 86-116° (general average 103.4°). It is smallest in the
triaenes of race I, form A (average 97°), and largest in those of races I and HI
(average 107 and 107.5°). Thus most of these spicules are plagioclades, some
orthoclades. (See table, p. 281.)

In race II some spicules, 310-330 in diameter, composed of two simple
(Plate 1, fig. 13) or branched (Plate 1, the right one in fig. 14) clade-like, and
two short, conical, blunt rudimentary rays, have been observed. These spicules
appear to be derivates of the triaenes described above.

The acanthtylasters (Plate 1, figs. 36-40c; Plate 2, figs. 14, 8-11; Plate 3,
figs. 7, 8, 25, 26c) are destitute of a central thickening and have from two to
fourteen, a few perhaps more than fourteen, rays. The rays are concentric and
nearly always uniformly distributed, simple, and equal in size. Acanthtylasters
with rays unequally long or branched (bifureate) have been only very rarely
observed. The rays of the larger acanthtylasters (Plate 2, figs. 1, 2, 8,9; Plate
3, figs. 7, 8) are cylindroconical and at the base 1-3 usually 1.5-2.3 yp thick.
They taper distinctly towards the end, and are, at their thinnest point a short
distance below the end, 0.7-1.5 « thick. The rays of the smaller ones (Plate 2,
figs. 10, 11) are somewhat more cylindrical and only 0.4-1.5 » thick. The
rays invariably bear spines, some of which always congregate at the end of the
ray and here form a terminal, acanthtyl cluster 1.3-3.5 yin diameter. The size
of the spines is on the whole proportional to the size of the aster. Apart from
this they are subject to considerable variation. Sometimes numerous small
and insignificant (Plate 3, fig. 7), sometimes numerous medium sized (Plate 2,
figs. 8, 9), and sometimes only one or a few very large spines, 1-1.6 » in length
(Plate 2, figs. 1, 2; Plate 3, fig. 8), arise from the sides of the rays. In the two-
rayed acanthtylasters a cluster of spines arises from the centre of the spicule.

ERYLUS SOLLASII. 281

DIMENSIONS OF TRIAENES.

Race
I II Ill all forms
Forms
general
pee =| aVCKAES \general
A B Cc D | limits pe average
| three
of all the spicules | 210- | 140- | 220— | 200- | 240— | 180— | 140- |
measured, 400 360 | 470 420 | 400 520 520
length |
av f the | :
ee ee | 378 | 200) sa7 | 380 | 380 || 490 385
three largest, 4 |
Rhabdome
of all the spicules | 19 15] s 10 | 10-22| 10-22] 12-20| 13-22| 8-22
measured, jt
thickness | |
average of the : Fs “ = e
three thickest, = ie fe = aM ae ue
P f all the spicules | 120- | 170— | 120- | 175— | 120—- | 225— | 120-
lades | le -
ee tet | casured, p 215 | 280 | 270 | 270 | 240 | 300 | 300
|
length of | of all the spicules 150— | 100- | 80— | 120- | 70- | 70-
i De 2 85 270 | 27
Branched main clades| measured, 230 | 200 140 185 70 | 270
lad 3 |
oo length of | of all the spicules 30-50 | 20-40 60- | 10-60 10= | 10-
end clades | measured, 1 ’ iz 160 130 | 160
of all the spicules | 160- | 370- | 215— | 290— | 170— | 280- | 160—
measured, /« 380 460 490 470 450 550 | 550
Cladome breadth |
average of the | 73 | 413 | 457 | 450 | 417 | 530 440
three largest,
of all the spicules} 89- | 100— | 89- 86— | 95- | 98— | 86—
l F ta 107 ll 112 116 115 116
Clade- (main clade) angle uecsured a ee Q i
| |
’ average of all 97 103 | 105 | 101 | 107.5 | 107 | 103.4

This appears as a rudiment of a third ray. Most of the spines are conical,
straight, and vertical, some conical and recurved (Plate 2, fig. 1), some irregu-
lar, cylindrical, terminally rounded or even thickened, and occasionally lobose
at the end.
; The acanthtylasters measure 10-38 in diameter. Their size is in inverse
proportion to the number of their rays. To obtain a clearer insight into this
correlation I measured (and counted the rays of) 207 of them and took the
means of the diameters of those with the same ray-numbers. There is no

982 ERYLUS SOLLASII.

difficulty in counting the rays of the large few-rayed acanthtylasters, but it is
impossible to ascertain the ray-numbers of the small many-rayed ones with
sufficient exactitude. I therefore calculated the means of the two- to nine-
rayed acanthtylasters by themselves, but combined the ten- to fourteen-rayed
ones in one group of which I took the mean. This mean can be taken as the
mean diameter of the asters with a ray-number equivalent to the mean of 10, 11,
12, 13, and 14, that is twelve. In this way I found that the average diameter

of the acanthtylasters with two rays is 38 4p,
of those with three DORs.
ahi | four “ Rs)
a «five oY 2 ns
Sie Sib : a)
car aS ““ seven * 235);
SO “eight es ZA
pone “nine o 19 “, and
“cc ce ce

ten to fourteen (mean twelve) rays 14 yp.

Apart from the two-rayed asters, which are so few that I was unable to
measure a number sufficient for attaining a reliable mean, the mean given
above shows that there is a very regular decrease in size with iereasing ray-
number, amounting in the asters with from three to five rays to 1 » and the
asters with six or more rays to about 2 per unit of difference of ray-number.

In all the forms four- to ten-rayed acanthtylasters have been observed.
The four- to six-rayed appearing to be the most frequent ones. In the forms
A, ©, and D of race I and in race II also three-rayed and in race II also a few
two-rayed acanthtylasters were found. Acanthtylasters with more than ten
rays have been found in all the forms except race I, form C. In the small speci-
mens the acanthtylasters are not smaller than in the large ones. The largest
acanthtylasters occur in the form A of race I and in the races II and III. (See
table, p. 283.)

The aspidasters. The dise-shaped spicules of the cortical armour of the
species of Erylus have hitherto been designated, like the ovoid spicules of the
armour of Geodia, as sterrasters. Closer examination of these spicules in the
species of Erylus of the ‘‘Albatross”’ collection has shown, however, that they
differ from the sterrasters found in the species of Geodia and allied genera not
only in their shape, but also in their mode of development, to such an extent
that it is advisable to give them another name. Aspidaster, the name selected
for them, has reference to their shield (do7is)- like shape.

ERYLUS SOLLASII. 283

DIMENSIONS OF ACANTHTYLASTERS.

Race
I Il Ill all forms
Forms
general average
A B C D limits of the laeeste
three
2 2s |
Number of rays 4-12 | 3-13 | 3-10 | 3-14 | 2-14 | 4-12 | 2-14
of all asters
measured wit :
B : 20-36 | 21—32 | 21-32 | 22-30 | 22-38 | 25-36 | 15-38
Six rays. or
fewer,
: of all asters
Boeancler measured with
, , 10-27 | 6-20 | 17-22 | 12-27 | 12-24 | 13-24] 10-27
seven or more
rays, /
average of the
a ap || 2a) 31 | 29 35 | 33 31.7
three largest,

The adult aspidasters of Erylus sollasii (Plate 1, fig. 36d; Plate 2, figs. 5-7
12-15, 19-26; Plate 3, figs. 9-11, 16, 17, 25d, 28) are 95-156 yp long, 55-82 It
broad, and 74-14 » thick. The general average proportion of their length to
their breadth to their thickness is 100:55.8:8.8. Optical transverse sections
show that these disc-shaped spicules are peadunlis attenuated towards the mar-
gin, which is usually quite sharp. The shape of their outline is variable; some
of them (Plate 2, figs. 13, 23, 26; Plate 3, fig. 9) are quite regularly oval, some
rounded rhomboidal (Plate 2, fig. 22; Plate 3, figs. 10, 17), and some irregular
(Plate 2, figs. 12, 24, 25) with lobose marginal protuberances of which one or
a few broad ones, or a larger number of narrow ones may be present. On one
face of the disc there is a very shallow, more or less circular depression 20-30
in diameter. In this depression, which is obviously homologous to the umbili-
cus of the sterrasters of Geodia, the surface is either quite smooth, or bears only
a few small rays or spines. From all the other parts of the surface (Plate 2,
figs. 5, 6; Plate 3, fig. 28) and also from the margin, rays usually about 1-2
thick protrude. Those on the margin are about 1.5 y long, those on the faces
appear to be shorter. These rays are scattered rather irregularly and (meas-
ured from centre to centre) 2-6 # apart. Hach ray bears a terminal verticil
of four to ten lateral spines. The centre of the aspidaster is occupied by a

small group of granules, from which very numerous and perfectly straight radial

284 ERYLUS SOLLASII.

lines extend towards the margin. In some adult aspidasters this radial structure .
is well defined (Plate 2, fig. 7), in others it can hardly be made out.

There appears to be a certain degree of correlation between the size of the
sponge and the size of its aspidasters, the latter being smaller in the small speci-
mens of race I, forms A and B (average length of the three largest of these forms
118 and 120 # respectively), larger in the middle-sized specimen, race I, form D
(that average 124 ,), and still larger in the large specimens of race I, form C,
and races II and III. Among the latter those of races II and III (that average
150 and 152 ») are very considerably larger than those of race I, form C (that
average 128 ,), and the other smaller forms of race I. Also in their shape the
aspidasters of the different forms differ to a certain extent, those of race II
being much more slender and those of race I, form C (Plate 2, fig. 25) much more
irregular than those of the others. Also in the number of the ray-spines differ-

» ences are observed, the rays of the aspidasters of race II bearing up to ten, those 7
of the aspidasters of the races J and III only from four to six lateral spines. ,

DIMENSIONS OF THE ASPIDASTERS.

Race

| I Il Ill all forms q

general ayer-
A B Ce D | limits age of the
| three largest

general
average

Senne | 102— | 95—| 105— | 107— | 120- | 128-| 95—
Wo assured, ft
: 121 |) 120) |) 130) |) 1261560 eosin eloG)

Length — —-
average of the
three longest, 1

120) |) S| 128) 124: 152 150 132 j
:

of all measured, pz
|
ea) ———

average of the
three broadest,

of all measured, ear her el Rie ee
: 12.2 | 10.4 12 11.4 14
Thickness} ————______}
average of the
three thickest,

Average of proportion of} 100: 100: 100: | 100: | 100: 100: 100:
length to breadth to thick- | 55.3:| 59.1: ie 56: | 49.8: | 57.7: 55.8:
ness, | 8.2 9 8.4 9.8 7.6 10 8.8

ERYLUS SOLLASILI. 985

Young aspidasters were found in considerable numbers scattered through-
out the choanosome in several specimens. They are imbedded in the ground
substance. A special membrane or plasmatic sheath enclosing them could not
be made out even in sections strongly stained with aniline-blue.

The youngest (smallest) aspidasters observed (Plate 2, fig. 14) were oval
dises, about 25 » long and 10 # broad, and composed of numerous exceedingly
slender and perfectly straight rays which radiated from a granular centrum
about 3 # in diameter. These rays are, at first, quite isolated. They grow in
length and in thickness and so the whole aspidaster increases in size (Plate 1,
fig. 36d; Plate 3, fig. 25d) and the basal parts of the rays become united. The
solid, central mass of silica thus produced forms a disc from the margin of which
the still isolated, distal parts of the rays protrude (Plate 2, fig. 19; Plate 3,
fig. 16). When this stage is reached the longitudinal growth of the rays slows
down or ceases altogether, while the transverse growth of the rays, that is their
increase in thickness, continues. In consequence the marginal spines become
joined more and more (Plate 2, figs. 20, 21) until they entirely lose their individ-
uality, the margin of the aspidaster becoming quite smooth. In young forms
of regular aspidasters this smooth, non-serrated margin is continuous (Plate 3,
fig. 11), in young forms of irregular ones lobose (Plate 2, fig. 15). Not only the
margin but also the two broad faces of such young aspidasters are smooth.
Their smoothness in this stage constitutes the chief difference between them
and the sterrasters of Geodia, which do not pass through a smooth stage during
their development. Later small, spine-like protuberances make their appear-
ance on the surface of the smooth young aspidaster. These develop into the
protruding rays with terminal verticils of lateral spines, which have been
described above.

LOCALITIES AND NATURE OF ENVIRONMENT.

Pat]
oo ,
3 Locality Date Depth Bottom ee
D
9
S. coast of Molokai, Lae-o Jel an Sand and Ronis
3847 te were iets N. 64°| April 8, 1902 (23-24 £.) = Cas AB
aa 22 (Race I)
S. coast of Molokai, Lae-o F 9] 7° : 2
A S0-133 m. 21. Sand and ;
384g |Ka Laau Light. N.68°15’,| April 8,1902 | Ss | otro | oravel Form (
W. 22.4’ a eta IN Ae aaa (Race I)

286 ERYLUS SOLLASII.

LOCALITIES AND NATURE OF ENVIRONMENT (continued).

35 ge
3 Locality Date Depth = a5 Bottom apeuneee
Li ae
i i
la eas; ‘
S. coast of Molokai, Lae-o 5 Coarse sand
| 133-78 m. 19.8° d 1
3849 | Ka Lak aight. N. 71°,| A 8, 1902 2 brok hells,
3849 | Ka ! san Light 7 pril (73-43 £.) (67.6° F.) oroken shells Race IT
W. 21.9 and corals
|N. E. coast of Hawaii, 91-121 m Fine gray sand 1
1055 | Alia Point Light. Hilo | July 16, 1902 (50. 62 £) ; —— and Foraminif- Race II
7 = oU-De I. oy
Bay. N. 20°, W. 3.5’ | era =
N. E. coast of Hawaii, a ee oe aren 1
1062 | Kauhola Light. S. 69°15’,| July 18, 1902 nee ¥ —_ pee! ... | Form D
Saas eke ~ (83-113 f.) and Foraminif-
| E. 6.9 en (Race I)

There can be no doubt that the seven sponges described above are very
closely allied, still they differ to a certain extent in size, shape, colour, and
spiculation. As to the size it is to be noted that the forms A and B of race
I are very much smaller than the others. Since, however, the spiculation of
these small specimens exhibits immature characters, there can be no doubt
that they are young forms, that their small size is merely due to their age
and of no systematic importance whatever. Neither can any importance be
attached to the differences in shape, since they lie well within the limits of
individual variation usual in sponges of this kind. The differences in colour,
which ranges from dirty white to dark chestnut-brown, are indeed great. If,
beginning with the lightest coloured one, we arrange the forms in the order of
the degree of their pigmentation, we get, 1) race I, form C, 2) race I, form D,
3) race II, 4) race III, 5) race I, form A, and 6) race I, form B. If, beginning
with the deepest, we arrange them in the order of the depth of the water at the
place where they were found, we get, 1) race I, form D, 152-207 m.; 2) race I,
form C, race II and III, 78-133 m.; and 3) race I, forms A and B, 42-44 m.
Although it is unknown which of the specimens of race I, form C, race II, and
race III grew in deeper, and which in shallower water, we see that the depth
of the locality is, on the whole, roughly in inverse proportion to the degree of
pigmentation. The darkest forms, the partly or wholly dark chestnut-brown,
race I, forms A and B, grew in the shallowest water, at a depth of 42-44 m., to
which, in clear tropical sea water, considerable day light penetrates. This .
and the fact that in some of the specimens the upper protruding parts are
darker than the basal, lead me to suppose that these differences of colour are

ERYLUS SOLLASII.

287

merely due to differences in the amount of light to which the different speci-

mens were exposed during life.

I am therefore inclined to consider these

differences as direct individual adaptations of the simplest kind to which no
systematic importance whatever can be attached.

The chief differences in the spiculation of the six forms is shown in the

following tabular view of the averages of the three largest observed of the most

important spicule dimensions and of the character of the triaene-cladomes and

aspidasters.
Race I Il III
Form A B Cc D
length, » 630 710 TAT 853 743 913
Rhabd megascleres
thickness, 12 13 23 20 19 21
length, 65 54 56 62 58 44
Microrhabds higlaness
ae 5.8 48 | 57 5.7 5.2 5.4
of tyle,
length, 373 290 397 380 380 490
rhabdome
thickness. 13 10 18 21 19 21
Triaenes cladome breadth, 373 413 | 457 450 417 530
most clades most clades
all clades | . most clades | . most clades
character : simple, some | }, simple, some
simple ranched : branched
branched branched
Acanthtylasters diameter, y 33 29 31 29 35 33
length, 120 118 | 128 124 152 150
Sey Tae mn 100: nin Lid 100: 100: 100:
length to breadth |... 0. | 99-1:| 97: 56:9.8 49.8: 7.6 57 7:10
to thickness Boo s8 2 9 8.4 :
Aspidasters many
ures |few irregu- f || few ixregu-
few irregular, four wee lar, four to} ©Y "TS | tay four to
character ‘ a LOU) || laruptoten| 2”
to six ray-spines «| SIX ray- epee a (SER TERY
to six : ray-spines cee 5
spines spines
ray-
spines

This table shows that in race I, form A, all the triaene-clades are simple;
and the rhabd megascleres shorter and thinner, the microrhabds longer and

thicker, and the triaene-cladomes smaller than in any of the others.

