K

*;i

fMi

m \.

;':Jf.«(''i

■iiiu;

);(/;

f\etou.nd /9V/

Alkx. Agassiz.

^ibntru of tijc Piuseum

OF

COMPARATIVE ZOOLOGY,

AT HARVARD COllECE, CAMBRIDGE, MASS.
JFouuBEti by pnliatt substiiption, in 18(51.

Deposited by ALEX. AGASSIZ.

No. ^^^^^

THE

EAY SOCIETY

INSTITUTED MDCCCXLIV.

^U

This Volume is issued to the Subscribers to the Ray Society /or the Year 1869.

LONDON

MDCCCLXXI.

A MONOGEAPH

OF THE

GYMNOBLASTIC OR TUBULARIAN

HYDROIDS.

IN TWO PARTS.

T.-THE HYDROIDA IN GENERAL.
II.-THE GENERA AND SPECIES OF THE GYMNOBLASTEA.

GEORGE JAMES ALLMAN, M.D.

DUBLIN AND OXON.,

KELLOW OF THE EOYAL COLLEGE OP SURGEONS IN IKELAND, F.E.S., F.ll.S.E., M.K.I. A., ETC.; EJIEKITUS
PROFESSOR OP NATURAL HISTORY IN THE UNIVERSITY OF EDINBURGH.

LONDON:

PUBLISHED FOR THE RAY SOCIETY BY

ROBERT HARDWICKE, 192, PICCADILLY.

imcccLxxi.

MCZ LIBRARY
HARVARD UNIVERSITY
CAMBRIDGE. MA USA

Freuet euch des wahken Scheins,
EucH des eensten Spieles ;
Kein Lebenpiges rsT ein Eins,
Immer ist's ein Vieles.

Goethe, " Ejnrrhema," in ' Gott und Welt.

VEIKTED BY J. E.
BARTHOI.OUEW CJ.C

kVxcxk i\b §oolv

GEOEGE BUSK, ESQ., E.E.S., E.L.S., &c.,

WnOSE ABLE AND CONSCIENTIOUS LABOURS IN THE FIELD OP IJESEAKCH TO WHICH ITS PAGES ABE

DEVOTED HAVE LARGELY CONTRIBUTED TO OUR KNOWLEDGE OP HYDROID ZOOLOGY,

WHOSE VARIED ACQUISITIONS, GATHERED PROM MANY' A REGION OP BIOLOGICAL SCIENCE,

HAVE EVER BEEN, WITH GENEROUS DISINTERESTEDNESS,

PLACED AT THE DISPOSAL OP ALL AYHO MAY BE WILLING TO USE THEM FOR

THE ADVANCEMENT OF KNOWLEDGE,

ONE OF MY EARLIEST FRIENDS AND MOST VALUED ASSOCIATES IN SCIENTIFIC WORK.

1871.

P E E F A C E,

The present work contains the result of many years' study of the remarkable
group of animals to whose elucidation it is devoted, a group in the investigation of
which ready access to the sea has afforded me special facilities.

My object has been to work out as exhaustively as possible the general natural
history of the Htdroida, and besides this to give a complete descriptive Zoology of the
Gymnoblastic or Tubularian forms of this Order.

The work is thus divided into two parts — the first devoted to the Morphology,
Physiology, Distribution, and other general considerations bearing on the entire Order
of the Htduoida ; the second, to descriptions of all the known genera and species
which compose one of its most important and interesting Sub-orders — that of the
Gymnohlastea.

A very large proportion of the observations here recorded are entirely original,
while it has, moreover, been my aim, in giving an account of the observations of
others, to take nothing for granted which it was possible for me to subject to personal
verification. It will be seen that the amount of labour thus involved is far from
slight. Indeed, it is only by constant and widely extended explorations of the coast,
both within the tidal zone and in the deeper sea regions, followed up by laborious
microscopic investigations, that results of any value are to be expected.

The plates have all been drawn from nature by myself, and are from the living
animal. The soft parts, which constitute the chief interest in these wonderful organ-
isms, are thus represented as they show themselves while the animal is still beneath the
waters of its native seas. This is all the more important in animals which, like the
gymnoblastic hydroids, retain in their dried state not a single character of value, and
which even in specimens preserved in spirits lose almost all their beauty and many of
their important zoological characters. The figures of the species, too, are all coloured
from life, so that not only will a more adequate idea of the beauty of these creatures

viii PREFACE.

in their living state be thus conveyed, but greater facilities will be afforded to the
practical zoologist in the comparison and determination of species. The plates,
moreover, contain numerous anatomical and embryological details ; and, besides the
magnified drawings of each species, I have in every case given a figure of the
animal in its natural size.

It was originally my intention to restrict the descriptive portion of the work
to the British representatives of the group. Further consideration, however, has led
me to believe that its value would be much increased by including descriptions of all
the known Gi/mnoUastea, whether British or foreign. The plates, however, are
necessarily confined to British species. Indeed, independently of other reasons, this
course was inevitable so long as I had resolved to make all my drawings from living
specimens. Full reference, however, is always given to the places where published
figures of the foreign species are to be found.

The same reason has obliged me to leave a few British species unfigured, as
I have hitherto failed in my attempts to obtain living specimens of them. Eeferences,
however, are here, as in the case of foreign species, always made to the works in
which figures of them are given.

Besides the plates, numerous woodcuts are introduced into the text. Though a
few of these have already appeared in my published memoirs, they are all from original
drawings of my own, and will, it is hoped, serve to render clear various points of
structure which it would be difficult to make intelhgible without the aid of figures.'

For obvious reasons it is only those species whose trophosomes have been dis-
covered which form the subject of the descriptive portion of the present work.
There are still known to zoologists a large number of hydroid medusae which have
not yet been traced to their trophosomes. Since Forbes's Monograph, published
among the earlier volumes of the Bay Society, much additional matter has been
accumulated regarding these beautiful organisms, and many of them have been
figured with structural details in the first part of the present work. I have still
many unpublished notes on them, and, though it was impossible to treat them here
systematically, I cannot dismiss the hope of being yet able to supplement the present
volume by another which would be devoted to the natural history of these free
hydroid medusae, whether they have been traced to their trophosomes or not.