In race I,

288 ERYLUS SOLLASII.

form B, most of the triaene-clades are simple; and the microrhabds thinner and
shorter, and the aspidasters absolutely shorter and relatively broader than in
any of the others. In race I, form C, most of the triaene-clades are simple; and
the rhabd megascleres thicker and the aspidasters more irregular than in any
of the others. In race I, form D, most of the triaene-clades are branched; and
the triaene-rhabdomes relatively thicker than in any of the others. In race II
most of the triaene-clades are simple; and the acanthtylasters larger, the aspid-
asters absolutely longer and relatively considerably narrower and thinner and
their rays provided with a larger number of lateral spines than in any of the
others. In race III most of the triaene-clades are branched; and the rhabd
megascleres longer, the microrhabds very considerably shorter, the triaene-
rhabdomes much longer, the triaene-cladomes much broader, and the aspid-
asters relatively thicker than in any of the others.

» That the megascleres of race I, form A, are smaller in size and more simple
in character than those of the other forms and that there are other differences
of this kind, appears to be due to differences in the age (size) of the specimens.
Some peculiarities, as for instance the irregularity of the aspidasters of race I,
form C, may be pathological. Some are, no doubt, to be accounted for by
differences in the external forces which acted on the different specimens. All
these can be considered as mere somatic non-germinal characters, destitute
of systematic significance. There remain however some, the nature of which
is more doubtful and which might well be germinal. These peculiarities are
the exceptional narrowness and thinness of the aspidasters and the richness of
their rays in spines in race II, and the exceptional shortness of the micro-
rhabds and the exceptionally large size of the triaene-cladomes in race III.
If these peculiarities are considered germinal three systematic groups must be
distinguished, one for the forms A, B, C, and D of race I, one for race II, and
one for race III.

There can, I think, be no doubt that these three groups must be united in
one and the same species; it is another question, however, whether or not
varieties should be established for them. After a careful consideration I have
decided that these differences are probably germinal and systematically impor-
tant, but sufficient only for racial distinction, and I distinguish three races,
designated I, II, and III, in this species accordingly.

ERYLUS SOLLASII. 989

Race I.

Rhabd megascleres 425-880 by 8-23 4; centrotyle microrhabds 30-78 by
2-5 #« and a tyle 3-6 yw; triaenes with simple clades only, or with simple and
branched clades, either the former or the latter predominating; rhabdome 104—
470 by 8-22 yp, cladome 160-490 » broad; acanthtylasters with from three to
fourteen rays, 20-36 » in diameter; aspidasters, regular or irregular, with from
four to six ray-spines, 95-130 by 55-75 by 7.4-12.2 p, average proportion of
length to breadth to thickness 100 : 56.8 : 8.8. This race comprises four forms,
designated, A, B, C, and D.

South Molokai, northeast Hawaii.

Race II.

Rhabd megascleres 450-760 by 9-20 »; centrotyle microrhabds 39-61 by
3.4-4.3 mw, tyle 3.5-5.5 mw; triaenes with simple and with branched clades, the
former predominating ; rhabdome 240-400 by 12-20 y, cladome 170-450 p broad;
acanthtylasters with from two to fourteen rays, 22-38 in diameter; aspidasters
mostly regular, with up to ten ray-spines, 120-156 by 60-76 by 9.2-11.4 yp,
average proportion of length to breadth to thickness 100 : 49.8 : 7.6.

South Molokai.

Race III.

Rhabd megascleres 720-980 by 12-22 »; centrotyle microrhabds 30-44 by
3-4.5 mw, tyle 3.5-5.5 ; triaenes with simple and branched clades, the latter
predominating; rhabdome 180-520 by 13-22 y, cladome 280-550 y broad;
acanthtylasters with from four to twelve rays, 25-36 in diameter; aspidasters,
mostly regular, with from four to six ray-spines, 128-153 by 71-82 by 12-14 p,
average proportion of length to breadth to thickness 100 : 57.7: 10.

Northeast coast of Hawaii.

The character of the canal-system and the spiculation show that these
sponges belong to the genus Erylus. From all the species of this genus, with the
exception of the one from Freemantle, S. W. Australia, which Carter’ erroneously
identified as Erylus (Stelletta) euastrum O. Schmidt and for which Sollas* estab-
lished Erylus lendenfeldi, they differ very considerably. Sollas’s description
indicated, and a reexamination of the type, kindly placed at my disposal by

1 H. J. Carter. Report on specimens dredged up from the Gulf of Manaar. Ann. mag. nat. hist.,

1880, ser. 5, 6, p. 136.
2W.J.Sollas. Tetractinellida. Rept. voy. ‘Challenger,’ 1888, 25, p. 239.

990) ERYLUS ROTUNDUS.

Mr. Kirkpatrick, clearly shows, that this species also differs from Hrylus sollasit.
The chief differences between the two are the presence of asters 100 in diam-
eter and exceedingly irregular aspidasters in E. lendenfeldi, and their absence in
EP. sollasti. These differences are quite sufficient for specific distinction.

Erylus rotundus, sp. nov.
megarhabda, var. nov.

Plate 5, figs. 18-23, 32; Plate 6, figs. 14, 18, 24, 33-35; Plate 7, figs. 22-30, 57-73; Plate 8, fig. 13.

typica, var. nov.

Plate 5, figs. 1-4, 11-17, 30; Plate 6, figs. 15-17, 25,

bo
~]

, 30-32; Plate 7, figs. 16-21, 46-56.

cidaris, var. nov.
Plate 5, figs. 5, 26-28, 31; Plate 7, figs. 1-10, 42-45, 75, 76, 79; Plate 8, fig. 14.

» [establish this species for thirteen specimens obtained at five different sta-
tions among the Hawaiian Islands. The aspidasters of these sponges are nearly
circular in outline and to this the name refers.

The thirteen specimens represent seven distinct forms which fali into
three groups. The latter I consider as varieties. One of these varieties pos-
sesses remarkable rhabd-clusters resembling certain Cidaridae in appearance,
hence the varietal name cidaris. The other two varieties, which are destitute
of these spicules, differ in respect to their microrhabds, these being very much
larger in one of them than in the other. The former I name megarhabda; for
the latter, which is the most frequent of the three, I have selected the name
typica.

Two forms belong to the var. megarhabda, four to var. typica, and one to
var. cidaris. "The number of specimens and the habitat of each form are tabu-
lated below: —

Erylus rotundus
var. megarhabda
form A: 2 specimens from the south coast of Molokai (Sta-
tion 3849);
“B: 1 specimen from the coast of Kauai (Station 3982);
var. typica
form A: 4 specimens from the south coast of Molokai (Station
3849),
“B: 3 dark-coloured specimens from the coast of Kauai
(Station 4024);

——————— OCLC eee

ERYLUS ROTUNDUS. 291

form C: 1 light-coloured specimen from the coast of Kauai
(Station 4128);

D: 1 specimen from the northeast coast of Hawaii
(Station 4061);

“ee

var. cidaris
1 specimen from the south coast of Molokai (Station 3849).

Shape and size. The larger of the two specimens of var. megarhabda, form A
(Plate 6, fig. 32), is upright, somewhat flattened, ellipsoid, 45 mm. high. Its
largest and smallest horizontal diameters are 38 and 27 mm. respectively. The
surface is rugose. The protruding ridges are high and irregular on the apex,
lower and arranged in a more regular manner longitudinally, at the sides of the
sponge. On the apex and the upper parts of the sides numerous circular or
oval apertures, 0.1—-0.7 mm. in diameter, are observed. The smaller specimen
of this form is irregular, massive, 32 mm. long, and in part covered with foreign
bodies, attached to the partly undulating, partly rugose surface. There are a
few groups of apertures up to 0.8 mm. in diameter. The single specimen of
var. megarhabda, form B, is irregular, massive, and 23 mm. long. To its rugose
surface foreign bodies are attached. There is one group of six conspicuous aper-
tures 0.7-1.4 mm. wide on the surface.

The largest of the four specimens of var. typica, form A (Plate 5, fig. 30),
appears as an upright bunch of thick lobose parts, joined for the greater part
of their length to form a continuous mass, from the upper side of which their
free lobes protrude. The whole sponge is 67 mm. high; its largest and smallest
horizontal diameters measure 69 and 72 mm. respectively. The lobose parts
are 16-36 mm. thick and distally rounded. The surface is slightly rugose. On
and near the summits of the lobes a few larger apertures, 1-2 mm. wide, are
observed; the sides are occupied by numerous small pores. Considerable parts
of the surface are covered by an incrusting composite ascidian (Plate 5, fig. 30).
The other three specimens of this form are similar, but smaller, the smallest
only 33 mm. high. To the surface of one of them numerous foreign bodies,
fragments of shells, ete., are attached. The largest of the three specimens of
var. typica, form B, is an irregular lobose mass, measuring 55 by 51 by 46 mm.
It appears to be composed of more or less coalesced gyriform parts up to 10 mm.
in thickness. Some of these terminate in slightly protruding digitate excres-
cences. The surface is penetrated by numerous small apertures and partly
covered with symbiotic sponge-crusts and foreign bodies (fragments of shells,
etc.). The other two specimens are similar and only slightly smaller. The

992 ERYLUS ROTUNDUS.

smallest is 52 mm. long. One of them possesses, besides numerous small pores,
two larger apertures (oscules) 1.5 and 2 mm. in diameter. The single specimen
of var. typica, form C, is irregular, massive, and 34 mm. long. Several short,
lobose protuberances arise from it. The surface is perforated by numerous
small pores and foreign bodies are attached to parts of it. The single specimen
of var. typica, form D, is an elongate mass, attenuated at one end to a digitate
process, 5 mm. thick. The total length of the sponge is 44 mm. Small pores
are scattered over its surface.

The single specimen of var. cidaris (Plate 5, fig. 31) is an upright, lobose
mass, 67 mm. high. Its largest and smallest horizontal diameters measure
57 and 45 mm. respectively. Small irregular grooves are observed on its other-
wise smooth, undulating surface. Here and there two adjacent grooves extend
for some distance in parallel directions, enclosing a gyriform fold. The surface
is perforated by numerous small pores, and a few crusts of symbiotic organisms,
but no dead foreign bodies, are attached to it.

The colour of the interior of these spirit specimens varies from dirty white
to light brown, that of the surface is subject to considerable variations. The
upper part and the sides of the large specimen of var. megarhabda, form A, are
dark purplish brown, the base and the interior being light dirty brown. Where
the dark colour of the sides gradually merges into the light colour of the base,
numerous whitish spots, marking the position of the — mostly closed — pores,
are observed on the surface. The smaller specimen of this form is coloured in
the same way, but the light-coloured part of the surface is here relatively more
extensive. The single specimen of var. megarhabda, form B, is rather dark
purplish gray.

Three of the specimens of var. typica, form A, are purplish brown above and
light dirty brown below; one is bluish gray. The three specimens of var. typica,
form B, are dark purplish black above and much lighter purplish brown below.
The single specimen of var. typica, form C, is light purple with a small, consider-
ably darker patch. The single specimen of var. typica, form D, is dirty white.
The single specimen of var. cidaris is whitish with a large brown patch, in
which numerous whitish spots, marking the position of the mostly closed pores,
are observed.

The differences in the degree of pigmentation of these sponges are probably
due to differences in the amount of light that fell on their surface during growth.
I think that in the specimens not uniform in colour, the upper parts, which were
more exposed to the light, became more strongly pigmented than the lower parts,

ERYLUS ROTUNDUS. 293

which were more or less in the shade, and am inclined to ascribe the differences
in the degree of pigmentation of the darkest parts of different specimens to
differences in the amount of light due to differences in the depth at which they
grew. Unfortunately the information about the depths given is not sufficiently
exact to allow of a definite conclusion on this point.

The superficial part of the body is differentiated to form a cortex, composed
of an outer and an inner layer. The outer layer is occupied by dense masses of
spicules and appears as an armour. Under the outer exposed parts of the sur-
face this armour usually is 65-90 y, in the walls of sheltered cavities, extending
farther into the interior, only 35 » thick, or even thinner. Pigment cells
occur in the armour between the spicules on the dark parts of the surface. The
inner layer of the cortex is usually 55-75 yw thick and contains hardly any
spicules. It is composed of paratangential fibres, pigment cells, and usually
contains also granule cells.

The pigment cells, the number of which is in proportion to the degree of
darkness of the surface, are nearly always elongate and usually extend para-
tangentially. They have one or, more frequently, several lobose or filiform
processes, appear irregularly amoeboid, and are very variable in size, 6-29
long. The transparent plasm of these cells contains numerous apparently
spherical granules, dark brown in transmitted light, which measure 0.3-0.8
in diameter. These granules are usually rather uniformly distributed through-
out the body of the cell and its processes, but sometimes parts of the cell are
free from them. Occasionally rows of single pigment granules, appearing like
strings of beads, have been observed in the sections. These probably lie in
(invisible) filiform processes of pigment cells.

In the distal part of the choanosome and in the lower layer of the cortex
of forms A and B of var. typica, and also in some of the others, remarkable
granule cells have been observed in large numbers. These cells appear to be
situated in spherical, oval, or irregular cavities of the ground substance, 15-20
in diameter, which in some places lie very close together. The granule cells
themselves are more or less spherical, measure 8-12 in diameter, appear hya-
line, and stain slightly with haematoxylin and aniline-blue. They are either
simple and structureless, or composed of a number, from ten to twenty or so, of
polyedrice parts 2-4 in diameter. The spaces between these parts appear to
be empty. Rarely a more strongly stained, superficial layer and a body, which
may be a nucleus, have been observed in the simple, undivided cells; and occa-
sionally minute pigment granules are attached to, or contained in, the ones

294 ERYLUS ROTUNDUS.

composed of polyedrie parts. It is possible that the spaces between these cells
and the walls of the cavities, within which they lie, and which appear to be empty,
are in reality thick, hyaline, cell walls. But as these spaces are not stainable
with any of the stains (eosin, malachite-green, magenta, aniline-blue, methyl-
violet, azure, haematoxylin, aurantia, Bismarck-brown), I think this improb-
able. The cells composed of parts are much more numerous than the simple,
undivided ones. The latter are scattered quite irregularly between the former
and do not inerease in number either towards the surface or towards the interior.
Although convinced that the undivided ones and the ones composed of polyedric
parts are merely different stages in the development of the same kind of cell,
[ am unable to say whether the simple ones arise from the composed ones or
vice versa.

In the sections of var. typica, form B, groups of broad, irregularly oval cells,
28-32 y long, were observed in the distal part of the choanosome. The plasm
of these cells is granular and each one contains a large nucleus, about 8 y in
diameter. These cells appear to be ova.

In the sections of var. cidaris young larvae were observed. Some of these
lay free on the canals, others appeared to be just on the point of emerging from
the cavities of the ground substance in which they were bred. These larvae
are spherical, measure 50-60 in diameter, and appear to consist of a central
granular mass, surrounded by a single layer of roundish, not elongate cells
about 8 in diameter.

In the choanosome of var. typica, form D, large numbers of monocellular
symbiotic Algae were observed. These are spherical or oval, measure 15-20 in
maximum diameter, and have a stout cell wall about 4 thick.

Canal-system. The uniporal entrances to the canal-system are usually
circular, and, when quite open, 100-250 # wide. Dilated pores of this width
are however not frequent, most of the pores being more or less contracted and
smaller, or closed altogether. The flagellate chambers are more or less spherical
and measure 14—23 in diameter. Those of var. cidaris are smaller (diameter
14-17 ys) than those of the others. The collar cells clothing them are not numer-
ous, distant, rather slender, and 4-6 #long. The larger canals are surrounded by
stout mantles of tissue free from megascleres and flagellate chambers. Some of
them are traversed by sphincter-membranes. Such have been particularly
observed in var. typica, form B. In the forms C and D of var. typica and in var.
cidaris no apertures much larger than the pores described above, were observed

on the surface. In these sponges the efferent openings (oscules) do not seem

ERYLUS ROTUNDUS. 995

to be of much greater width than the afferents. In both forms of var. mega-
rhabda and in some specimens of the forms A and B of var. typica on the other
hand, larger oscules, up to 1.4 mm. wide in the former, and up to 2 mm. wide
in the latter variety, have been observed. These larger oscules usually lie on
or near the summit of protruding parts of the sponge. In some eases, as for
instance in megarhabda, form B, large oscular tubes, up to 2.4 mm. in diameter,
lead up to the oscules. In other cases, as for instance in var. cidaris, a tract of
transparent tissue, about 1.7 mm. broad, free from megascleres and flagellate
chambers, extends from each of the here strongly contracted or evenly closed
oscules, down into the interior of the sponge. In the axis of this tract a row of
small cavities is observed. These cavities, which in the radial sections appear
to be isolated, are 100-150 # broad, up to 350 long, and situated close together.
Distally, towards the contracted oscule, they become smaller and scarcer. I
consider these rows of cavities as the remnants of the lumen of the strongly
contracted oscular tubes.

The skeleton consists of rhabd megascleres, microrhabds, triaenes, large
acanthtylasters with not very numerous rays, small oxyasters with numerous
rays, and aspidasters. In several forms also asters, resembling the acanthtyl-
asters In size and ray-number, but with conical rays, which become very slender
distally, have been observed. These spicules, which are particularly numerous
in var. megarhabda, form A, are in all probability merely young stages of the
ordinary acanthtylasters. I shall not therefore deal with them as a special
spicule form. In var. cidaris aster-like rhabd-clusters have been observed.

Some of the rhabd megascleres are isolated, others form more or less undu-
lating bundles (Plate 6, fig. 25a), which traverse the internal parts of the choano-
some in a radial or, in the digitate and lobose processes, longitudinal direction,
and, on nearing the cortex, tend to assume a position vertical to the surface.
These bundles are in var. megarhabda, form A, up to 100 4 broad. In the other
forms most of the bundles are 10-40 # thick. The broad bundles of var. mega-
rhabda, form A, appear to be flattened, band shaped; the narrow ones are cylin-
drical. The isolated rhabds are, in the interior, quite irregularly scattered;
near the surface, just below the cortex, most of them usually assume a position
more or less vertical to the surface. This radial arrangement of the subcortical
rhabds is particularly well marked in var. megarhabda, form A.

The asterose rhabd-clusters of var. cidaris are scattered in the choanosome.

In var. megarhabda the armour is composed chiefly of obliquely or radially
situated microrhabds, aspidasters being relatively scarce and confined to its

296 ERYLUS ROTUNDUS.

superficial part. In this variety numerous microrhabds also occur scattered
in the choanosome. In vars. typica and cidaris, on the other hand, the armour
is chiefly composed of aspidasters, and here the microrhabds are confined to its
superficial part, except in the vicinity of the pores, around which they form
mantles, extending right through the whole armour. Sometimes the superficial
microrhabd-bearing part of the armour-layer is stout and well developed, and
then it consists of an outer zone composed of paratangial microrhabds and an
inner zone of oblique and vertical (radial) microrhabds, lying above and be-
tween the outermost aspidasters. Often, however, this microrhabd-bearing
outer armour-layer is insignificant, and then composed only of relatively few,
mostly oblique microrhabds. It is possible that the superficial parts of the
sponges presenting this appearance have been rubbed off. The majority of the
microrhabds in the pore-canal mantles are situated so that one of their ends
points obliquely upward towards the centre of the pore. In some forms of
rar. typica, particularly in form B, a fair number of microrhabds were also found
scattered in the choanosome.

The cladomes of the triaenes extend paratangentially just below the cortex
or within its lower, fibrous layer; their rhabdomes are directed radially inward.
The triaenes occupy the interporal spaces and in some forms, as for instance
in var. typica, form A (Plate 6, fig. 27a), form well-defined groups, in which a
number of triaenes lie close together at the points of intersection of the inter-
poral zones.

The acanthtylasters and their oxyaster-like young are scattered through-
out the choanosome. They are most abundant in one of the specimens of
var. megarhabda, form A. In some forms, as for instance in var. typica, form B,
they are very much searcer in the subcortical region than in the interior of the
choanosome. A great many acanthtylasters lie in the walls of the choanosomal
canals. Some of the rays of these usually protrude into the canal-lumen. In
the walls of the remnants of the contracted oscular tubes of var. cidaris, above
referred to, the acanthtylasters stand particularly close together and here form a
veritable pavement. This local acanthtylaster-density is doubtless due to
the contraction of the surface on which, when normally extended, they are
probably distributed in the ordinary, not particularly dense manner.

The small oxyasters with numerous rays are confined to the roofs of the
subcortical cavities and the walls of the pore-canals. In the choanosome they
appear to be entirely absent. In the walls of the pore-canals, where they are
most numerous and sometimes form quite a dense layer, they extend right up

ERYLUS ROTUNDUS. 297

“

to within a short distance of the outer surface. In one of the specimens of var.
megarhabda, form A, I failed to find any of these asters in situ in the sections.

The aspidasters take part in the formation of the cortical armour and are
also found scattered in the choanosome. In vars. typica and cidaris the greater
part of the armour is composed of these spicules, which are here absent only in
the mantles surrounding the pore-canals. Apart from these mantles, the
proximal (internal) part of the armour in these varieties consists entirely of
aspidasters. In the distal (external) part of the armour microrhabds are usually
added to the aspidasters, and sometimes the outermost part of the armour con-
sists entirely of microrhabds. Most of the aspidasters of the armour are situated
paratangentially. Under exposed tracts of the surface they form many layers,
under the sheltered parts of it which limit the cavities, extending into the inte-
rior, only few layers or only a single layer. In var. megarhabda the aspidasters
form only a small part of the armour and are here confined to its distal (external)
part. Young and also adult aspidasters are usually found scattered in the
choanosome. Here they generally lie in cavities of the ground-substance, as
long and broad but much wider (thicker) than the aspidasters, so that the mar-
gins of the aspidasters are in contact with the ground-substance, while their
faces are separated from it by apparently empty spaces. Seen en face the aspid-
asters consequently appear to fill these cavities completely, seen in profile they
appear as narrow bars occupying the long axis of the oval cavities. The empty
spaces at their sides may of course have been produced by the shrinkage of the
tissue, during the preservation of the sponges in alcohol, but they may also be
natural, and in this case occupied possibly by some liquid, rich in silica secreted
from the surrounding tissue, from which the cell or cells building the aspidaster
draw their supply.

The rhabd megascleres (Plate 5, figs. 11-23, 26-28a) are for the most part
simple amphioxes or amphistrongyles, curved uniformly, or in the middle
more strongly than near the ends. Occasionally style (Plate 5, fig. 17¢) and
angularly bent or branched derivates of these rhabds have been observed. The
ordinary amphioxes and amphistrongyles are 310-650 long, and 6-15 yu thick.
Of the amphioxes and amphistrongyles occurring together in the same speci-
men, the former are on the whole longer than the latter. In the vars. mega-
rhabda and cidaris nearly all the rhabds are sharp-pointed amphioxes (Plate 5,
figs. 19-23, 26-28a), blunt amphioxes and amphistrongyles (Plate 5, fig. 18)
being rare. In the forms B and C of var. typica, sharp-pointed amphioxes
(Plate 5, figs. 14, 15, 17a) also preponderate; in the forms A and D of this
variety, on the other hand, the blunt rhabds (amphistrongyles) (Plate 5, figs.

298 ERYLUS ROTUNDUS.

11-13, 16b) are more numerous than the sharp-pointed amphioxes. Some
of the amphioxes, this was particularly observed in form B of var. typica, appear
as amphistrongyles, the blunt ends of which are surmounted by small conical
tips. These tips may be simple or terraced, telescope-like. Angularly bent or
branched rhabd-derivates have chiefly been observed in form A of var. mega-
rhabda and in var. cidaris. The spicules of this kind in the last-named variety
appear as transitional forms connecting its aster-like rhabd-clusters with the
ordinary rhabds.

The dimensions of the rhabd megascleres and the relative frequency of the
sharp-pointed and blunt amphioxes and amphistrongyles in the different forms
and varieties are tabulated below.

Erylus rotundus

yar. megarhabda var. typica var.

= cidaris
A B | A B c itl ies s
| | g a
| | a oe
la
| =e 2s
Rhabd megascleres \| | | as ag
| | | B = ar a @
|] I r=] ae oe
| | AS = = as ao
S| | eo eee a=
(ee Pe ck r= =a z ao ss
| a 3 aw 7) 3 E cI
3 anes “4 = | & I a
i) 1 ae 2] <ioel we oe
° 3 S oP ri ae) a5
= | = SI rs < em em
= 3 a 2 a » os Ze x 2
1] S = S a = | 5 =F rie
| 3 | | Ea
PR 4 | R s < Zz ae 23
| limits F || 330— | 350— | 330- || 320 330- | 350- | 310— | 310— || 449- |} 310—
| , 659 550 650 |) 570 5609 520 490 570 650 650
Length — || | eee
| 1]
|average of the a Par eeoelll Mees ai =>
| 603 5A3 573 || 556 533 507 477 | 518.3 617 5AS
longest three,
|
= = -| |
° . 1| _ e « * « 4
limits, yr || 6-13 | 6-12 | 6-18 || 6-11 | 6-12 | 6-11 | 7-15 | 6-15 || 8-12 || 6-15
|
——— Pe S| |
Thickness f th ||
average 0 1e |
Sones 13 ll 2 | 11 11 14 11.3 12 11.9
thickest three, yt
— = Pe | ' no = 1 too = ib
2/2 |2 |g |88 18 |8 (ss) 8 Wee
i = = ; @ of = s Sas at as
oa) 2 | = = =5 ra) , Ses a 2
| == a — tks :- e Bee = ce
ls | | | 4 (ee |e |S |) eee eee
| 4 | ee om w a Ss a Pd 3
s s g rs & & 4, 22 || s ie
= ze a | a : Sa = --2
| = = & | ?s |e | 6 seal cae ge
E E —E 3 Be gs seg
3 3 3 = ae 5 ial 2EE a 58
“he ce) es) (2 B =| 25
Shape A 3 5 teh || ae 3 we | e728 || s Be
>=) + = | 55 = 2 = as = poy
= S i} a) Sh = 2S ~—“e, |] & aS
3 o i) f) gare tl See leks sz | ee | ae ae
S s
iy ie | eae 25) & mizy ese || oe
B, a a | Gs | BS2] a Be gee 2 o&
5. c. Ste ||| BE 225 re es pn Fane 2a
| 22 | 28 | 3é |) 22 | See) S808] | e224 seq
= Os ms ns cle SES as a sae maa oso
Se ray) rag) ae red = n wn B aeza
es = > =o TER > Soh |Pos be TSE
os a Be beay} SA eas wa SA S55 Sa ano
|| 2g mE BE uo RAD BE Zo ase ae Boe
oe ce on on Sze on os Eos on B25
|) a2 Be Be) Ba | aGeak | 8S Fa | sao ES nao
!

ERYLUS ROTUNDUS. 999

The aster-like rhabd-clusters (Plate 7, figs. 4-10), which have been found
only in var. cidaris, are, in my opinion, to be considered as derivates of ordinary
rhabds. They appear as smooth oxyasters, composed of from about fifteen es
thirty concentric, straight rays fairly uniform in thickness, but differing exeeed-
ingly in length, and distributed very irregularly. From four to ten of the rays
appear properly developed; these are conical, more rapidly attenuated distally
than proximally, and pointed. The other rays are rudimentary, very short,
eylindrical, and terminally rounded. These rudimentary rays together form a
kind of lobose centrum, from which the longer, pointed rays arise. The rhabd-
clusters are 125-180 in total diameter. Their rays are 5-8 » thick and the
longest one of the whole cluster is 70-100 y long.

The branched amphioxes (Plate 7, figs. 1-3) also occurring in this variety,
which I consider as transitions between the clusters and the ordinary rhabds,
are 410-520 y» long and 9-13 » thick. They bear from one to four straight
branch-rays.

The microrhabds (Plate 5, fig. 27e; Plate 6, figs. 30-35; Plate 7, figs. 46-51,
50a, 54a, 55, 56a, 57a, 60a, 61-73, 75a, 79) are for the most part simple, isoactine,
gradually or rather abruptly pointed or, more rarely blunt, uniformly curved,
and usually slightly centrotyle amphioxes. The tyle, never large, is often so
insignificant as to be hardly visible, and many of these spicules seem to have no
central thickening at all. In the forms A, B, and C of var. typica (Plate 6, figs.
30-32; Plate 7, figs. 46, 48-51, 53a, 54a, 55, 56a) nearly all the microrhabds have
a distinct central tyle. In the microrhabds of var. cidaris (Plate 7, figs. 75a, 79)
the tyle is not so well developed, and in most of the microrhabds of var. mega-
rhabda (Plate 6, figs. 33-35, Plate 7, figs. 57a, 66-73) and var. typica, form D
(Plate 7, fig. 47), the tyle is hardly perceptible or absent altogether. Besides
these regular, simple, and isoactine microrhabds a few anisoactine ones, with one
actine reduced in length and rounded at the end, and a few with small branch-
rays near one end, have been observed in var. cidaris, in both forms of var.
megarhabda, and in form C of var. typica. The microrhabds are 30-98 long and
1.5-7.5 yu thick. Those of var. megarhabda are considerably larger (maximum
averages of three 82.3-95.7 by 5.3-6.7 #2) than those of the other two varieties
(maximum averages of three 54-61.7 by 3-3.8 and 66.1 by 4.2 » respectively).

300 ERYLUS ROTUNDUS.

DIMENSIONS AND SHAPES OF MICRORHABDS.

| Erylus rotundus

var, megarhabda var. typica var. =|
cidaris oe
| | os
B | s AaB c Digs S
|} A | 3° ° g
3 2 be
|| | bes | we a
s+ | PS eo
5 52 oa
2 | 2a Sn
Microrhabds ae zi =e || as we
| oe 2 Spoil =o os
haces SI E = os ae
a 3) ee 3 2S Eh
| Baa D 2 3 cu so
= | o2 = a a = So oc
g | £2 g = = = Ee EL
3 |e co rel ate to ed as =s
fond So S) 3 bo = TR n
Sit all less = ees 3 gs Se
=~ || ee = aaa ro)
= Scans S Sl ree Nimes Zs
i Ate Ss Ex iG) ot hoe 5 E= BS
|| wm 4 as D | Z pie] a5
== = | zs |
et
limits, 2 | 45-98 | 43-86 | 43-98 |) 35-60 30-55 | 31-58 | 46-66 30-66 || 32-55 || 30-98
!
|
Length f the |
average o mF 5 ae = a
. acl || 95.7 | 82.3 89 57.7 54 57 61.7 | 57.6 54 66.1
| longest three, y |)
| }
liens oes oemeea | aeeee lone 5 we >
| limits, 3-7.5 | 3-5.5 | 3-7.5 |/1.5-3.3) 1.5-3 | 2.5-4| 2-3.5 | 1.54 |] 2-65 |/1.5-7.5
Thickness Ss
av e
| eee Gz 5.3 6 3.1 3 3.8 3.3 3.3 4 4.2
thickest three,
= Airs | LP a . > Fastest eo —
—£ |23 |s< || 3 & |g | 83 5 2
r= aS on = s s EE 8 S
o a = roy n on =
) sees | --3 ° s ° 3 = S
|o@ | oo a Ss 2 s
Zs ore coats 5 3 ee > zy
| -2 = —= 2 fe 2 Es = 6
| & e5 | ge g = ae = sa =
3 Be aro fat |e Bog lis wba 3 4
By bo | bE 2 2 2 a = =
> = ae 5 S = begs & =
. ms a Ee Zz = ba) =. Re) a
Shape | Bs Pe |) zis = ® 2 aC} = BE >
I] oe feces” /c'O = = Sp |r =. 3
an mn | a ae > =a- os ne =
Ba Bo || = & cl = Gimp ||| eh = 5
|| o® "Ss || Os = a =, i ed ko BS So
ey, Res | es S = 3 == oa ee a
oo | See (2? |) Ba | 3) |) el) See) ook eee =
B= nea P 5s | 8 g 2 = =
2a |zes|ee | ¢e | 2 | » | Bo | See || Be |) Be
s& | dese | srs rote | zy = aa Sef s =
sa | 5*§ | Se || se a a | So | 380 |) Se z5
oleae || Geta | esp cae 3) =) dad | So il sa 25
Fis S20 | saZ sé S 2s aso eT EB
me | bac | bse £8 a = me SEE mS Eas)
|
|

Most of the triaenes (Plate 5, figs. 1-5; Plate 6, fig. 27a) are orthotriaenes,
some plagiotriaenes. The rhabdome is conical and generally slightly and irreg-
ularly curved in an undulating manner. It is 170-370 » long and, at the
cladomal end, 6-12 » thick. The clades of the same cladome are fairly equal in
length. They are usually slightly curved and 80-270 » long. The breadth of
the cladome is 155-440 . The triaenes of var. cidaris have the broadest, those
of var. megarhabda the narrowest cladomes. The angle enclosed between the
clades and the rhabdome is 90-107°. In form A of var. megarhabda I have

found a few triaenes with clades either bearing a small branch-ray or abruptly
bent down near the end.