As the descriptive portion of this Monograph is based upon the entire organism,
both trophosome and gonosome affording characters equally essential in the diagnosis,
I have never been contented with specimens in which the gonosome as well as the
trophosome was not present. It is only in one or two cases that I have failed in

^ The use of the blocks employed in the illustiation of my " Report on the Hydroida," pub-
lished in the ' Transactions of the British Association for the Advancement of Science/ has been
liberally granted by the Council of that body.

PREFACE. ix

procuring examples provided with their gonosomes, and that I have been obliged
to confine my figures to the trophosome alone.

The additions which the last few years have made to our knowledge of hydroid
morphology liave necessitated the introduction of new terms. Such terms as I have
found it necessary to construct have been made as far as possible etpuologically
significant of the ideas intended to be expressed by them, while I have endeavoured
to define them with a rigidity which may allow of no ambiguity in their application.
The advantages to be derived from a significant and rigidly defined terminology
are great, for it not only facilitates the recording and communication of scientific
truths, but it even becomes, like the symbols in algebra, a direct aid in original research.

With the view of making the terminology as perfect as possible, I have not
hesitated to alter some of the terms formerly introduced by myself. Terminology
differs from nomenclature in priority of use not necessarily giving a fixity of tenure ;
and while capricious change of terms must be deprecated, no one ought to be
precluded from substituting a better term for one already in use.

The labour of the drawings, which I could entrust to no hand but my own, and
the necessity of procuring in every case living specimens as the subjects of them,
have caused the work to be longer in preparation than I had originally anticipated,
and I cannot avoid here expressing my obligations to the Council of the Ray Society
for the patience with which they have borne the delay. One advantage, however, has
followed from it, for I have been thereby enabled to carry up to the present stand-
point of our knowledge this exposition of a rapidly developing department of research,
in which every year has been bringing out new facts and more or less modifying old
views."

The coasts of the British Isles have afforded me the chief fields for exploration,
and my dredgings and tidal coast work have extended from the south-western extre-
mity of Cornwall to the furthest outliers of the Shetland Isles. Some investigations,
however, have been also carried on in the Mediterranean, and I have thus obtained
many facts in hydroid zoology from the northern shores of the Adriatic, from the coast
of Naples, and from the eastern and western Riviera.

Continental museums, wherever accessible, have been consulted. These, on the
whole, are very poor in all that concerns the zoology of the Htdroida, and few of them
possess anything beyond some dried specimens of such common species as may be
casually picked up on the sea-beach.

Some, however, have repaid the trouble of consultation, and I must here express
my thanks to M. Milne-Edwards and to M. Lacase Duthiers for the liberal manner in

' Quite recent additions to our knowledge of hydroid life render necessary some modification of
the statements contained in pp. 22, 23 regarding our want of evidence of the direct development of the
medusa from the egg, without the intervention of a hydriform trophosome. The reader will accordingly
correct and supplement these statements by the results of later observations detailed in p. 100.

b

X PREFACE.

which they placed the collections of the Jardin des Plantes at my disposal, and for the
opportunity thus afforded me of critically examining the authentic specimens of
Lamarck, as well as other interesting hydroid collections in the museum. To
Professor Stossich, of Triest, I am also indebted for an opportunity of examining the
collection of hydroids in the museum of that town, one of the best collections of these
animals contained in any Continental museum which I have been able to consult.

To Professor Paolo Panceri, of the University of Naples, my thanks are especially
due, not only for the liberal way in which he placed in my hands specimens for
investigation, but for the valuable assistance I received from him in the examination
of the Neapolitan coast.

To my friend Professor Schiff, of Florence, I owe my hearty acknowledgments
for aid in consulting the museum of that city, and for many other ways in which he
has facilitated my researches.

To the Marquis Giacomo Doria, who, in the disinterested love of science, has
devoted his time and property to the advancement of natural history, pursued, at the
sacrifice of health, amid the malaria of East Indian jungles, and has thus added
another laurel to those which have already made the name of Doria illustrious in the
annals of the great Genoese Republic, as well as to Dr. Gestro, his assistant, and to
Professor Trinchesi, of the University of Genoa, I am indebted for much kindness, and
for valuable guidance to the zoological localities of the Gulf.

To Dr. Giglioli, of Florence, I owe the opportunity of inspecting an extensive
collection of drawings in which he records many important observations made on
hydroid planoblasts and other pelagic forms met with during the circumnavigatory
voyage of the " Magenta," which he accompanied as assistant-naturalist.

Dr. Du Plessis, of Nice, who has made the hydroids of the neighbouring coast
a subject of special study, and has been singularly sucessful in keeping them in a
healthy state in his vivarium, kindly acted as my guide to various hydroid localities
with which he had become familiar in the beautiful bay of Villafranca ; while I am also
under much obligation to Professor MecznikofF, of St. Petersburgh, who happened to
be at the same time residing at Villafranca, where he was engaged in researches on the
lower animals of the bay, and where he communicated to me some of the important
results to which he had arrived.

To Professor Van Beneden, of Louvaine, and to the late venerable Professor
Sars, I am indebted for presentations of their many important memoirs, and for the
communication of specimens, while my thanks are also due to Professor Agassiz and
to Mr. Alexander Agassiz, as well as to Professor Kolliker and to Professor Haeckel,
for copies of many valuable memoirs bearing more or less directly on hydroid zoology.
And still further, I must express my obligations to Professor Costa, of the
University of Naples ; to Sig. Pillppo Trois, of Venice ; to Professor Savi, of Pisa ;
to Professor Oscar Schmidt, of Gratz ; to Dr. Antoine Pritsch, of Prague ; to Dr.

PREFACE. xi

Brauer, of Vienna ; and to Dr. Marshal, of Leiden, as well as to many other conti-
nental naturalists, for friendly assistance, either by freely placing at my disposal
specimens which I could not elsewhere have obtained or by otherwise aiding me in
the object I had in view.