ERYLUS ROTUNDUS. 301

DIMENSIONS OF TRIAENES.

| Erylus rotundus

| var. megarhabda var, typica var.

| | cidaris | 3
| A B S AS | 8 c D = | :
| Bre | cee el
a D> 5
oe | | | 52 || 3
Triaenes I Be || ie - | $5 | 2g
| i | § § | Sé | 33
oz Eg q | oy 7
| 2 | 3° a | @ a 5 | 22 ==
| gs Be! |e 3 = & | es Se
4 3 | g3 2 Baal ae is 2 28
s | 8 | Bs 5 5 5 5 = Ee
R +} z a a - Z 3 leas
|} | mas | | | || ||
length (lim- || 170— | 350- | 170- || 200- | i | 200- || 180 || 170-
its), || 180 | 370 | 370 || 220 heres) 220 || 190 || 370
Rhabdome I i
thickness |, | |
eee 11 | 2 = 12 || 7 2
diirnttah eg 6-11 6-11 || 8-1 8 | 89] 8 | 812] 78 || 641
|
|
a 85— | 150-| 85- || 140-| 80- | 150- | 210-| 80- |] s0-|| s0-
Clade-length (limits | | 210- | | |
ade-length (limits), » 180 | 200 | 200 || 225 | 210 | 220 | 250 | 250 || 270 || 270

| 160- | 250- | 160- || 300- | 150- | 250- | 3s0- | 150- || s00- || 150-

limits, y 290 | 350 | 350

iv)
(Yo)
So
be
i=)
i=)
He
ix
oO
ns
is
o

os
or)
S
_
S
Ss
al
©

Cladome
breadth | average of |
the broadest | 231 333 282 || 347 | 357 | 310 | 370 346 |) 397 335

three, y
| | | ||
ae 92- | | 92- |} 92 | | ene | ye) Cre
limits ° \| 90 90 |
yy 2 7 1 1077 1] O7
Clade-angle 10: 10 | 107 I | | 107 | 104 I 107
average ° || 96.5 96.5 || 99.5 90 | 90 | 93.2 || 100.5 || 95.3

|

The acanthtylasters (Plate 6, figs. 14b, 15, 16; Plate 7, figs. 52-54b, 56b,
57b, 58, 59, 75b, 76b) have from two to fourteen concentric, regularly dis-
tributed rays. Two-rayed acanthtylasters are rare and have been observed
only in var. cidaris. Also the three-rayed, which have been found in var.
typica, form B, var. megarhabda, form A, and in var. cidaris, are not frequent.
Four- to eight-rayed acanthtylasters are abundant in all the forms. Acanthtyl-
asters with more than eight rays appear to be most frequent in var. typica,
form B, and in var. megarhabda, form A. The acanthtylasters measure 12-31 yu
in total diameter. Those of var. cidaris are somewhat smaller than those of the
other two varieties. The size of the acanthtylasters is on the whole in inverse
proportion to the number of their rays, those with from two to six rays being
16-31, those with from seven to nine rays 14-24, and those with from ten to

302 ERYLUS ROTUNDUS.

fourteen rays 12-19 in diameter. The rays are, at the base, 0.5-2.4 » thick
and taper distally. At their ends they are usually thickened to an acanthtyle,
rarely simply rounded off. The transverse diameter of the acanthtyle (rounded
end) is, inclusive of its spines, 0.5-4 ». The basal part of the rays is always
quite smooth. This smooth part may be quite short, or it may extend right up
to the acanthtyle (the rounded, spiny end). Acanthtylasters with rays smooth
right up to the acanthtyle have been observed chiefly in var. typica, form A.
Usually the proximal one to two thirds of the rays are spineless, the remaining
distal part being either rough or provided with smaller or larger spines. The
larger the spines, the fewer their number. Frequently an increase in the size of
the spines towards the end of the ray is noticeable. The acanthtyle is covered
with numerous, fairly large spines. Most of the spines appear to be conical and
pointed, but in var. megarhabda, form A, acanthtylasters with cylindrical spines,
rounded at the end, have also been observed. The spines of the acanthtyle are
usually somewhat recurved. The proximal spines are usually directed more or
less obliquely backward, the distal obliquely upward and outward. On the
whole the acanthtyle-spines have the appearance of short hair combed down
in all directions from the apex. The spines on the other parts of the rays are
vertical or, more rarely, directed obliquely outward.

DIMENSIONS OF ACANTHTYLASTERS.

Erylus rotundus
var. megarhabda var. typica var.
ee cidaris
A B A B Cc D
Acanthtylasters ae
Bisel ene 3
= = |
oS =] n n
& | é
s 2 os z fa zg
g w = = =) 7A n $
= =) Gis = = = q —]
as E Z = o 5 =
= = = = 3 <4 z m 5
a = a = a 3 s S $
= S = s 3 3 = = on
5 GS} = ° g a = = a
Hi < = Rn < 4 4 3 3
total diameter,
24 24
Wt
: basal thickness
with two 1.5 1.5
of rays, #2
rays }
| diameter of
| acanthtyle 3.3 3.3
(rounded end), |

ERYLUS ROTUNDUS. 303
DIMENSIONS OF ACANTHTYLASTERS (continued).
Erylus rotundus
var. megarhabda var. typica | var.
| cidaris
A B A B (] D
Acanthtylasters = S
3 3
E ig 2
3] oO 3
a = a & =
3 s D n
3 aoe | S| z
g Sree ee iitclip gs hha 5
s z é a 3 3 s 3 5
3 oe 3 & ad >
5 2 a a 2 2 z = =
Sate G3 23 23-29 23-29 | 25-28 || 23-29
fe
: basal thick- 1.2- 1.2—- 1.4— =
1.4 14 =
ee eof mays, 2 1.4 14 |} 19 |} 1.9
rays
diameter of
acanthtyle 2.7 2.7 ay Po ell eee
(rounded end), /« = of
total diameter, || 14 97| 1g 26) 16-27 || 19-30 | 17-31| 17-23| 18-26| 17-31 || 16-26|| 16-31
L
7 A = =
with four | basal thickness |! 9 5 5 | 9.1 |9.8-2 ||0.7-2| 9 jos2| 2" |o05-2 |]1-24 || 9
to six | of rays, + 1.7 1.5 24
rays
diameter of || 13] 13-| 1.3-|] 1.5-| 1.2- ae
acanthtyle , Ee ete i! 2-3 14
ous || Teh Gig ||) Si eA 2.2
(rounded end),
total diameter, || 16 91 | 14-93] 14-23 || 15-19| 16-24| 15-24| 17-18 | 15-24 16-21 |] 14-24
“
with seven basal thickness 0.8— 0.7- 0.7— 12 0.9- 0.6— 12 0.6— 1-1.7 0.6—
to nine | of rays, riloe|| Sales als 2 | Tse I) ee 7 | ape} lest
rays a
diameter of ie is
acanthtyle 1-23 |1.3-2|1-23]] 15 [132/052] 15 052] 50 || os
(rounded end), y«
total diameter, }! j3_19| 12-17| 12-19 || 12-17| 15-19] 13-16| 16 |15-19|| 13 || 12-19
: ,
with ten to | basal thickness || 0.6- | 0.5-| 0.5- || 0.6-| 0.5-| 08-| 9, | 05-]) | 0.5-
fourteen | of rays, 1s 1% 1.5 1 Nas | es 13%) 123 135)
rays a
diameter of =
1.2- Tee a : 0.7
peantutyie 51 [P18 [1-21 ]146]072] 5 | 12 [022] 12 | oy
(rounded end), y«

304 ERYLUS ROTUNDUs.

It has been stated above, that oxyaster-like spicules similar in size and
ray-number to the acanthtylasters, which I consider a young form of the latter,
also occur in these sponges. These spicules are rather numerous in var. mega-
rhabda, form A, and met with in smaller numbers in var. megarhabda, form B,
and in the forms A and C of var. typica. The distal parts of the rays of these
asters are exceedingly slender. Proximally the rays thicken considerably and
abruptly, so that their basal part appears bulbous. Besides these spicules,
which I consider as the earliest known stages, others similar to them, but with
thicker and distally rough rays, representing a later developmental stage, are
observed. Finally various asters of this kind occur, in which a more or less
pronounced spiny thickening crowns the end of each ray. These asters connect
the slender-rayed oxyasters with the true acanthtylasters.

The small many-rayed oxyasters (Plate 6, fig. 14c; Plate 7, figs. 52c, 60c, 76c)

»2re without centrum or have a slight central thickening, in var. typica, form A,
up to 4 in diameter. There are from eight to twenty-two, or more, usually
from fourteen to twenty, equal, concentric, and regularly distributed rays.
The rays are, at the base, 0.4-0.9 y thick and conical, either throughout or only
at the end, and then nearly cylindrical in their basal part. They are always
sharp pointed and more or less spiny. Sometimes the spines are too small to be
discerned as such and their presence is indicated only by a certain roughness of
the rays. More often, however, particularly in the larger oxyasters, the spines
are large enough to be clearly made out. The larger the spines, the fewer their
number. Some of the spines frequently form a verticil some distance below the
end of the ray. Oxyasters of this kind were chiefly observed in var. typica,
form A. The total diameter of the oxyasters is 7-17 p. Those of var. typica
are a little smaller than those of the other two varieties. A few asters were
observed which appeared as transitions between these oxyasters and the acanth-
tylasters.

ERYLUS ROTUNDUS. 305

DIAMETERS OF SMALL OXYASTERS.

I Erylus rotundus

|
| var. megarhabda var. typica var.

|
\| = = ir —_ —= | eidaris
|
\ A B | A B C D }]
1 ||
— —— |
|
Small oxyasters (oxy- | | |]
sphaerasters) with || | =
numerous rays | || 5 |
| | a =| |
I} ey D 3
ic Cue ra R=) a || B
S ye bibles = =) a ae il 3
ey E ) ra ~ a 3 | rs) | o
= = a | = ran oa o | 5 | res
S 3 5 | <4 3S 3 a Ee =|
5 = = S = 3 = = | >
5 = 8 3 ae BS
R i a B i i Z rae | a
== |= === — = | | =
Total diameter, Le 7.5-14| 8-17 | 7.5-17 | 10-13} 8-14 | 8-16 | 7-12 7-16 | 8-14 7-17

The aspidasters (Plate 5, figs. 27f, 28f; Plate 6, figs. 17, 18; Plate 7, figs. 16-
30, 42-45; Plate 8, figs. 13, 14) are broad-oval or circular dises, often with a
somewhat irregular outline. This irregularity of outline generally does not
exceed that of the aspidasters represented in figs. 18, 19, and 25 on Plate 7;
occasionally, however, quite irregular aspidasters, with one or more deep inci-
sions reaching far into the interior, have been observed in all varieties. The
aspidasters are 50-77 y long, 46-70 » broad, and 4.4-8.8 y thick. Those of
var. megarhabda (50-66 by 46-59 by 4.6-8.8 4) are smaller than those of the
other two varieties. The general average proportion of length to breadth to
thickness of the aspidasters of all the forms (varieties) is 100: 93.3: 10.2. The
aspidaster-dise is either of uniform thickness throughout, or slightly thickened
in the middle. Its margin is simply rounded off. An umbilicus could not be
detected. All parts of the surface, the margin as well as the two faces, are
covered with protuberances. The largest protuberances are 0.7-2 # thick and
about as high. Most of them bear a terminal verticil of usually from three to
seven exceedingly small lateral spines. The large protuberances are usually
scattered rather irregularly over the surface. Occasionally some of the pro-
tuberances of the central part of the aspidaster lie in straight lines, radiating
from the centre of the disc. Between the larger protuberances small ones, just
perceptible with the strongest lenses, lie singly or in small groups on the other-

wise smooth surface of the disc.

306 ERYLUS ROTUNDUS.

DIMENSIONS OF ADULT ASPIDASTERS.

Erylus rotundus

var. megarhabda var, typica var. =|
| cidaris |; =
1] oo
| °
A B = A B Cc D 3 2
— | g eo
\| aD g =| aS =
|| nS ® © aS 2
Aspidasters \| Se E E = Se 22
Aspldasters | Se 3 3 3 Zs as
feos oo ‘3 a a z 2 om
} a as =< a4 = as 2a
i) oS =i 2 nS ra = Fics So
| = as ) 3 Cy 4a ss c=
|| = 2 = i] — 77 2 eS
hers) EZ = > A 5) ES =
= = = Ss
| a = I a 3 I 2 E °
S = =e + =| s Ss i 7
| 5 = ee = 3 =z B ee S 5
| 68 & = i) 3g S 3 s 3
wn hee bow z n =| a) ae
} 8) 8) 22) Bol se) See es
3 re} | =
| & & cate) & 8 & Sa alec}
—= = —— = a a ——" — —
ties a=!) Saal < yam eas > z 5
| limits, » | 50-62 | 53-66 | 50-66 |) 65-77 | 50-67 | 60-72 | 64-70 | 50-77 || 65-75 || 50-77

Length |
| average of the

three largest, ys

60 64 62 72 63 71 69 | 68.8 73 67.4

= 7 zm |

Breadth, limits, 46-56 | 49-59 | 46-59 || 54-68 | 48-61 | 57-70 | 61-69 | 48-70 || 62-69 || 46-70

Fess 3 Weeks 6.2— x 4.6— x e ~po| 4:4-| 44— 4.4—
Thickness, limits, » 88 4.6-5 88 5-7.8 | 5-6.4 | 6 6.8 76 78 5-7 88

Average proportion of || 100: | 100: | 100: 100: | 100: | 100: | 100: | 100: 100: 100:
length to breadth to || 91: 90: | 90.5: ots 94: 96: 95: 94: 96: || 93.3
thickness 12.3 8.2 | 10.3 7) Le 9.8 | 11.2 | 10.5 9.2 10.2

Thickness ing || 0.7—| 0.7- i |} ‘ }
lhieckness of protruding || 0 Ta | 0.7 0.7 1-2 |0.8-1 |0.7-2 | 1-1.9 | 0.7-2 || 0.7-1 || 0.7-2
rays, jt OM | ey 1

The youngest aspidasters observed were oval or circular discs, about half
the size of the adult, and composed of numerous, long, exceedingly slender,
perfectly straight rays, lying nearly in the same plane, and radiating from an
irregular, lobose structure 2-4 # in diameter. This lobose centrum is appar-
ently composed of short, terminally rounded concentric rays, arising obliquely
to the plane occupied by the long rays. The rays, which in this young stage
appear to be isolated throughout, grow in length and in thickness. Their
longitudinal growth leads to an increase in the size of the aspidaster, their
transverse growth (thickening) to a coalescence of the rays themselves, which,
as the growth continues, progresses from the centre towards the margin of the
dise. Thus the aggregations of isolated, radial rays become larger, solid dises
with serrated margins. This goes on until the spicule has attained nearly its
full size. Then the longitudinal growth of the rays ceases, while their lateral
growth continues. This leads to a filling up of the serrations and to the forma-

EE —

ERYLUS ROTUNDUS.

307

tion of smooth dises with more or less continuous margin (Plate 7, figs. 22S):

On the whole of the surface, the margin as well as the faces, of the disc, small

excrescences then make their appearance, and these grow out to form the pro-
tuberances above described.

LOCALITIES AND NATURE OF ENVIRONMENT.

VAR. MEGARHABDA LENDENFELD.

3349 | Ka Laau Light.

infeed files

April 8, 1802 |

(73448 f.)

(67 6° F.)

and corals.

38 ' Bee
os Locality Date Depth $a5 Bottom _ No. of
An = = + specimens
S. coast of Molokai, Lae-o 133-78 m 19.8° Coarse sand, a
3849 |Ka Laau Light. N. 71°,| April 8, 1902 (73-43 Fi ) (67 oS F) broken shells, Rene
Wie21:9'. < andicorals orm /
Vicinity of Kauai Island, 426-73 m 92° Coarse brown
3982 | Nawiliwili Light. N. 68°,| June 10, 1902 ay wee ee corals, — sand, | |, :
W. 1.6’ (233-40 f.) (48.5° F.) orale Form B
VAR. TYPICA LENDENFELD.
ae Hes 0
33 Locality Date Depth = Bs Bottom Bolas
v7 mae
S. coast of Molokai, Lae-o 133-78 m 19.8° Coarse — sand, 4
3849 | Ka Laau Light. N. 71°,| April 8, 1902 Ear ae Rc broken shells,
ey (73-43 f.) Grier | yer
Vicinity of Kauai Island, 44-79 m 93.99 Coarse coral, 3
4024 | Mokuaeuae Islet. S. 83°, E.| June 23, 1902 inte eae er |sand, and [o- ;
(24-43 f.) (73.72 F.), | ae Form B
7.0. raminifera
| Corals, sand,
Be eset of Mawaii, 44-152 m lacsnnise a 1
ie se zOe 8 1902 Me = 2 .
satel — men) tuly 18,1902 (24-83 f.) jules, and Fo-| Form D
S657 raminifera
Vicinity of Kauai Island, 165-327 m. g.g° | Coarse brown
4128 | Hanamaulu Warehouse. | August 1, 1902 253-68—90— (47 50 F.) | coral, sand, and Form C
N. 44° 30’, W. 2.6’. 179 f.) “* "? | Foraminifera.
VAR. CIDARIS LENDENFELD.
Ss | Eee ; No. of
gs Locality Date Depth | SRS Bottom specimens
i Qo
a |= <= = ==
S. coast of Molokai, Lae-o ee a Coarse _ sand, |
= ese | SES broken shells, 1
|

W. 21.9.

308 ERYLUS ROTUNDUS.

The thirteen sponges described above obviously form a systematic, though
far from a homogeneous, group, the specimens comprising it differing not incon-
siderably from each other in several respects. In three specimens the cortical
armour is chiefly composed of microrhabds, the aspidasters in it being but few
and confined to its external part. In the other ten the cortical armour is
composed chiefly of aspidasters, and the microrhabds which take part in its
formation are confined to the external part. In the three specimens the
microrhabds are considerably larger and the acanthtylasters and aspidasters
smaller than in the ten. Among these ten there is one which has smaller
oxyasters and relatively thinner aspidasters than the others, and which pos-
sesses aster-like rhabd-clusters, a kind of spicule not observed in any of the
others. Thus three secondary groups, megarhabda (armour chiefly composed
of microrhabds, microrhabds large, aspidasters small), typica (armour chiefly
composed of aspidasters, microrhabds small, aspidasters large, without rhabd-
clusters), and cidaris (armour chiefly composed of aspidasters, microrhabds-
small, aspidasters large, with rhabd-clusters) can be distinguished.

Two of the specimens of group megarhabda, which come from the south
coast of Molokai, are fairly identical, while the third, which comes from the
coast of Kauai, has a more grayish colour, smaller rhabd megascleres and micro-
rhabds, considerably larger triaenes, larger oxyasters, and thinner aspidasters.
Thus two somewhat different forms (A and B) are contained in the group
megarhabda.

In five of the nine specimens of group typica most of the rhabd megascleres
are amphistrongyles or very blunt amphioxes; in the other four most of these
spicules are sharp-pointed amphioxes. Four of the five specimens with chiefly
amphistrongyle rhabd megascleres, which come from the south coast of Molo-_
kai, are fairly identical with each other; the fifth, which comes from the north-
east coast of Hawaii, is not like these, massive, lobose, but elongate, digitate
in shape, having a much lighter colour, thicker rhabd megascleres, less centro-
tyle, for the most part simple amphiox microrhabds, somewhat larger euasters
and relatively thinner aspidasters. Of the four specimens with chiefly sharp-
pointed rhabd megascleres, which all come from the coast of Kauai, three (from
Station 4024) are fairly identical, while the fourth (from Station 4128) has a
lighter colour, slightly larger microrhabds, much smaller acanthtylasters and
longer, considerably broader and thinner aspidasters. Thus four forms (A, B,C,
and D) are contained in the group typica.

The surroundings of all the different forms of the same group must have

ERYLUS ROTUNDUS VAR. TYPICA. 309

been different to a certain extent, since they were found at different stations.
For this reason, and because the differences between them are not great and
their peculiarities more in the character of individual (somatic) adaptations
than of germinal qualities, I think that no greater systematic value than that
of local forms, two, A and B, in the group megarhabda, and four, A, B, C, and
D, in the group typica, can be attached to them.

The differences between the groups are much greater, and can hardly be
directly due to differences in the surroundings, since specimens belonging to
different groups were repeatedly captured together, at one and the same sta-
tion. This and their general nature lead me to consider the peculiarities, by
which these groups differ, not as mere somatic adaptations but as germinal
characters. Although most probably germinal in nature and certainly not in-
considerable, these differences are, in my opinion, nevertheless insufficient for
more than varietal distinction.

The characters of the three varieties are the following :—

Var. megarhabda.

Cortical armour composed chiefly of microrhabds. Rhabd megascleres
mostly sharp pointed, blunt forms rare; 330-650 by 6-13 4. Rhabd-clusters
absent. Microrhabds gradually and sharply pointed, central tyle small or ab-
sent; 43-98 by 3-7.5 ». Triaenes; rhabdome 170-370 » long; cladome 160-
350 » broad. Acanthtylasters with three or more rays; 12-27 yp in diameter.
Oxyasters 7.5-17 win diameter. Aspidasters 55-66 by 46-59 by 4.6-8.8 1; aver-
age proportion of length to breadth to thickness 100: 90.5: 10.3.

South coast of Molokai; coast of Kauai.

Var. typica.

Cortical armour composed chiefly of aspidasters. Rhabd megascleres
sharp pointed or blunt amphioxes, or amphistrongyles; 310-570 by 6-15 y.
Rhabd-clusters absent. Microrhabds variously pointed or blunt, with or with-
out central tyle; 30-66 by 1.5-4 ». Triaenes; rhabdome 200-220 yp long; clad-
ome 150-400 » broad. Acanthtylasters with three or more rays; 15-31 y in
diameter. Oxyasters 7-16 yp in diameter. Aspidasters 50-77 by 48-70 by
4.4-7.8 »; average proportion of length to breadth to thickness 100: 94: 10.5.

South coast of Molokai; coast of Kauai; northeast coast of Hawaii.

310 ERYLUS CALICULATUS.

Var. cidaris.

Cortical armour composed chiefly of aspidasters. Rhabd megascleres
chiefly sharp-pointed amphioxes; 440-650 by 8-12 4. Rhabd-clusters 125-
180 # long. Microrhabds gradually and sharply pointed, more or less centro-
tyle; 32-50 by 2-4.5 4. Triaenes; rhabdome 180-190 y long; cladome 300-
440 » broad. Acanthtylasters with two or more rays; 12-31 y in diameter.
Oxyasters 8-14 » in diameter. Aspidasters 65-75 by 62-69 by 5-7 #1; average
proportion of length to breadth to thickness 100: 96: 9.2.

South coast of Molokai.

The structure of the canal-system and spiculation of these sponges clearly
show that they belong to Erylus. They differ considerably from all the species
of this genus previously described, by the nearly circular shape and the small
size of their aspidasters and other characters. Their nearest ally is the species
here described as Hrylus caliculatus. By its shape being ealiculate, by its micro-
rhabds being amphistrongyle instead of amphiox, by its aspidasters being larger,
relatively much thicker, and partly reniform in shape, and by its spicules gener-
ally being larger and much stouter, this sponge differs from H. rotundus to

such an extent, that it must be considered specifically distinet from it.

Erylus caliculatus, sp. nov.
Plate 5, figs. 6-10, 24, 25, 29; Plate 6, figs. 1-13, 19-23, 26, 28, 29; Plate 7, figs. 11-15, 31-41, 74, 77,
78, 80; Plate 8, figs. 1-12, 15-20.

I establish this species for a specimen obtained on the northeast coast of
Hawaii (Station 4062). The name refers to its caliculate shape.

Shape and size. The sponge (Plate 5, fig. 29) appears as a broad, low,
truncate, inverted cone. It is 33 mm. high. The base of the cone, which forms
the upper side of the sponge, is irregularly oval in outline, depressed in the
middle, 47 mm. long, and 36 mm. broad. Its elevated margin is rounded, about
6 mm. thick, and partly divided into lobes. The base of attachment, which
corresponds to the truncate summit of the cone, measures 30 by 18 mm. The
surface is uneven and covered with shallow grooves, 0.5-1 mm. broad. These
grooves are particularly well marked on the protruding marginal lobes. Nu-
merous small circular pores, up to 0.8 mm. in diameter, are scattered over the
sides of the sponge. Apertures occur also on its depressed upper face, but
these are not so numerous and less uniform in size than those on the sides.
The largest of these apical apertures, which I am inclined to consider as oscules,

measure | mm. in diameter. The margin is free from pores.

ERYLUS CALICULATUS. 311

The colour of the sponge (in spirit) is dirty white.

The superficial part of the body is differentiated to form a cortex which con-
tains an aspidaster-armour and is about 65 # thick.

Canal-system. The, probably efferent, apertures on the depressed, terminal
face of the sponge are surrounded by sphincter-membranes. The flagellate
chambers (Plate 6, figs. 28, 29) are more or less spherical and 15-22 in diam-
eter.

The skeleton consists of rhabd megascleres, microrhabds, triaenes, acanth-
tylasters, slender-rayed oxyasters, small oxysphaerasters, and aspidasters.
Some of the rhabd megascleres form bundles (Plate 6, fig. 26a), others are iso-
lated and scattered. The bundles extend from the base upward and outward;
on nearing the surface they curve, where necessary, so as to abut steeply or
vertically on the cortex. The cladomes of the triaenes extend paratangentially
just below the cortical armour, their rhabdomes being directed radially inwards.
The triaenes occupy the interporal spaces (Plate 6, fig. 23) and often form well-
defined groups at the intersections of the interporal zones. The microrhabds
form a thin superficial layer overlying the aspidaster-armour and occupy the
mantles surrounding the cortical canals, chiefly their outer parts farthest from
the lumen. A few microrhabds are also found scattered in the choanosome.
The acanthtylasters and the slender-rayed oxyasters, which latter I consider as
young acanthtylasters, are numerous in all parts of the choanosome and extend
right up to the cortex and even into the mantles of the cortical canals. The
small oxysphaerasters are confined to the subcortical region and the mantles
of the cortical canals, on the inner surface of which they are often quite numer-
ous. The aspidasters occupy the proximal and middle parts of the cortical
armour in dense masses, leaving only the mantles of the cortical canals free.
Some aspidasters also occur scattered in the choanosome. The aspidasters in
the armour are mostly situated paratangentially. Next the mantles of the
cortical canals, however, they often assume other positions.

Most of the rhabd megascleres (Plate 5, figs. 24, 25) are very blunt amphi-
oxes (Plate 5, figs. 24a, 25a), many indeed so blunt that they can be considered
as amphistrongyles. A few blunt styles (Plate 5, fig. 24c) and branched rhabd,
derivates have also been observed. Most of the rhabds are more or less curved,
usually more strongly in their central part than at their ends. The rhabds are
410-850 » long and 10-19 » thick, the average measurements of the three
longest and thickest being 723 by 18 yp.

The microrhabds (Plate 6, figs. 1, 2, 4a, 5a; Plate 7, figs. 74a, 77a, 78, 80)

312 ERYLUS CALICULATUS.

are centrotyle amphistrongyles, usually more or less attenuated towards the
rounded ends. They are generally slightly curved, the curvature being uniform
or, rarely, greater near the ends than in the middle. Most of these spicules
are isoactine, anisoactine forms with one actine reduced in length being met
with only exceptionally. The microrhabds are 39-52 long and, near the middle,
close to the tyle, 3-5 u thick, the average measurements of the three longest
and thickest being 51 by 4.8 «. The tyle is 0.3-1 » more in transverse diameter
than the adjacent parts of the spicule.

Besides these regular microrhabds, branched microrhabd-derivates have
been observed in small numbers. In these spicules two or, rarely, more, short,
terminally rounded or, exceptionally, pointed branch-rays arise from a point
a little below one end of the spicule. In one of these spicules such branch-rays
were observed at both ends. When there are two branch-rays they usually
stand opposite each other in a straight line which intersects the axis of the
spicule at an angle of 40 to 60°.

The triaenes (Plate 5, figs. 6-10; Plate 6, fig. 23) are orthotriaenes. Their
rhabdome is straight, conical, 200-300 » long, and, at the cladomal end, 13-20 yu
thick. The clades are 125-220 » long, usually simple, and slightly and irregu-
larly curved. Rarely one clade bears a short branch-ray. The clades enclose
angles of 87-98° with the rhabdome. The breadth of the cladome is 210-380 y.

The acanthtylasters (Plate 6, figs. 4b, 5b, 6-13, 29; Plate 7, figs. 74, 77b;
Plate 8, figs. 2-12) usually have from four to eight, rarely only two or three,
concentric and regularly or, more rarely, irregularly distributed rays. The
rays are equal or one or more of them reduced in length. The properly
developed rays are, at the base, 1.54.5 « thick and taper towards the distal end,
which is usually crowned by an acanthtyle, rarely simply rounded off. The
acanthtyle or rounded end is 1-5 » in transverse diameter. The basal part of
the rays is either quite smooth or bears a few spines; from their central and
distal parts a good many spines arise. The acanthtyle (rounded end) is densely
covered with spines. The spines are usually conical, sharp pointed, and of con-
siderable size. The largest one observed was 1.5 » long. The spines arising
from the acanthtyle (rounded end) are generally very markedly recurved and in
their position resemble short flexible hair combed down in all directions from
the apex of the acanthtyle (rounded end). Those arising from the rays farther
down are either vertical or directed obliquely outward or inward. The total
diameter of the acanthtylasters is 17-50 ». This dimension and also the size
of the rays and acanthtyles are, on the whole, in inverse proportion to the ray-

ERYLUS CALICULATUS. 313

number. This proportional correlation is, however, not a uniformly regular one,
for while the three- to five-rayed acanthtylasters and their parts are nearly equal
in size, and the same applies to the seven- and eight-rayed ones, there are con-
siderable dimensional differences between the two- and three-rayed, the five-
and six-rayed, and six- and seven-rayed ones. In a curve representing this
correlation two steep, step-like falls would interrupt the general descent. The
dimensions of the two-rayed, three- to five-rayed, six-rayed, and seven- to eight-
rayed acanthtylasters are tabulated below.

ACANTHTYLASTERS.
Ray-number 2 3-5 6 7-8
Total diameter, 50 | 23-39 19234.) |) 17223
jt | =
Basal thickness of rays, 3 | 2-4.5 Te 3) elo
Transverse diameter of acanthtyle 45 | 2-5 1-43 125
(rounded end), y |

Many oxyasters, similar in diameter and ray-number to the acanthtylasters,
occur in the choanosome. The rays of these spicules are quite slender in their
distal part but usually thickened in a very marked manner at their base. Simi-
lar spicules with slightly thicker, rough rays, and others with still thicker and
rougher rays and a slight terminal thickening, connect the oxyasters with the
acanthtylasters. I consider the former as young stages of the latter.

The small oxrysphaerasters (Plate 6, fig. 3; Plate 7, figs. 11-15, 74c) have
a spherical centrum, 4-5.5 in diameter, from which from ten to twenty or more
equal, concentric, regularly distributed, conical, blunt or sharp-pointed rays arise.
The rays are, at the base, 0.7-1.4 » thick, and smooth. Their middle and distal
parts are covered with small spines. Occasionally a few spines, larger than the
rest, form a loose verticil some distance from the end of the ray. The whole
aster is 9-18 in diameter. A correlation (inverse proportion) between size and
ray-number is not discernible.

The aspidasters (Plate 6, figs. 19-22; Plate 7, figs. 31-41; Plate 8, figs. 1,
15-20) are dises varying from broad-oval to circular or reniform. Very rarely
aspidasters quite irregular in outline, with several deep incisions, have been
observed. The broad-oval to circular aspidasters (Plate 6, figs. 19, 20; Plate 7,
figs. 31-37), which are much more numerous than the reniform ones, are fairly
regular in outline, 72-88 y long, 67-77 broad, and 10.6-12.8 p thick, The

314 ERYLUS CALICULATUS.

average proportion of length to breadth to thickness is 100:95:15. In the
reniform aspidasters (Plate 6, figs. 21, 22; Plate 7, figs. 38-41), which are other-
wise similar to the broad-oval to circular ones, the margin is incised at a point
usually lying on one of the broader sides. This incision may be quite insigni-
ficant (Plate 7, fig. 39) or it may extend far into the interior of the spicule
(Plate 6, fig. 21). The reniform aspidasters have similar dimensions to the
broad-oval or circular ones, but are somewhat narrower, some not more than
63 » broad. The surface of the adult aspidaster is covered with short, truncate,
protruding rays, 1.8-4 p thick. These rays bear terminal verticils of lateral
spines, and usually also some spines arise from their apical faces. The number
of protruding rays is very variable. On some aspidasters (Plate 6, fig. 20; Plate
7, figs. 32-35, 38-40; Plate 8, figs. 15, 16) they are few and far between, in
others (Plate 6, fig. 19; Plate 7, figs. 36, 37, 41; Plate 8, figs. 17-20) very
numerous and, although usually distributed rather irregularly, nowhere very
far apart.