To specify here the names of our own countrymen from whom I have received
assistance woidd be to extend this list of obligations to a much greater length than
space will allow. Reference to them in other parts of the work will show that I
have not been unmindful of the aid they have afforded me. I cannot, however,
avoid expressing in this place my obligations to Professor "Wyville Thomson, Dr.
Carpenter, and Mr. J. Gwyn Jeffreys, for having placed in my hands the whole of the
hydroids procured in the deep-sea dredgings of the " Porcupine " expedition ; and to
Mr. Busk, for allowing me the free use of his collection of hydroids obtained from
various parts of the world, and affording facts of much value in the geographical
distribution of the order.

The earlier sheets of the present Monograph had been already printed before the
publication of Mr. Hincks's work on the British Hydroids.^ This will account for
the absence of all allusion to it in the section devoted to the history of those labours
which have contributed to bring our knowledge of the Htdroida to its present stand-
point. And yet the literature of hydroid zoology demands a special reference to this
valuable work. Eminently critical, with the descriptions accurate and lucid, and with
the figures abundant and expressive, it is the most complete systematic work on
the Hydroida hitherto published. The large amount of original observations gives it
a special value, and its fulness of description and illustration renders it indispensable
to every student of the Htdeoida. The delay which has occurred in the publication
of the second part of the present Monograph will enable me to cite unreservedly Mr.
Hincks's work, without which the synonomy and literature of many of the species
here described would be very deficient.

' ' A History of the British Hydroid Zoophytes.' By Thomas Hincks. London, Van Voorst, 1868.

GLOSSOLOGY.

Imaginary generalised Diagrams of Ilydroids to illustrate the terminology.

In order better to distinguish the various parts and to render the respective limits of the trophosome and
gonosome at once apparent, the endoderm and ectoderm of the gonosome are in both figures represented by
simple outline, while in the trophosome the endoderm is throughout distinguished by a heavy line, the ectoderm
by transverse hatching, and the perisarc by a simple wavy line.

I. ir.

I. Imaginary generalised Diagram of a Gt/mnoh!astie Egdroid.—?' ?'J\^'.ji ^y?^'"^'"^"^- 1 Hydropbyton. c, c, c— Somatic cavity.

e,'e,'e.— Ectoderm of hydrophy ton."' I Coenosarc. /,/,/.— Perisarc. j.—Hydranth extended. /.—Hydranth contracted.
h, h. — Hypostonie bearing the mouth at its extremity. k. — Sacciform gonopbore (sporosnc) springing from the hydrocaulus.
f .— Sporosac springing from a blastostyle. In J, f , the spadix is seen to occupy the axis of the sporosac, and round the spadix
are developed the generative elements. I. — Medusiform gonopboi'e (planoblast). A manubrium occupies its axis, and in the

walls of this the generative elements are directly developed. m. — Blastostyle.

II. Imaginary generalised Diagram of a Calyptohastic Bydroid.— The letters o to A indicate the same parts as in I. i, !.— Hydrotheca.

k. — Sporosac springing from a blastostyle, with the generative elements developed round a spadix which occupies the axis of
the sporosac. I. — Planoblast springing from a blastostyle. m, m. — Blastostyle. ». — G

GLOSSOLOGY.

TERMS APPLICABLE TO THE HYDEOID COLONY IN GENERAL.

Hydrosoma {vSpa, hydra, mythological monster; uw/ta, body). The entire hydroid colony.

Ectoderm (kro'c, outside; Sep/ua, skin). The more external of the two organized layers of
which the body of every hydroid is composed. (Diagrams I and II, e, e, e.)

Endoderm {IvBov, within ; Sip/na, skin). The more internal of the two organized layers of
which the body of every hydroid is composed. (Diagrams I and II, d, d, d.)

Perisarc {inpl, around; aap^, flesh). The unorganized chitinous excretion by which the soft
parts are to a greater or less extent invested. (Diagrams I and 11,/,/,/.)

Zooids {Iwov, animal; 'tiSoc, form). The more or less independent products of non-sexual
reproduction ; the members more or less individualized of which the hydroid colony is composed.
(Diagrams I and II, g, (/, k, ¥ , /, m)

Trophosome (rpo^i), nourishment; aw^ia, body). The entire assemblage of zooids with their
common connecting basis, destined for the nutrition of the colony.

Gonosome (yoroc, offspring; <7(J^a, body). The entire assemblage of zooids destined for the
sexual reproduction of the colony.

Thread-cells. Peculiar bodies consisting of a containing capsule and contained filament
destined for urtication, and universally present as a histological element of the ectoderm. (Fig.
52, page 118.)

Palpocils [paljm, I feel; cilium, an eyelash). Microscopic, hair-like, non-vibratile processes
of the ectoderm, probably organs of touch. (Pig. 48, page 112.)

Heteromorphism {inpoq, diverse; /uopffj, form). Diversity of form among the component
zooids of the colony.

Homomorphism {ofwwq, similar; ixQp(^>'u form). Similarity of form among the component
zooids of the colony.

Polymerism (ttoXuc, many ; /ifpoc, part). Simple multiplicity of the component zooids of the
colony.

TERMS APPLICABLE TO THE TROPHOSOME.

Hydranth (u'Spa, hydra; avQoq, flower). The proper nutritive zooid, or that part of it which
carries the mouth and proper digestive cavity. (Diagrams I and II, g, /.)

Hypostome (utto, under; <TTO(ua, mouth). The distal prolongation of the hydranth, which
carries the mouth on its summit. (Diagrams I and II, h.)

Hydrotheca {v^pa, hydra; e»)/c>/, receptacle). The cup-like chitinous receptacle which
protects the hydranth in the calyptoblastic genera. (Diagram II, i, i.)

Hydrophyton (uSpn, hydra; (^vtuv, plant). The common basis of the trophosome by which
its zooids are connected into a single colony. (Diagrams I and II, a, a, a, a, b, h.)