The youngest stages of the aspidasters (Plate 7, fig. 31; Plate 8, fig. 1)
appear as discs composed of slender rays radiating from a common centre.
The great majority of these rays are as long as the radius of the disc, lie nearly
in one plane, and extend from the centre to the margin of the disc. A few are
much shorter, and these are situated obliquely to the plane of the dise occupied
by the long rays (Plate 8, fig. 1). These ray-aggregations grow in the same
way as in Hrylus rotundus, described in detail, p. 306, and become smooth dises
(Plate 6, fig. 22), on the surface of which protruding rays later make their appear-
ance. It seems, a priori, probable that the aspidasters with few protuberances
(Plate 6, fig. 20; Plate 7, figs. 32-35, 38-40; Plate 8, figs. 15, 16) are young
forms which later, by the accession of further protuberances, are converted into
the aspidasters with numerous protruding rays (Plate 6, fig. 19; Plate 7,
figs. 56, 37, 41; Plate 8, figs. 17-20). Since, however, the protruding rays of the
aspidasters with but few of them seem to be on the whole larger than those of
the aspidasters with many of them, this is somewhat doubtful.

This sponge was caught with the tangles on the northeast coast of Hawaii,
Station 4062, Kauhola Light, 8. 69° 15’, E. 6.9’ on July 18, 1902; depth 152-
206 m. (83-115 f.); it grew on a bottom of coral, voleanic sand, shells, and
Foraminifera.

The structure of the canal-system and the spiculation of this sponge clearly
show that it belongs to Erylus. From all the species of this genus previously
described it differs considerably by the partly broad-oval to circular, partly

ERYLUS CALICULATUS. 315

reniform shape, and the small size of its aspidasters. It is nearest allied to
Erylus rotundus. From this E. caliculatus is distinguished by the shape, which
is massive lobose to digitate in H. rotundus and caliculate in EH. caliculatus;
by the microrhabds, which are amphiox in the former and amphistrongyle in
the latter; by the aspidasters which are absolutely smaller, relatively thinner,
and oval or circular in H. rotundus and absolutely larger, relatively thicker, and
in part also reniform in L. caliculatus; and by the spicules generally which are
smaller and much less robust in H. rotundus than in LE. caliculatus.

Il. GENERAL SYSTEMATIC ACCOUNT OF THE GENERA, SPECIES,
AND VARIETIES FROM THE PACIFIC OCKAN.

Erylidae.

Tetraxonida with rhabd and teloclade megascleres, and a superficial armour
composed of aspidasters and microrhabds. Euasters are always present in the

» choanosome.
For the present I place only one genus, Erylus, in this family.

ERYLUS Gray.

With uniporal afferents and uniporal efferents or larger oscules. Without

ana- or protriaenes.
Twenty-two species are known. Eight of these occur in the Pacific Ocean.

SUMMARY OF THE SPECIES FOUND IN THE PACIFIC OCEAN.

A, The large choanosomal euasters are oxyasters.
A, The teloclades are orthoplagiotriaenes.
A, The microrhabds have pointed ends.
E. placenta Thiele. E. monticularis Kirkpatrick.
B; The microrhabds are amphistrongyle centrotyles.
E. decumbens Lindgren.
B, The teloclades are chiefly dichotriaenes.
E. oxyaster Lendenfeld.
B, The large choanosomal euasters are acanthtylasters.
A, The aspidasters are about twice as long as broad.
E. nobilis Thiele. E. sollasii Lendenfeld.
B, ‘The aspidasters are nearly as broad as long.
E. rotundus Lendenfeld, var. megarhabda Lendenfeld, var. typica
Lendenfeld, var. cidaris Lendenfeld. E. caliculatus Lendenfeld.

316

ERYLUS DECUMBENS. 317

Erylus placenta THIeLr.

Zoologica, 1898, 24, p. 5, plate 1, fig. 1; plate 6, fig 1 a-h. Lenpenrecp, Tierreich, 1993, 19, p. 87.

Incrusting, 2-3 mm. thick. In spirit: reddish gray.

Amphioxes: mostly 700-800 by about 15 »; sometimes much shorter,
only half as long. Orthotriaenes: rhabdome 500, clades 270 » long.

Microrhabds: slightly curved, abruptly pointed centrotyle amphioxes;
25-50 long. Oxyasters: three to six, most frequently four rays; each ray 20-
30 » long. Oxysphaerasters: centrum about 5, whole aster 10 » in diameter.
Aspidasters: oval, very frequently with incised margin, irregular; 170-200 by
80-90 by 18 p.

Northwestern Pacific. Japan: Kagoshima Bay.

Erylus monticularis Kirkparrick.

Ann. mag. nat. hist., 1900, ser. 7, 6, p. 351, plate 14, fig. 3 a-h.

Thin, incrusting. Pale brown.

Amphioxes: 210 by 10 ». Orthotriaenes, rhabdome 6 y» thick; clades 186 p
long.

Microrhabds: amphiox; 40-80 by 1-3 p. Oxyasters: about 6 rays; total
diameter 18-30 2. Small tylasters (chiasters, Kirkpatrick): with small centrum;
about 12 rays; total diameter 10 4. Aspidasters: 150 long, 114 » broad.

Central Pacific. Funafuti Islet.

Erylus decumbens LinpGREN.
Zool. anz. 1897, 20, p. 485. Zool. jahrb. Syst., 1898, 11, p. 338, plate 20, fig. 1.
Erylus euastrum (Schmidt) LENDENFELD, Tierreich, 1903, 19, p. 86.

Incrusting, 3mm. high. Surface black, interior gray.

Amphiozxes: one end often blunt, 0.8 mm. by 24 p. Orthoplagiotriaenes:
rhabdome 420 by 28 p; clades curved, concave to rhabdome, 280 «long; clade-
angles nearly 90° (according to text), 104° (according to figure).

Microrhabds: curved, centrotyle amphistrongyles, 60 by 6 yp. Oxyasters:
two to five smooth rays; each ray 24 “long. Sphaerasters: numerous rays;
total diameter 10 4. Aspidasters: oval, some with incisions; 182 by 120 by 28

Formerly I was inclined (loc. cit.) to consider this species as identical with
Stelletta euastrum Schmidt (1868) and Erylus cylindrigerus Ridley (1884), but the
reexamination has made me doubtful on this point, so that I now revert to
Lindgren’s name.

Western Pacific: Java.

318 ERYLUS SOLLASII.

Erylus oxyaster LeNpDENFELD.

Ante, p. 268.

Massive with lobose or digitate protuberances. In spirit: brown, part of
the surface lighter than the rest.

Amphioxes: pointed, rarely blunt: 1.8-2.9 mm. by 60-85 yp. Styles: rare;
1.9-2.3 mm. by 60-105 ». Angularly bent and branched rhabd-derivates: in
dimensions similar to the amphioxes, rare. Plagiotriaenes: rare; rhabdome 0.9
mm. by 75-90 »; clades 0.7 mm. long; clade-angles 109-112°. Dichotriaenes:
rhabdome 0.6-1.6 mm. by 70-105 #; main clades 250-400, end clades 40-450 pz
long; clade-angles 109-120°. Irregular dichotriaene-derivates: with the clades
reduced in number or the rhabdome reduced in length; rare.

Microrhabds: more or less curved, centrotyle, generally isoactine amphi-
strongyles; 31-47, rarely up to 93 y long, 3.5-4.5 » thick. Oxyasters: usually

* with a slight central thickening; one to twenty or more perfectly smooth, coni-
cal rays; total diameter 10-90 4. Aspidasters: oval, rarely roundish or irregu-
lar; the ordinary oval ones 208-243 by 125-150 by 30-40 yp.

fastern Pacific. Galapagos Islands: 0° 50’ S., 89° 36’ W. ‘‘Albatross”’
Station 2809. ;
Erylus nobilis THre.e.
Abhandl. Senckenb. gesellsech., 1900, 25, p. 48, plate 2, fig. 17. LeNDENFELD, Tierreich, 1903, 19,

p. SD.

Irregularly cylindrical. White; in the interior brownish.

Amphioxes: rather abruptly pointed; nearly 1 mm. by 30 4. Orthotriaenes:
rhabdome 600 by 40 4; clades 250 long, slightly curved.

Microrhabds: centrotyle amphistrongyles; about 48 by 6. Acanthtylas-
ters: most frequently seven rays; each ray 20 wlong. Aspidasters: oval, outline
sometimes irregular; 190 by 90-100 by 40 p.

Western Pacific. Ternate.

Erylus sollasii LeNDENFELD.

Ante, p. 272.

Irregularly massive, with lobose, gyriform, or short digitate processes. In
spirit: whitish to chestnut-brown or purplish brown; one part of the surface
sometimes darker than the rest.

Blunt amphioxes: 425-980 by 8-24 p. Sharp-pointed amphioxes, amphi-
strongyles, and styles of similar dimension; rare. Orthoplagiotriaenes: rhabdome

ERYLUS ROTUNDUS. 319

140-520 by 8-22 p, rarely reduced in length and thickened; cladomes very poly-
morphic; clades simple or bifurcate; simple clades 120-300 » long; bifurcate
clades, main clades 70-270, end clades 10-160 long; clade-angles 86-116°;
in some specimens all the triaene-clades are simple, in some the majority are
simple, the minority bifurcate, in some the majority are bifurcate.

Microrhabds: curved, centrotyle, pointed; 30-78 by 2.5-5 yp. Anisoactine
and branched microrhabd-derivates: of similar dimensions; rare. Acanthtyl-
asters: two to fourteen or more rays; total diameter 10-38 p. Aspidasters:
oval, rounded rhomboidal, or irregular with lobose marginal protuberances ; 95—
156 by 55-82 by 7.4-14 yp.

Central Pacific. Hawaiian Islands: south coast of Molokai. ‘‘ Albatross”
Stations 3847, 3848, 3849; northeast coast of Hawaii. ‘‘ Albatross”? Stations
4055, 4062.

Erylus rotundus LenpeNnreLp.

Ante, p. 290.

Massive, oval or somewhat irregular, with lobose, gyriform, or digitate pro-
tuberances. Surface usually more orless rugose. In spirit: dirty white to light
brown in the interior; surface dirty white to purplish brown or purplish gray or
purplish black, one part of the surface often much darker than the rest. Some-
times with whitish spots, marking the position of the pores, on the darker parts
of the surface.

Rhabd megascleres: sharp-pointed amphioxes, blunt amphioxes, or am-
phistrongyles, variously combined, 310-650 by 6-15 y. Styles, angularly bent
and branch-bearing rhabd-derivates: of similar dimensions; rare. Aster-like
rhabd-clusters: four to ten conical, irregularly distributed, longer, and a number
of very short rudimentary, cylindrical, terminally rounded rays; total diameter
125-180 y; only in variety cidaris. Orthoplagiotriaenes: rhabdome 170-370
by 6-12 »; clades often slightly, irregularly curved; 80-270 long; clade-angles
90-107°.

Microrhabds: slightly curved, variously but generally sharply pointed ;
30-98 by 1.5-7.5 »; with or without central tyle, the latter when present small.
Acanthtylasters: two to fourteen rays; total diameter 12-31 ». Small oxyasters:
eight to twenty-two or more, conical, spined rays; total diameter 7-17 y.
Aspidasters: broad oval to circular; 50-77 by 46-70 by 4.4-8.8 y..

Central Pacific. Hawaiian Islands: south coast of Molokai. ‘‘ Albatross”
Station 3849; coast of Kauai. ‘Albatross” Stations 3982, 4024, 4128; north-
east coast of Hawaii. ‘‘ Albatross” Station 4061.

290 ERYLUS ROTUNDUS VAR. CIDARIS.

Erylus rotundus var. megarhabda LeNpENFELD.

Ante, p. 309.

Cortical armour composed chiefly of microrhabds. Rhabd megascleres:
mostly sharp pointed; blunt forms rare; 330-650 by 6-13 y. Ehabd-clusters
absent. Orthoplagiotriaenes: rhabdome 170-870 » long; cladome 160-350 4
broad. Microrhabds: gradually and sharply pointed; 43-98 by 3-7.5 4; cen-
tral tyle small or absent. Acanthtylasters with three or more rays; 12-27
in diameter. Small oxyasters: 7.5-17 in diameter. Aspidasters : 55-66 by 46-
59 by 4.6-8.8 yu.

Central Pacific. Hawaiian Islands: south coast of Molokai. ‘‘Albatross”’
Station 3849; coast of Kauai. ‘‘ Albatross” Station 3982.

Erylus rotundus var. typica LeENDENFELD.

Ante, p. 399.

Cortical armour composed chiefly of aspidasters. Rhabd megascleres:
sharp pointed or blunt amphioxes, or amphistrongyles; 310-570 by 6-15 yp.
Rhabd-clusters: absent. Triaenes: rhabdome 200-220 y long, cladome 150-
400 broad. Microrhabds: variously pointed or blunt, with or without central
tyle; 30-66 by 1.5-4 . Acanthtylasters: with three or more rays; 15-31 yp in
diameter. Oxyasters: 7-16 # in diameter. Aspidasters: 50-77 by 48-70 by
44-7.8 py.

Central Pacific. Hawaiian Islands: south coast of Molokai. ‘Alba-
tross”’ Station 3849; coast of Kauai. ‘‘Albatross”’ Stations 4024, 4128; north-
east coast of Hawaii. ‘‘ Albatross’? Station 4061.

Erylus rotundus var. cidaris LeNDENFELD.

Ante, p. 310.

Cortical armour composed chiefly of aspidasters. Rhabd megascleres: chiefly
sharp-pointed amphioxes; 440-650 by 8-12 ». Rhabd-clusters: 125-180 y long.
Triaenes: rhabdome 180-190 » long, cladome 300-440 y» broad. Microrhabds:
gradually and sharply pointed, more or less centrotyle; 32-50 by 2-4.5 yp.
Acanthtylasters: with two or more rays; 12-31 in diameter. Small oxyasters:
8-14 in diameter. Aspidasters : 65-75 by 62-69 by 5-7 p.

Central Pacific. Hawaiian Islands: south coast of Molokai. ‘Albatross ”’
Station 3849.

ERYLUS CALICULATUS.

OO
bo
pt

Erylus caliculatus LenpENFELD.

Ante, p. 310.

Inverted conical, caliculate. In spirit: dirty white.

Blunt amphioxes and amphistrongyles: 410-850 by 10-19 4. Orthotriaenes:
rhabdome 200-300 by 13-20 yp; clades 125-220 y long; clade-angles 87-98°.

Microrhabds: centrotyle amphistrongyles, usually attenuated towards
the ends; 39-52 by 3-5 yw. Acanthtylasters: usually four to eight, rarely only two
or three rays; total diameter 17-50 ». Small oxysphaerasters: ten to twenty
or more rays; centrum 4-5.5, whole aster 9-18 y in diameter. Aspidasters:
broad-oval to circular, or, not so frequently, reniform; the broad-oval to circular
ones 72-88 by 67-77 by 10.6-12.8 ; the reniform ones often narrower.

Central Pacific. Hawaiian Islands: northeast coast of Hawaii. ‘‘Alba-
tross”’ Station 4062.

WI. DISTRIBUTION.

If the fauna of the Pacific so far as it relates to the Erylidae is compared with
that of other regions, it is seen that its single genus Erylus occurs in both.
Some of the Pacific species are similar to species found outside the Pacific, but
not asingle one appears to be really identical with any ultra-Pacific one.

Right of the twenty-two known species, that is 36 %, occur in the Pacific.

» The range of the Pacifie species is, so far as at present known, not great.
With the exception of Hrylus rotundus and E. sollasii, which have been obtained
at various points on the coasts of the Hawaiian Islands, all the Pacifie species
are recorded from one locality only.

The majority of species, the Pacific as well as those of other regions, are
tropical or subtropical. It is remarkable that no specimen of Erylus has hitherto
been found on the west coast of the American continent.

The horizontal distribution of the Pacifie species is:—
Eastern Pacific Islands.
Erylus oxyaster. Galapagos.