GLOSSOLOGY. xv

Hydrorhiza {iSpa, hydra; f,tta, root). The proximal end of the hydrophyton by which the
colony fixes itself to other bodies. (Diagrams I and II, 6, b, b, b.)

Hydrocaulus (uSpa, hydra; kouXo'c, stem). All that portion of the hydrophyton which
intervenes between the hydrorhiza and the hydranth. (Diagrams I and II, a, a, a, a.)

Coenosarc [kowoq, common ; aap^, flesh). The common organized fleshy portion of the
hydrophyton ; the living bond by which the zooids are organically united to one another.
(Diagrams I and II, d, d, d, e, e, e.)

Nematophores (i-n/ua, thread; fopiut, I carry). Peculiar bodies developed in certain genera
from definite points of the trophosome (and of the corbulse in the genus Aglaophenid), and con-
sisting of a chitinous receptacle with sarcode contents in which thread-cells are usually immersed.
They are characteristic of the family of the Plumularidce. (Figs. 50 and 51, pages 116 and 117.)

TERMS APPLICABLE TO THE GONOSOME.

Gonophore (ydi'oc, offspring; ^opku),\ bear). The ultimate generative zooid which gives
origin directly to the generative elements, ova or spermatozoa. (Diagrams I, l\ k', I, and II, k.)

Sporosac {a-Kopa, sexual product, offspring; auKiiQ, a sack). A sack-shaped gonophore
destitute of obvious umbrella. (Diagrams I, k, k', and II, k.)

Planoblast [wXavoQ, wandering ; jSXoimj, a bud). A generative bud with a structure fitting
it for a free locomotive life when detached from the hydrosome. (Diagrams I and II, I.)

Gonocheme (yoVoe, offspring; oy^nun, chariot). A medusiform planoblast which gives origin
directly to the generative elements. (Diagram I, /.)

Blastocheme (/3Xa<7Tj(, bud ; oy^vf^a, chariot). A medusiform planoblast which gives origin
to the generative elements, not directly, but through the medium of special sexual buds which
are developed from it. (Diagram II, I, and Figs. 9 and 10, page 35.)

Blastostyle (jSXaiTTi), bud; arvXag, column). A columniform zooid destined to give origin to
generative buds. (Diagrams I and II, m, m, in.)

Perigonium [inpi, around; -yococ, offspring). The walls of a sporosac by which the genera-
tive elements are confined, and in which, when fully developed, three laminae may be demonstrated.
(Fig. 7, page 32.)

Ectotheca (f'/cro'c, outside; 0>iK»), sheath). The most external of the three laminae of the
perigonium. (Fig. 7, c, page 32.)

Mesotheca [^liaoq, middle; 0»(k»), sheath). The middle one of the three lamina; of the
perigonium. (Fig. 15b, b, page 44.)

Endotheca (t'l/Soi^, within; Qw, sheath). The most internal of the three laminae of the
perigonium. (Fig. 7, b, page 32.)

Spadix ((TTraSit, the fruit-shoot of a palm tree, a term used by botanists for a form of inflo-
rescence). The hollow body which projects from the floor of the sporosac into its cavity, and round
which the generative elements are developed. (Diagrams I and II, k, k', and Fig. 7, a, page 32.)

Umbrella. The gelatinous bell of a medusiform planoblast. (Diagrams I and II, /, and
Fig. 8, c, page 33.)

xvi GLOSSOLOGY.

Manubrium {maniihrinw , handle). The axial portion which, in a raedusiform planoblast,
hangs from the summit of the umbrella, carrying the mouth at its extremity. (Diagram I, /, and
Fig. 8, page 33.)

Atrium {atrium, a hall). An enlargement of the somatic cavity which occurs in many
medusae. It is situated at the base of the manubrium, and from it the radiating canals proceed.
(Fig. 17, page 46.)

Codonostome (kwIwv, bell ; arofxa, mouth). The orifice of the umbrella through which its
cavity communicates with the external water. (Fig. 8, page 33.)

Velum {velum, a veil). The membranous perforated diaphragm which stretches transversely
across the codonostome. (Fig. 8, h, page 33.)

Ocellus (diminutive of oculus, eye). A heap of pigmentncells accompanied or not by a
refracting body, and forming a coloured spot on definite points of the umbrella-margin in certain
planoblasts. (Fig. 5(3, t;, page 139.)

Lithocyst {X'Soc, stone ; kiIcttic, bladder). A sack-like body containing concretions, developed
on definite points of the umbrella-margin in certain planoblasts. (Figs. 57, c, and 58, //, pages 140
and 141.)

Phanerocodonic {<j>avipoQ, manifest; kuiSujv, bell). The condition of a gonophore when it
possesses a developed umbrella. (Diagram I, /.)

Adelocodonic (aSijXoc, not manifest; kmSwv, bell). The condition of a gonophore when no
developed umbrella is present. (Diagram I, I; Ic , and II, k.)

Gonangium (yo'i-oc, off'spring; a-yytiov, vessel). An external chitinous receptacle within
which, in the calyptoblastic genera, the sporosacs or planoblasts are developed. (Diagram II, «.)

Gubernacidum [ffitbernaculum, rudder, director). A common sack-like membrane which
surrounds the generative buds within the gonangium, and aids in directing them or their contents
towards the orifice of the gonangium. (Figs 18, d, and 19, d, page 48.)

Acrocyst {aKpog, on the top ; Kvan^, bladder). An extenial sac which in certain hydroids is
formed upon the summit of the gonangium, where it constitutes a receptacle in which the ova
pass through some of the earlier stages of their development. (Figs. 21 and 22, page 50.)

Meconidium (diminutive from /miKuv, a poppy). Peculiar sporosacs, somewhat resembling
a poppy capsule in form, and borne upon the summit of the gonangium in the genus GonothyrcBa.
(Fig. 28, page 57.)

Corbulae {corbula, a basket). Basket-shaped receptacles which enclose groups of gonangia
in certain plumularian hydroids. (Fig. 30, page 60.)