Central Pacific Islands.

Erylus monticularis. Funafuti.
“ ~ sollasii. Hawaiian Islands.
“~~ rotundus. eS a
“~~ caliculatus. 7

Southeastern Asiatic Islands.

Erylus decumbens. Java.

(zs

nobilis. Ternate.
Northeastern Asiatic Islands.

Erylus placenta. Japan.

All the Pacific species have been found in rather shallow water.

322

————

iv. sLISt OF

STATIONS.

||| ees
fe |§8/ sé
Locality Lat. Long. Date =e 3 z g g Bottom Instruments used
BS ee | Se
o Ao
mn - &
\s 7 7 hee! 7, Ol aw,
Gala S - ‘
fe BDBEO S|$.005000 W.893600| April 4,1888) 45 79° | 741° | Gray sand Small beam trawl
'S. coast of Molo- Tete,
kai, Lae-o Ka| N.643000 W. 230000| April 8,1902| 23-24 | 76 Sand and stones | Hand lead, 6 foot
: Light hemp tangles
‘sg. coast of Molo- Sigsbee sounding ma-
‘kai, Lae-o Ka|N.681500 W.22 400} “ “ “ | 44-73 | 76 | 71.1 | Sand and gravel Cutts SEES, D7
-Laau Light net, 5} foot Blake
beam trawl, ete.
coast of Molo- Coarse sand, ee = =
Lae Ka|N.710000 W.210900| “ “ “ | 73-43 | 76 | 67.6 | brokenshells,and | CU"® 10 foot Blake
‘Laau Light Banas beam trawl, surface
¢ tow net
nity of Kauai Gonrsetabroral Sigsbee sounding ma-
and, Nawili- | N.680000 W. 010600 | June 10,1902) 233-40 | 78 48.5 | corals sand shells | Cline, 9 foot hemp
Light J 4 tangles
icinity of Kauai Gane my anal Sigsbee sounding ma-
Island, Mokuaeae | S.830000 E. 07 0600 | June 23,1902 24-43 75 73.7 | snd Foraminifera | Cine: 9 foot hemp
Islet tangles
eee ast yor Tanner sounding ma
| Hawaii, Alia Fine gray sand,| 7). =
Point Light, Hilo N.200000 W. 0305 00 | July 16,1902) 50-62 76 and Foraminifera chine, 8 foot hemp
Day tangles
3; ca E. coast of Corals, sand, | Tanner sounding ma-
awaii, Kauhola | $.790000 E. 060700} July 18,1902) 24-83 77 coralline nodules, | chine, 8 foot hemp
Light and Foraminifera | tangles
=
_ |N. E. coast of Coral, voleanic | Tanner sounding ma-
Hawaii, Kauhola | 8.691500 E.060900; “ “ “ | 83-113 77 sand, shells, and | chine, 8 foot hemp
Light Foraminifera tangles
‘Vicinity of Kauai Goarme Brawn Sigsbee sounding ma-
253-68-— chine, 8 foot Albatross
Island, Hanama-| N.443000 W. 020600 | Aug. 1, 1902 77 47.8 | coral, sand, and Blare be:
: 90-179 eae pattern Blake beam
_ | ulu Warehouse Foraminifera ie

323

=
cA
of

‘PLANATION OF THE PLATES.

32

PLATE 1.

Erylus sollasii LENDENFELD.
Figures 1-48.

1.— Transverse section of a lamellar part of a large dark specimen from South Molokai (race II); mag-
nified 7.5; phot. Zeiss, planar 50 mm.:

a, cortex; b, choanosome; ce, monaxonid symbiont attached to one side of the Erylus.

2.— Part of the surface of a large whitish specimen from South Molokai (race I, form C); magnified 7.5;
phot. Zeiss, planar 50 mm.

3.— Subcortical portion of a radial section occupied by numerous large spherical granular cells, of a
large dark specimen from South Molokai (race II); haematoxylin; magnified 209; phot. Zeiss,
apochr. 8, compens. oe. 6.

4.— Part of a radial section through the choanosome of a large dark specimen from South Molokai (race
II); haematoxylin; magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6.

5-12.— Apical views of cladomes of triaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:

5, of a triaene with one simple and two bifureate clades of a large dark specimen from South Molo-
kai (race IT);

6, of a regular dichotriaene with very short end clades of a large dark specimen from South Molokai
(race IT);

7, of a regular plagiotriaene of a large dark specimen from South Molokai (race IT);

8, of an irregular dichotriaene-derivate in which one end clade of each of the three pairs is reduced,
of a middle-sized light-coloured specimen from northeast Hawaii (race I, form D);

9, of a regular dichotriaene with long end clades of a middle-sized light-coloured specimen from
northeast Hawaii (race I, form D);

10, of a triaene with one trifurcate and two bifureate clades of a middle-sized light-coloured specimen
from northeast Hawaii (race I, form D);

11, of a regular plagiotriaene of a large whitish specimen from South Molokai (race I, form C);

12, of a regular plagiotriaene of a middle-sized light-coloured specimen from northeast Hawaii
(race I, form D).

13-26.— Side views of triaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:

13, of a regular plagiotriaene with a rhabdome reduced to a short, conical protuberance, of a large
dark specimen from South Molokai (race IT) ;

14, of a triaene with one bifurcate and two simple clades, and a rhabdome reduced to a short,
conical protuberance, of a large dark specimen from South Molokai (race II) ;

15, of a plagiotriaene with simple, unequal, stout, blunt clades and a regular blunt rhabdome, of a
large whitish specimen from South Molokai (race I, form C) ;

16, of a plagiotriaene with equal, cylindrical, terminally rounded clades and a rhabdome bearing
small rounded protuberances near the end, of a large whitish specimen from South Molokai
(race I, form C);

17, of a plagiotriaene with slightly unequal, slender, blunt clades and a regular pointed rhabdome,
of a large whitish specimen from South Molokai (race I, form C);

18, of an irregular orthotriaene with unequal stout clades partly rounded at the end, and a reduced
cylindrical, terminally rounded rhabdome, of a middle-sized light-coloured specimen from north-
east Hawaii (race I, form D);

19, of a plagiotriaene with fairly equal, slender, pointed, regular clades and a fairly pointed rhabd-
ome, of a middle-sized light-coloured specimen from northeast Hawaii (race I, form D);

20, of a plagiotriaene with slender, regular clades, one of which is strongly curved at the end, of a
middle-sized light-coloured specimen from northeast Hawaii (race I, form D); -

21, of a fairly regular plagiotriacne with short pointed clades, of a large dark specimen from South
Molokai (race IT);

22, of a triaene with one bifurcate clade, of a large dark specimen from South Molokai (race IT);

23, of a protriaene with unequal, cylindrical clades, reduced in length and terminally rounded,
of a large dark specimen from South Molokai (race IT);

24, of a plagiotriaene with one clade terminally recurved and a rhabdome bearing slight protuber-
ances near the end, of a large dark specimen from South Molokai (race IT) ;

25, of a rather stout plagiotriaene with relatively long clades, of a large dark specimen from South
Molokai (race IT);

26, of a slender plagiotriaene with relatively long clades of a large dark specimen from South Molokai
(race IT).
27.— A large whitish specimen from South Molokai (race I, form C); natural size; phot. Zeiss, anastig.
480/412 mm.
28.— A large dark specimen from South Molokai (race IT); natural size; phot. Zeiss, anastig. 480/421
mm.
29-35.— Rhabd megascleres; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
29-31, more or less amphiox-rhabds tapering towards both ends, of a large whitish specimen from
South Molokai (race I, form C);
32, stout and short, somewhat irregularly cylindrical amphistrongyle, of a large whitish specimen
from South Molokai (race I, form C);
33, 34, sharp-pointed amphioxes, of a small dark specimen from South Molokai (race I, form B);
35, pointed amphiox of a large dark specimen from South Molokai (race IT).
36-41. Groups of microscleres from centrifuge-spicule preparations; magnified 300; phot. Zeiss,
apochr. 4, compens. oc. 6:
a, isoactine centrotyle microrhabds; b, anisoactine microrhabds; c, acanthtylasters; d, young
aspidaster ;
36, 39, 41, of a large whitish specimen from South Molokai (race I, form C);
37, of a small dark specimen from South Molokai (race I, form B);
38, of a large dark specimen from South Molokai (race IT);
40, of a middle-sized light-coloured specimen from northeast Hawaii (race I, form D).
42-48.— Rhabd megascleres; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
42, 43, large, more or less pointed amphioxes, of a middle-sized light-coloured specimen from north-
east Hawaii (race I, form D);
44, style. abruptly bent close to the blunt end, of a middle-sized light-coloured specimen from
northeast Hawaii (race I, form D);
45, amphistrongyle, of a middle-sized light-coloured specimen from northeast Hawaii (race I,
form D);
46, small blunt amphiox, of a middle-sized light-coloured specimen from northeast Hawaii (race I,
form D);
47, 48, small, sharp-pointed amphiox, of a small, partly light, partly dark, specimen from South
Molokai (race I, form A).

SPONGES OF THE PACIFIC, II. ERYLIDAE.

Fig. 1-48 Erylus sollasii n. sp.

4, 3-7, 13, 14, 21-26, 28, 35, 38 race II; 2, 11, 15-17, 27, 29-32, 36, 39, 41 race I, for
33, 34,37 vace I, form B; 47,48 race I, form A.

PLATE 2.

a
_

PLATE 2.

Erylus sollasii LenpeNrELp.
Figures 1-26.

1, 2.— A large acanthtylaster of a large dark specimen from South Molokai (race II); magnified 1800;
u. v. phot. Zeiss, q. monoehr. 1.7, q. oc. 10:
, focused higher; 2, focused lower.
3, 4.— Group of acanthtylasters from a centrifuge-spicule preparation; magnified 540; u. v. phot. Zeiss,
q. monochr. 6, q. oc. 10:
3, of a large whitish specimen from South Molokai (race I, form C);
4, of a large dark specimen from South Molokai (race IT).
5-7.— The central part of an aspidaster of a large dark specimen from South Molokai (race II); mag-
nified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
5, focused high (on the upper surface); 6, focused intermediate; 7, focused low (on the centrum).
8, 9.— A large acanthtylaster of a large eee specimen from South Molokai (race II); magnified 1800;
> u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
8, focused higher; 9, focused lower.
10, 11.— A small acanthtylaster of a large dark specimen from South Molokai ee II); magnified 1800;
u. v. phot. Zeiss, q. monochr. 1.7, q. oe. 10:
10, focused lower; 11, focused higher.
12-15.— Aspidasters; magnified 540; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
12, somewhat irregular adult aspidaster, of a large whitish specimen from South Molokai (race I,
form C);
13, regular, oval adult aspidaster, of a large dark specimen from South Molokai (race II);
14, very young aspidaster composed of radiating rays, of a large dark specimen from South Molo-
kai (race IT);
15, elongate, young, perfectly smooth aspidaster, of a large dark specimen from South Molokai
(race IT).
16.— Group of isoactine microrhabds, of a large dark specimen from South Molokai (race II); magnified
540; u. v. phot. Zeiss, q. monochr. 6, q. oe. 10.
17.— Isoactine microrhabd, of a large whitish specimen from South Molokai (race I, form C); magni-
fied 540; u. v. phot. Zeiss, gq. monochr. 6, q. oc. 10.
18.— Isoactine microrhabd, of a large dark specimen from South Molokai (race II); magnified 1800;
u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.

19-26.— Aspidasters; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
19, young aspidaster, of a small, partly light, partly dz fs specimen from South Molokai (race I,
form A);

20, young aspidaster, of a large whitish specimen from South Molokai (race I, form C);

21, young aspidaster, of a middle-sized light-coloured specimen from northeast Hawaii (race I,
form D);

22, adult, lozenge-shaped aspidaster, of a small, partly light, partly dark, specimen from South

Molokai (race I, form A);

23, adult, oval aspidaster, of a middle-sized light-coloured specimen from northeast Hawaii (race I,
form D);

24, adult, irregular aspidaster, of a small, partly light, partly dark, specimen from South Molokai
(race I, form A);

25, adult, irregular aspidaster, of a large whitish specimen from South Molokai (race I, form C);

26, adult, broad oval aspidaster, of a middle-sized light-coloured specimen from northeast Hawaii
(race I, form D).

SPONGES OF THE PACIFIC, IL. ERYLIDAE. PLATE 2.

0;
|
f
;
;

om) 50 oo

erect tae oe

Be ye
og oe >
ay

G5
Ge, e8-
2
CNG: “

)
SA Pa 4
moro 9?

‘

Fig. 1-26 Erylus sollasit n. sp. : 2
1, 2, 4-11, 13-16, 18 race II; 3, 12,17, 20, 25 race I, form C; 19, 22, 24 race I, form A; 21, 23

RS
Xa
a

S

8

ont
>
ef

3

S

PLATE 3.

PLATE 3.

Erylus sollasii LenpeNFELD.
Figures 1-28. Race III. Large dark specimen from northeast Hawaii.

1.— Side view of a triaene with branched clades, and a reduced, cylindrical, terminally rounded rhabd-
ome; magnified 100; phot. Zeiss, apochr. 16, compens. oe. 6.
2-6.— Apical views of cladomes of triaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
2-4, with all clades branched dichotomously or in a more complicated manner;
5, 6, with only two clades thus branched and the third simple.
7, 8.— Parts of large acanthtylasters; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7, q. oe. 10.
9-11.— Aspidasters; magnified 350; phot. Zeiss, apochr. 4, compens. oc. 6:
9, a perfectly adult one;
10, a nearly adult one;
11, a young, still smooth one.
12.— Side view of a triaene with branched clades; magnified 100; phot. Zeiss, apochr. 16, compens.
oc. 6.
13.— Two isoactine microrhabds; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6.
14, 15.— Two isoactine microrhabds; magnified 540; u. v. phot. Zeiss, q. monochr. 6, q. oc. 10.
16, 17.— Aspidasters; magnified 540; u. v. phot. Zeiss, q. monochr. 6, q. oc. 10:
16, a young one;
17, an adult one.
18.— Promonaene-like style, abruptly bent near the pointed end; magnified 100; phot. Zeiss, apochr.
16, compens. oc. 6.
19-22.— Rhabd megascleres; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
19, 20, large amphioxes;
21, style;
22, small amphiox.
23, 24.— Side views of triaenes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
23, with three-branched clades;
24, with simple and two-branched clades.
25, 26.— Groups of microscleres from centrifuge-spicule preparations; magnified 540; u. v. phot. Zeiss,
q. monoehr. 6, q. oc. 10:
a, microrhabd; b, small, most probably foreign, sphaeraster; ec, large acanthtylasters; d,
very young aspidasters.
27.— View of the largest specimen; natural size; phot. Zeiss, anastig. 480/412 mm.
28.— Part of the surface of an adult aspidaster; magnified 1800; u. v. phot. Zeiss, q. monochr. 1.7,
q. oc. 10.

Erylus oxyaster LeNDENFELD.
Figures 29-35.

29-35.— Microscleres and groups of such from centrifuge-spicule preparations; magnified 600; u. v.
phot. Zeiss, q. monochr. 6, q. oc. 10:
a, (Figs. 29-32, 35) microrhabds;
b, (Fig. 35) large diactine oxyaster;
c, (Fig. 34) large triactine oxyaster with one ray reduced;
d, (Figs. 32, 33, 35) large oxyasters with three or more fully developed rays;
e, (Fig. 32) small oxyaster with numerous rays. ;

PLATE 3.

- SPONGES OF THE PACIFIC, Il. ERYLIDAE.

Brylus oxyaster n. Sp.

Fig. 1-28 Erylus sollasit n. sp. (race [11].
Fig. 29-35

<i
a
Eo} ?
|
=

PLATE 4.

Erylus oxyaster LENDENFELD.
Figures 1-43.