Planula (a diminutive noun, suggested by a supposed resemblance to a Planarid). The
locomotive infusorium-like embryo into which the egg of most hydroids becomes directly
developed. (Fig. 39, K, page 86.)

Actinula (a diminutive noun found from okt/c, a ray). The locomotive polypoid embryo into
which, in certain genera, the egg becomes directly developed. (Plate XXI, fig. 6, Plate XXIII,
fig. 16, &c.)

GLOSSOLOGY.

TERMS APPLICABLE TO CERTAIN CONDITIONS OF THE HYDROSOMA —
NAMES OF LEADING SYSTEMATIC GROUPS.

Gymnoblastic (-yujiivoc, naked ; ftXdarri, bud). The condition of a hydroid when no external
protective receptacle (liydrotheca or gonangium) invests either nutritive or generative buds.
Gtmnoblastea, the name of one of the sub-orders of Hydrotda. (Diagram I, and the various
plates illustrating the present Monograph.)

Calyptoblastic ((coAutttoc, covered; ^Xaart,, bud). The condition of a hydroid when an
external protective receptacle (hydrotheca or gonangium) invests either the nutritive or generative
buds. Calyptoblastea, the name of one of the sub-orders of Hydroida. (Diagram II, and
fig. 2, page 23.)

Eleutheroblastic {iXivOipoe, free; /BAauDj, bud). The condition of a hydroid when the
nutritive buds, instead of remaining permanently attached, become free and enjoy an independent
existence. Eleutheroblastea, the name of one of the sub-orders of Hydroida.

MoNOPSEA ((uo'i'oc, single; o'l^ic, appearance). The name of one of the sub-orders of the
Hydroida, in which development from the egg takes place without the intervention of a hydri-
form trophosome.

Rhabdophora {pufiSoQ, rod ; fop^w, I bear). The name of one of the sub-orders of the
Hydroida. It corresponds to the extinct group of the GraptoUtes, in which a soHd rod is
developed in the walls of the chitinous perisarc.

CONTENTS.

PAGE

PREFACE ......•• V

GLOSSOLOGY . . . . • • xi

Terms applicable to tlie Hydroid Colony in general . . . xii

Terms applicable to the Tropliosome . . • • . xi

Terms applicable to the Gonosome . . ■ • • xm
Terms applicable to certain Conditions of the Hydrosome. — Names of Leading

Systematic Groups ....•• xvii

PART I.

THE HYDROIDA IN GENERAL.

PHYSIOGNOMY OP THE HYDROIDA.— DESIGN OF THE MONOGRAPH . 1
SYSTEMATIC POSITION OF THE HYDROIDA . • -2

HISTORY OF THE PROGRESS OF OUR KNOWLEDGE OF THE HYDROIDA 3
MORPHOLOGY OF THE HYDROIDA.— TERMINOLOGY .

I. The Hydrosoma in General

1 . Generalised Conception of a Hydroid .

2. General Structure.— Ectoderm and Endoderm

3. Composite Character of the Hydroida. — Trophosome and Gonosome

4. Orientation . . . • •

II. Morphology of the Trophosome

1. Hydranths . . . . •

2. Hydrophyton . . . ■ •

Cffinosarc and Perisarc . . . •

Hydrorhiza and Hydrocaulus

21
21

22
24

24

24
26
26

27

CONTENTS.

III. Morphology of the Gonosome ....

1. General View of the Gonosome

2. The Adelocodonic Gonophore (Sporosac)

3. The Phanerocodonic Gonopliore (Gonocheme, Sexual Medusa)

4. The Blastostyle .....

5. The Blastochcme (non-Sexual Medusa)

6. Honiological Parallelism between Sporosac and Medusa

7. Honiological Parallelism between Hydranth and Medusa

8. Further Modifications of the Gonosome

9. The Generative Elements ....

Ova .....

Spermatozoa .....

IV. Development

1. Development of the Bud. — Zooidal Development . . .67

a. Development of the Hydranth . . . .68

Hydrantb-bud in the Gymnoblastca and Eleutheroblastea . 68

Spontaneous Decapitation and Re-formation of successive

Hydranths . . . . .69

Polarity of the Hydroid . . . .70

Hydranth-bud in the Calyptoblastea . . .71

b. Development of the Blastostyle . . . .73

c. Development of the Sporosac . . . .74

d. Development of the Medusa . . 76

Formation of Buds by the Medusa . . .82

2. Development of the Ovum. — Embryonal Development . . 85

Development of the Embryo from the commencement of the Segmen-
tation of the Vitellus to the attainment of the Free Locomotive
Stage . . . . . .85

Further Development of the Planula to the attainment of the Adult

Form . . . . .89

Development by Actinulse . . . .90

3. Significance of the Medusa in the Life-series of the Hydroid . 93

Direct Development of the Medusa from the Egg . . 98

4. Relation between Zooidal and Embryonal Multiplication in the

Hydroida ; Polymerism and. Heteromorphism ; Genetic Suc-
cession of Zooids . . . . .101

PAGE

29

32
32
33
35
38
40
43
64
64
64

CONTENTS.

V. Histology

1. The Ectoderm ....

General Structure of the Ectoderm .

Palpocils ....

Fibrillated Tissue

Umbrella of Medusa

Nematopliores ....

Thread-cells . . . •

2. The Endoderm ....

Cellular Structure of Endoderm. — Contents of the Cells
Canaliculation of Eudoderra
Endoderm of Tentacles

PAGE

110

110
110

111

112
113
115
116

122
122
124
126

PHYSIOLOGY OF THE HYDROIDA

1. Digestion ....

2. Circulation, Nutrition, and Respiration

3. Secretion . . • ■

4. ContractOity . . . •

.5. Sensation . . . •

The Ocellus
The Lithocyst
Touch ....

6. Phosphorescence

7. Reproduction

a. Sexual Reproduction. — Generation

Comparison of the Sexes in the Hydroida
Origin of the Generative Elements .

h. Non-Sexual Reproduction .
Gemmation
Fission

128

128

130

139

136

137
138
139

144

145
146

147
147
148

150
150
151

CONTENTS.

DISTRIBUTION

1. Distribution in Space ....

a. Horizontal Distribution .
h. Vertical Distribution

General Considerations
Bathimetrical Zones
The Surface Zone
The Littoral Zone
The Laminarian Zone
The Coralline Zone
The Deep-water Zone
The Abyssal Zone
Deep Lacustrine Fauna

2. Distribution in Time ....

Fossils referred to the Hydroida on insufficient Evidence

Fossil Ilydroid Trophosomes .

Fossil liydroid Medusse

Graptolites ....

Table of the Distribution of the Hydroida in Time

PAGE

155

155
155
162
162
166
166
167
168
168
169
169
170
170
170
172
174
175
183

SUPPLEMENTARY.
CLASSIFICATION ....
HOMOLOGIES OF THE CCELENTERATA

TERATOLOGY AND PATHOLOGY

1. Modifications resulting from Parasitism

2. Other Abnonnal Conditions

184
189

194

194
196

CONTENTS.

ANATOMY OF SPECIAL FORMS

TLtl)ularia iiidivisa
Coryiiiorpha nutans
Clavatella prolifera
Cladonemu ladiatum
Ilydractinia echinata
Gemmaria implexa
Dicoryne conferta

PAGi:
205

205
-'05
•212
210
220
223
220

SUPPLEMENTARY NOTE.
STRUCTURE OF CORDYLOPHORA
THE GONANGIUM ....
FUNCTION OF THREAD-CELLS
EVOLUTION OF THE HYDROIDA

228
229
220
230

PART II.

THE GENEEA AND SPECIES OF THE GYMNOBLASTEA.
GENERAL PRINCIPLES OF HYDROID ZOOGRAPHY
SYNOPSIS OF FAMILIES AND GENERA

235
239

DESCRIPTIONS OF THE FAMILIES, GENERA,
CLAVID^ .....
TURRID.E ....

CORYNID^ ....

SYNCORYNIDiE ....

AND SPECIES.

242
259
264
274

CONTENTS.

DICORYNID^

BIMERID^

BOUGAINVILLIDiE

EUDENDRIDJ]:

HYDRACTINIDiE .

PODOCORVNID.E

CLADONEMIDiE

NEMOPSID^

PENNARID^

CLADOCORYNIDiE

MYRIOTHELID^

CLAVATELLlDiF:

CORYMORPHID^

ISIONOCAULIDtE

TUBULARIDiE

HYBOCODONID^

HYDROLARIDiE

I'AGE

2!) 2

. 294

. 31(1

. 330

. 342

. 34 S

. 35G

. 360

. 363

. 37!)

. 381

. 383

. 386

. 395

. 398

. 421

. 424

BIBLIOGRAPHY . . . . ... 429

INDEX TO THE FAMILIES, GENERA, AND SPECIES 01" GY.MNOBLASTEA 435
LIST OF WOODCUTS . . . . . .439

GENERAL INDEX . . . . . . .441

PART I.

THE

HYDEOIDA IN GENERAL.

THE

HYDROIDA IN GENERAL

PHYSIOGNOMY OF THE HYDROIDA.— DESIGN OF THE
PRESENT MONOGRAPH.

Rooted in the transparent reservoirs wliicli the retiring tide has left behind it in the rocky
shore, or spreading over the fronds of the sea-weeds, or fringing the reef at low water with a
mimic vegetation, or brought up by the dredge of the naturalist and the lines of the fisherman
from the deeper regions of the sea, there may be obtained, on perhaps every coast and in every
latitude, certain singular organisms which repeat with such unerring fidelity the forms of the
vegetable kingdom that we can scarcely bring ourselves to believe that the hundred plant-like
shapes which root themselves in that marvellous sea-garden, and stretch forth their branches, and
unfold their buds, and spread abroad their starry flow-ers, have not the structure and the life as
well as the form and the habit of the plant. And yet they are no plants, these strange plant-like
dwellers in the sea, but genuine animals in all that constitutes the essence of animality.

When Marsigli, more than a century and a half ago, fished up from the Mediterranean Sea
a piece of living coral, and for the first time in the history of science its branches were seen
clustered with starry polypes, he believed that he had before him a blossoming plant ; for in the
branching stem which he had plucked from the rock M-here it had been rooted, and in its living
bark and eight-petalled flowers he saw nothing but the evidence of vegetality, which surely proved
that the great botanists of the day — Ray and Tournfort, and Ca3salpinus, and Bauhin, and Lobel,
were right when they called corals plants, and assigned them to the Floivi rather than to the Fauna
of the sea. And so, also, the organisms with w^hich the present monograph is to be occupied are
no less plant-like than their relatives the corals, for they are rooted, and branch, and bud and
blossom like them.

But more than this, the sea is filled with living and moving forms, floating bells of crystal,
whose beauty no description can convey, whose multitude no thought can estimate. Unlike those

1

2 THE HYDROIDA IN GENERAL.

animated flowers wliich root themselves to the sea-bed, these no less wonderful Medusa^, with
functions higher and more varied, lead a life of freedom. They love the upper regions of the
sea, and wherever over its wide surface the conditions suited to their welfare are to be found
there will the towing-net encounter them. A thousand leagues away from land, where the shij)
lies motionless in the calm, there they are abroad in their unnumbered hosts ; and where the
gale is strong, and the wave breaks upon the rocky headland, there too they congregate and
sport imharmed in the surf. And yet for days together the towing-net may sweep the sea
without a trace of them, for they are sensitive to every changing mood of the atmosphere above
them ; they feel the gathering cloud and the summer shower; and when the sea freshens beneath
the falling rain-drops, or the air rests upon its surface with influences unfavorable to their well-
being, they sink into salter waters and find shelter in more genial depths.

But their life was not always one of freedom as it now is, for they once grew as buds upon
those strange hydroids, which, with the life of the animal, root themselves to the sea-bed like a
plant; they sprung forth from their sides, and drew their nourishment from the parent branch,
and expanded and developed themselves until they became fitted for an independent existence,
and then, full of a new and higher life, they broke away from their supporting stalk, active and
energetic beings, unrivalled in the gracefulness of their motions and in the symmetry and beauty
of their forms.