1-23.— Megascleres; magnified 30; phot. Zeiss, planar 20 mm.:
1, apical view of the cladome of a dichotetraene;
2-5, apical views of cladomes of dichotriaenes;
6, 7, pointed, straight, or slightly curved amphioxes;
8, pointed, angularly bent amphiox;
9, blunt, slightly curved amphiox (amphistrongyle) ;
* 10, 11, styles;
12-14, amphiox-derivates, strongly angularly bent near one end, like monaenes;
15, 16, amphiox-derivate with a vertically arising branch-ray, like mesomonaenes;
17-19, amphiox-derivates with two or three branch-rays, like meso- or amphiclades;
20, side view of a plagiotriaene;
21, side view of a triaene with one simple and two bifureate clades;
22, 23, side views of dichotriaenes.
24.— View of the sponge; natural size; phot. Zeiss, anastig. 480/412 mm.
25.— Part of a superficial, paratangential section showing a pore; magnified 200; phot. Zeiss, apochr. 8,
compens. oc. 6.
26-34.— Mieroscleres and groups of such from centrifuge-spicule preparations; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6.:
a, (Figs. 26, 28-33) microrhabds;
b, (Fig. 27) large monactine oxyaster;
c, (Fig. 28) large diactine oxyaster;
d, (Figs. 26, 30) large triactine oxyasters;
e, (Pigs. 30, 32-34) large oxyasters with four or more rays;
f, (Figs. 29, 30) small oxyasters with numerous rays.
35.— The umbilical face of an aspidaster; magnified 300; phot. Zeiss, apochr. 4, cor ~ons. oc. 6.
36, 37.— Part of the surface of the umbilical face of an aspidaster; magnified 2000- .. v. phot. Zeiss,
q- monocehr. 1.7, q. oc. 10:
36, focused lower; 37, focused higher. F
38.— A ray of a large oxyaster; magnified 2000; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10.
39, 40.— A small oxyaster with numerous rays; magnified 2000; u. v. phot. Zeiss, q. monochr. 1.7, q.
oc. 10: . e y;
39, focused lower; 40, focused higher.
41—43.— Aspidasters; magnified 150; phot. Zeiss, apochr. 8; compens. oc. 6:
41, rare, nearly circular aspidaster;
42, 43, ordinary, elongate aspidasters.

~

SPONGES OF THE PACIFIC, II. ERYLIDAE. PLATE 4.

Fig. 1-43 Erylus oxyaster 2. Sp.

PLATE 5.

PLATE 5.

Erylus rotundus LENDENFELD.

3, 11-13. — var. typica LeNDENFELD. Form D. Northeast Hawaii.
4, 17.

I
Figs. 2, — var. typica LENDENFELD. Form B. Kauai.

Figs. 5, 26-28, 31. — var. cidaris LENDENFELD.

Figs. 14, 15. — var. typica LeENDENFELD. Form C. Kauai.

Figs. 16, 30. var. typica LENDENFELD. Form A. South Molokai.

Figs. 18-22, 32. — var. megarhabda LENDENFELD. Form A. South Molokai.
Fig. 23. — var. megarhabda LENDENFELD. Form B. Kauai.

Erylus caliculatus LeNnDENFELD.
Figures 6-10, 24, 25, 29.

1-7.— Apical views of triaene-cladomes; magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6:
1, 3, of Brylus rotundus var. typica from northeast Hawaii (form D);
2, 4, of a dark specimen of Hrylus rotundus var. typica from Kauai (form B);
5, of Erylus rotundus var. cidaris;
6, 7, of Erylus caliculatus.
S-10.— Side views of triaenes of Prylus caliculatus; magnified 100; phot. Zeiss, apochr. 16, compens.
oc. 6.
11-15.— Amphioxes of Erylus rotundus var. typica; magnified 100; photo. Zeiss, apochr. 16, compens.
oc. 6:
11-13, of a specimen from northeast Hawaii (form D);
14, 15, of a light specimen from Kauai (form C).
16, 17.— Groups of megascleres from spicule-preparations of HZrylus rotundus var. typica; magnified
100; phot. Zeiss, apochr. 16, compens. oc. 6:
16, of a specimen from South Molokai (form A);
17, of a dark specimen from Kauai (form B);
a, sharp-pointed amphioxes; b, blunt amphiox (amphistrongyle); c, style.
18—23.— Rhabd megascleres of Erylus rotundus var. megarhabda; magnified 100; phot. Zeiss, apochr.
16, compens. oe. 6:
18, an amphistrongyle of a specimen from South Molokai (form <A);
19-22, amphioxes of a specimen from South Molokai (form A);
23, an amphiox of a specimen from Kauai (form B).
24-28.— Groups of spicules from spicule-preparations; magnified 100; phot. Zeiss, apochr. 16, com-
pens. oe. 6:
24, 25, of Erylus caliculatus;
26-28, of Brylus rotundus var. cidaris;
a, (Pigs. 24-28) amphioxes;
c, (Fig. 24) style;
d, (Fig. 26) triaene;
e, (Fig. 27) microrhabd;
f, (Figs. 24, 27, 28) aspidasters;
g, (Figs. 24, 27) acanthtylasters.
29-32.— Views of whole specimens; phot. Zeiss, anastig. 480/412 mm.:
29, Erylus caliculatus; natural size;
30, Erylus rotundus var. typica from South Molokai (form A); magnified 1: 1.08.
31, Erylus rotundus var. cidaris; magnified 1: 1.13;
32, Erylus rotundus var. megarhabda from South Molokai (form A); magnified 1: 1.08.

SPONGES OF THE PACIFIC, I]. ERYLIDAE. tae

~ ~ ee

ee
:

A
bo
tho
—
bo
na)

16 17

26
é t
x o
5
z
a ; i
4\
6
f
x
~
rs
> it |
27
30
'
f 2
a)
x ij j
baa \ // y
x \f ay
fi IV
Pi /
[' /|
/ a al
/ AH |
; Ze if
\ AT \VZ%,,}
x
y
i
hy
OD
yf
c
rt
Y,
25 4
2

f

1, 3, 11-13 E. r. var. typica, form D; 2, 4,17 4 ar. typice
16, 30 E. r. var. typica, form A; 18-22, 32 E. r.

=)
is)
| =
z

PLATE 6.

Erylus caliculatus LENDENFELD.

Figures 1-13, 19-23, 26, 28, 29.

Erylus rotundus LeNDENFELD.

Figs. 14, 18, 24, 33-35. — var. megarhabda LENDENFELD. From A. South Molokai.
Figs. 15-17, 25, 30-32. — var. typica LENDENFELD. Form B. Kauai.
Fig. 27. — var. typica LeNDENFELD. Form A. South Molokai.

-

— Microrhabds of Erylus caliculatus: magnified 600; u. v. phot. Zeiss, gq. monochr. 6, q. oc. 10.
3.— Small oxysphaeraster of Lrylus caliculatus; magnified 600; u. v. phot. Zeiss, q. monochr. 6, q. oc. 10.
4, 5.— Groups of microscleres from a centrifuge-spicule preparation of Erylus caliculatus; magnified 600;
. u. v. phot. Zeiss, q. monochr. 6, q. oc. 10:
a, microrhabds; b, acanthtylasters.
6-13.— Acanthtylasters of Erylus caliculatus; magnified 600; u. v. phot. Zeiss, q. monochr. 6, q. oc. 10.
14.— Group of microseleres from a centrifuge-spicule preparation of Hrylus rotundus var. megarhabda —
from South Molokai (form A); magnified 600; u. v. phot. Zeiss, q. monochr. 6, q. oc. 10:
b, acanthtylasters; ec, small oxysphaeraster.
15, 16.— Acanthtylasters of a dark specimen of Erylus rotundus var. typica from Kauai (form B);
magnified 600; u. v. phot. Zeiss, q. monochr. 6, q. oc. 10.
17-22.— Aspidasters; magnified 600; u. v. phot. Zeiss, gq. monochr. 6, q. oc. 10:
17, nearly adult aspidaster of a dark specimen of Erylus rotundus var. typica from Kauai (form B);
18, adult aspidaster of a specimen of Erylus rotundus var. megarhabda from South Molokai Kierra A);
19, adult, circular aspidaster with numerous protruding rays of Erylus caliculatus;
20, young, circular aspidaster with few protruding rays of Lrylus caliculatus;
21, adult, reniform aspidaster with numerous protruding rays of Erylus caliculatus;
22, young, smooth, reniform aspidaster of Hrylus caliculatus. meee
23.— Superficial, paratangential section of Erylus caliculatus; magnified 30; phot. Zeiss, planar 20 mm.
24.— Radial section through the superficial part of Zrylus rotundus var. megarhabda from South Molokai
(form A); magnified 10; phot. Zeiss, planar 50 mm.
25.— Radial seetion through the choanosome of a dark specimen of Erylus rotundus var. typica from
Kauai (form B); magnified 20; phot. Zeiss, planar 20 mm.:
a, radial bundle of rhabd megascleres. ,
26.— Radial section through the superficial part of Hrylus caliculatus; magnified 30; phot. Zeiss, —
planar 20 mm.: a
a, radial bundle of rhabd megascleres. :
27.— Paratangential section, a small distance below the surface, of Erylus rotundus var. typica from —
South Molokai (form A); magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6: )
a, group of triaene-cladomes.

28, 29.— Radial sections through the choanosome of Erylus caticulatus; aniline-blue: ‘=
28, magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6;
29, magnified 100; phot. Zeiss, apochr. 16, compens. oc. 6. ‘
30-35.— Microrhabds of Erylus rotundus; magnified 600; u. v. phot. Zeiss, q. monochr. 6, q. oc. 10:

30-82, of a dark specimen of var. typica from Kauai (form B); : 7
33-35, of a specimen of var. megarhabda from South Molokai (form A).

SPONGES OF THE PACIFIC, II. ERYLIDAE.

Fig. 1-13, 19-23

(0-35 Erylus rotundus n.. sp.

T7, 18, 24, 33-35 E. r. var. megarhabda, form A; 15-17, 25, 30-32 E. r. var. typica, form B

a4

E.

FoD

PLATE 6.

(OT ORT ne el

var. typica, form A.

PLATE 7.

PLATE 7.

Erylus rotundus LENDENFELD.

, 42-45, 75, 76, 79. — var. cidaris LENDENFED.

6, 48, 54. — var. typica LENDENFELD. Form B. Darkspecimen. Kauai.
8, 49-51, 53. — var. typica LeENDENFELD. Form A. South Molokai.

21, 55, 56. — var. typica LENDENFELD. FormC. Light specimen. Kauai.
24, 26, 27, 30; 57-59, 66-73.— var. megarhabda LENDENFELD. Form A. South Molokai.
28, 29, 60-65. — var. megarhabda LENDENFELD. Form B. Kauai.

— var. typica LENDENFELD. Form D. Northeast Hawaii.

Erylus caliculatus LenpeNFELD.
Figures 11-15, 31-41, 74, 77, 78, 80.

1-10.— Branched amphiox-derivates and oxyastrose rhabd-clusters of Hrylus rotundus var. cidaris;
magnified 200; phot. Zeiss, apochr. 8, compens. oc. 6:
1-3, long amphioxes with two or three shorter branches;
4-10, oxyastrose rhabd-clusters.
W15.— Oxysphaerasters of Erylus caliculatus; magnified 2000; u. v. phot. Zeiss, q. monochr. 1.7, q.
oc. 10:
11, an oxysphaeraster with thicker rays, focused higher;
12, the same, focused lower;
13, an oxysphaeraster with thinner rays, focused high;
14, the same, focused intermediate;
15, the same, focused low.
16-21.— Aspidasters of Erylus rotundus var. typica; magnified 300; phot. Zeiss, apochr. 4, compens.
oc. 6:
16, of a dark specimen from Kauai (form B);
17, 18, of a specimen from South Molokai (form A);
19-21, of a light specimen from Kauai (form C).
22-30.— Aspidasters of Erylus rotundus var. megarhabda; magnified 300; phot. Zeiss, apochr. 4,
compens. oc. 6:

24, 26, 27, 30, adult aspidasters of a specimen from South Molokai (form A);
25, 28, 29, adult aspidasters of a specimen from Kauai (form B).
31-41.— Aspidasters of Erylus caliculatus; magnified 300; phot. Zeiss, apochr. 4, compens. oc. 6:
31, a very young, circular aspidaster showing the fine radial rays forming the disc;
32-35, young, circular aspidasters with few protruding rays;
36, 37, adult, circular aspidasters with numerous protruding rays;
38—40, young, reniform aspidasters with few protruding rays;
41, adult, reniform aspidaster with numerous protruding rays.
42—45.— Aspidasters of Hrylus rotundus var. cidaris; magnified 300; phot. Zeiss, apochr. 4, compens.
oc. 6:

46-80.— Microseleres and groups of such from centrifuge-spicule preparations; magnified 300; phot.
Zeiss, apochr. 4, compens. oc. 6:
46, 48, 54, of a dark specimen of Erylus rotundus var. typica trom Kauai (form B);
47, 52, of Erylus rotundus var. typica from northeast Hawaii (form D);
49-51, 53, of Erylus rotundus var. typica from South Molokai (form A);
55, 56, of a light specimen of Erylus rotundus var. typica from Kauai (form C);
57-59, 66-73, of Erylus rotundus var. megarhabda from South Molokai (form A);
60-65, of Erylus rotundus var. megarhabda from Kauai (form B);
74, 77, 78, 80, of Prylus caliculatus;
75, 76, 79, of Erylus rotundus var. cidaris;
a, (Pigs. 46-51, 53-57, 60-75, 77-80) microrhabds; b, (Figs. 52-54, 56-60, 75-77) acanthtylas-
ters; c, (Figs. 52, 60, 74, 76) small oxysphaerasters; d, (Figs. 56, 60) very young aspidasters;
e, (Fig. 56) adult aspidasters.

os

ee

Y

\ , are
66 \67\\68 | 69 \ 70. 71 72 73

37

76,79 Erylus rotundus n. Sp.

var. typica, form Be 17, 1899S: Z,
66-73 E. r. megarhabda, form A;
:megarhabda, form B; 47, 52 E.r. var. typica, form Dd.

S of ae ay we 77 72 20 Beylus calculatus 2. Sp.

Fig. 1-10, 16-30, 42-73, 75,
1-10, 42-45, 75, 76, 79 E.r. var. cidaris; 16, 46, 48, 54 E.r.v

(19-21, 55, 56 E. r. var. typica, form G= 22-24, 26, 27,30, 57-59,

E. r. var. typica, form A;
5, 28, 29, 60-65 E. r. var.

Set
2

—,

mics]
| g
S|
Ay

PLATE 8.

Erylus caliculatus LeNDENFELD.

Figures 1-12, 15-20.

Erylus rotundus LENDENFELD.

. Fig. 13.— var. megarhabda LeNpENFELD. Form A. South Molokai.
Fig. 14.— var. cidaris LENDENFELD. -

1.— Central part of a young aspidaster of Lrylus caliculatus, showing the rays which form the aise !
nified 2000; u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10. a"
2-12.— Acanthtylasters and parts (rays) of such of Erylus caliculatus; magnified 2000; u. v.
Zeiss, q. monochr. 1.7, q. oc. 10: .
2, a ray of a very large acanthtylaster; —i7
3, a small, rather regular acanthtylaster, focused high; 4, the same, focused low;
5, a large regular acanthtylaster, focused high; 6, the same, focused low;
7, a small acanthtylaster, some of the rays of which are considerably shortened, focused hi
the same, focused low; .
9, 10, single rays of ual acanthtylasters; ==
11, part of a small, particularly thin-rayed acanthtylaster; ;
12, part of a small, particularly thick-rayed acanthtylaster.
13-17.— Marginal parts of aspidasters; magnified 2000; u. v. phot. Zeiss, q. monoehr. 1.7, q.
13, of an adult aspidaster of Erylus rotundus var. cyanea (form A) from South Molokai;
14, of an adult aspidaster of Hrylus rotundus var. cidaris;
15a, of a young, still quite smooth aspidaster of Erylus caliculatus; -
15b, 16 of young aspidasters with only few protruding rays, of Brylus caliculatus;
17, of an adult aspidaster with numerous protruding rays, of Zrylus caliculatus.
18-20.— Portions of the surface of the central parts of aspidasters of Erylus caliculatus; magnific
u. v. phot. Zeiss, q. monochr. 1.7, q. oc. 10:
18, of an aspidaster not quite fully developed;
19, of a fully developed aspidaster, focused high;
20, the same, focused low.

_ SPONGES OF ‘THE PACIFIC, II. ERYLIDAE.
} i PLATE 8.

es et Ee TEAS. 8S

Fig. 1-12, 15-20 Erylus caliculatus n. Sp.
Fig. 13, 14 Erylus rotundus n. Sp.
13 E. r. var. megarhabda, form A; 14 E. r. var. cidaris.

BINDING SECT. JUN 6 Ot

QL Harvard University. Museum
ul of Comparative Zoology

H35 Memoirs

veAl

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

STORAGE

senteesse reeset te
eeesi pert iete
ery

y
et
ears
es

eoee
peers

erhiicoate

ras
sue

ry
ors
es

~
ye
erry

ry
Mscistriat ach
ler
eethaytad

seated

pear i
aL ee tare! etusearsre pent ite

3

fteeet
ane 3
Ryritngese a:

cats

at
tt

at
a

3

saci
ini

ahs
eetona ers