The true significance of all this budding and blossoming, of this imitation by the animal
of the form and growth of the plant, lies at the foundation of a scientific knowledge of the
Hydroida, and constitutes one of the most interesting and marvellous chapters in the
morphology and physiology of animals.

It is my intention to devote the present work to an examination of the Hydroida in their
general morphological and physiological relations as a great natural group ; while to one large
and important subdivision of this group, the Tubulari/ice, a more special consideration will be
given, and all the genera and species of which it is composed will be described in detail.
Thus, a purely descriptive zoology of the Tubulariiice will be combined with a careful study of
their structure and physiology, and of the structure and physiology of the entire order of the
Hydroida, that more comprehensive group under which the T/ibularince are immediately
included.

When thus investigated, it will be found that the study of the Hydroida possesses an
interest far beyond what we may at first be inclined to attribute to beings so simple in their
structure and so apparently insignificant in the place allotted to them in the economy of nature,
for we shall then learn that some of the most important facts in morphology and some of the
highest laws in physiology find in them their expression and elucidation.

SYSTEMATIC POSITION.

The Hydroida of the present monograph include the Hi/drin(P, TiibidarinfB, CampanularincB,
and Sertidari/ia, being so far exactly coextensive with the Hydroida of Johnston.' The group
Hydroida, however, as here understood, necessarily embraces most of the so-called naked-eyed
or gymnophthalinic Medusae, for a large proportion of these are known to be the free generative

' George Johnston, ' A History of the British Zoophytes.' Second Edition, 1847.

HISTORY OK RESEARCH. 3

zooiils of the TubulariiHc and CaMpaiiulariiuc, wliile those which have not yet been so traced —
provided we have no reason to regard them as the free zooids of the S'iphonophora — and even
those which may be proved to be developed directly from the egg, cannot, in a philosophical
system, be separated from the others.

1 accept without hesitation the group CcELENTFaiAT.\, witii the characters assigned to it by
Leuckart; and I further adopt the division of this group into two primary sections, with the
names of Adbwzoa and Hydrozoa, as projjosed by Huxley. The following table will indicate
the place of the Hydroida among the other members of the llijdrozoa .'

ACTINOZOA.

CCELENTERATA

Cteuopliora."
Discopliora.
Hydhozoa . . / LuceniariBe.

Hydroida.
y Siplionophora.

HISTORY OF THE PROGRESS OF OUR KNOWLEDGE OF THE HYDROIDA.

The history of the successive stages through which any important branch of human
knowledge passes in its development from the first dawnings of its truths upon the mind to that
more perfect phase which in the lapse of time it has attained, constitutes one of the most instruc-
tive subjects upon which the philosophic student can be engaged ; and a history of this develop-
ment, as it shows itself in the j)i-ogress of our knowledge of the Hydroida may, therefore, with
advantage precede that exposition of the present state of oiu- knowledge of them which is the
chief aim of the present work.

In order to avoid extending our historical sketch to an inconvenient length, the record of many
unportant anatomical and physiological discoveries must be postponed to that part of the volume
where these discoveries can be described with sufficient detail ; and I shall here confine myself
chiefly to the more important steps which have been made towards the determination of the
systematic position of the Hydroida, and their recognition as a distinct group with the limits
assigned to them in this monograph.

' I must for the present hesitate to include among the liydrozoal orders the tabulate and rugose
corals. The hydrozoal affinities of these groups have been recently claimed for them by Agassiz as the
result of an examination of living specimens of MiUepora alcicornis, a tabulate coral, in which, if there
be no error of observation, Agassiz has detected a true hydrozoal structure, while he believes himself
supported by analogy in attributing this structure, not only to all the other genera of tabulate corals,
whether living or extinct, but even to the entirely extinct group of Rugosa. (See his ' Cont. Nat.
Hist. United States,' vol. iv.) The observations, however, on which this view has been based are
plainly not yet as complete as could be desired for a determination so important, and even startling.
Of the generative system more especially we are entirely ignorant. Under these circumstances I
believe it will be safer to wait for such verification as may be expected from further researches.

' In adopting the more usual view, in accordance with which the Ctenophora are placed among
the Hydrozoa rather than among the Actinozoa, as originally, indicated by Leuckart, and more deci-
dedly insisted on by Huxley, I believe myself borne out by a careful study of the structure of Beroe.

4 THE HYDROIDA IN GENERAL.

We have no evidence whatever to show that the Greek and Roman naturahsts were
acquainted with any member of the Hydroida. Aristotle and the naturahsts of Greece and
Rome who followed him had some knowledge of corals, sea-anemones, and steganophthalmic
Medusa; ; but this was very imperfect, while no mention is made by them of a single hydroid, and
it is not until the eighteenth century that we find in the writings of naturalists anything beyond the
most obscure indications of an acquaintance with the animals now included in the order Hydroida.

It was in the beginning of the eighteenth century that the fresh-water Hydra was discovered
by Leeuwenhoeck, and its faculty of buddhig like a plant accurately described. Leeuwenhoeck
communicated a notice of this discovery to the Royal Society of London in 1803.'

The first grand impulse, however, to the study of the Hydroida was given some years later
by Trembley. Abraham Trembley was born in Geneva in 1700, and in 1743 was awarded the
Copley medal by the Royal Society of London, of which he had been elected a Fellow. It was
while residing at the Hague with his two pupils, the sons of the Count de Bentinck, that he
obtained, in the pond at Sorgvliet, the coimtry house of the count, the hydras which enabled
him to make that remarkable series of observations on the reproductive powers of these animals
which resulted in the discovery of jjhenomena hitherto unsuspected in the animal kingdom,
and of the highest significance in physiology ; for they established the fact that the animal
organism may not only midtiply itself by budding in the manner of a plant, as Leeuwenhoeck
had already demonstrated, but that it may possess the power of enduring repeated subdivision,
and may suffer with impunity the most extensive mutilations, the fragments of the divided Hydra
not only recovering from the operation, but becoming endowed, after a time, with all the parts of
which they had been deprived by the act of division.

The discoveries of Trembley were communicated to Reaumur, and recorded by him, in 1742,
in the preface to the sixth volume of his ' History of Insects;'" and in 1744 an extended account
of them was published by Trembley himself, in his celebrated treatise on ' Fresh-water Polypes.''
In this remarkable work the species of Hydra known to Trembley are described with copious
details of their general structure and habits, and of the curious experiments to which he subjected
them. The work consists of four memoirs, and is abundantly illustrated with figures of great
beauty exhibiting the Hydra in various conditions and under various modes of treatment, all
from the pencil and most of them from the graver of the celebrated Lyonet ; while the quaint but
expressive vignettes from another hand, which are placed at the heads of the four memoirs, and
which represent various parts of the grounds of Sorgvliet, with the author and his two pupils
engaged in the capture and observation of the Hydras, give an additional charm to a work which
must be regarded as the most important step yet made towards a scientific knowledge of the
Hydroida.

The progress of discovery in the natural history of the Hydroida, however, is so intimately

connected with various observations which had been about this period made on certain corals

and other Actinozoa, that it is impossible to follow the one without some knowledge of the other.

The researches of Trembley were preceded by Peysonelle's demonstration of the true nature

of the polypes of coral. The coral polypes were discovered towards the beginning of the last

^ Ant. de Leeuwenhoeck in ' Phil. Trans.' for 180.3.

- Rene-Antoine Ferchaud de Reaumur, ' Histoire des Insectes.' Paris, 1743.
' 'Memoires pour servir il I'Histoire d'un genre de Polypes d'eau douce il bas en forme de
Comes.' Leyden, 1744.

HISTORY OF RESEARCH. 5

century by the Count de Marsigli/ Marsigli, however, regarded them as the flowers of the coral,
and saw in thera a proof of the vegetable nature of the supposed sca-pUuit which bore theni ;
and his discovery was at once received as a full confirmation of the views entertained by the
leading botanists of the time, who all regarded the corals as genuine members of the vegetable
kingdom.

Jean Antoine Peysonelle, however, during a residence at Marseilles and on the Mediterra-
nean shores of Africa, and subsequently at Guadaloupe, applied himself to the study of living
corals and madrepores, and soon became convinced that the coral flowers of Marsigli were truly
animals closely allied to the Actuiia, or " TJrticce vtarina," as they were called l)y the naturalists
of that day.

Peysonelle's views were communicated by Reaumur to the Academy of Sciences in 1727,'
where they were received with discredit, and even contempt, Reaumur himself, who Ijelieved in
the vegetable nature of coral, not even mentioning the name of the author whose communication
he undertook to present, so that Peysonelle's discovery remained almost unknown until 1742,
when he forwarded to the Royal Society of London a long memoir, which was published in
abstract in the ' Philosophical Transactions' of that year.

The discoveries of Peysonelle, however, had arrested the attention of the celebrated botanist,
Bernard de Jussieu, and, with the view of verifying them, he determined to visit the sea-coast of
Normandy. Though the shores of Normandy afforded to Jussieu no true corals, he found there
the nearly allied Alcyonium, which enabled him to confirm the views of Peysonelle. He at the
same time convinced himself that the plant-like Flustras wei-e truly animals ; and, what has a
more direct importance in its bearing upon the present history, he observed the polypites of
Tuhularia indivim, and was thus enabled to refer this hydroid to the animal kingdom. The
results of de Jussieu's visit to Normandy were published in the ' Memoires de rAcadumie' for
1742,* where he gives a figure of Tahtdaria ind'msa, which in truthfulness and expression has
never since been surpassed.

Reaumur, unable to resist the accumulated evidence of the animality of corals and hydroids,
now fully accepted the views of Peysonelle, which he had some years before scarcely deemed
worthy of a serious thought.

At this time Linnaeus was carrying out those wonderful reforms in classification and nomen-
clature which were destined to exert an influence on the progress of natural history greater than
anything which had been effected since the days of Aristotle, and which mark out the eighteenth
century as the most significant in the history of the natural sciences.

The discoveries of Peysonelle, of Jussieu, and of Trembley, however, had not yet brought
conviction to the great systematist, and in 1745 we find him, in a dissertation on the fossil
corals of Sweden,^ after contrasting the various opinions regarding the nature of coral in accord-
ance with which it was assigned either to the mineral, the vegetable, or the animal kingdom,
candidly confessing that he was unable to decide between these rival views.

^ See Luigi Ferd. Marsigli, ' Histoire physique de la Mer.' Amsterdam, 1725. Translated,
under the care of Boerhaave, from the original Italian edition of 1711.

" Bernard de Jussieu, " Exameu de quelques productions Marines qui out ete mises au nombre des
plantes, et qui sout I'ouvrage d'une sortc d'Inscctes dc Mer;" 'Mem. de I'Acad. Roy. des Sciences,'
Paris, 1742, p. 392.

* Carolus Linua3us, ' De Coralliis Balticis.' UpsaliK, 1745.

6 THE IIYDROIDA IN GENERAL.

The only hydroid which, up to this time, had been examined in a living state with results
of any value to seience was the Tiibularia indivisa, which, as already mentioned, had been studied
by De Jussieu on the coast of Normandy ; a most important accession, however, to our know-
ledge of the IIydroida was now about to be made by the observations of Ellis,

John Ellis was a London merchant devoted to the study of natural history, which he pur-
sued in the intervals of his mercantile labours, with an enthusiasm and a success which renders
his name famihar to every student of the Hydroida. He was elected a Fellow of the Royal
Society, and was awarded the Copley Medal in recognition of the esteem in which his researches
were held by that body.

An examination of dried specimens of various hydroids had already led Ellis to suspect that
these plant-like productions really belonged to the animal kingdom, and determined him to study
them in a living state. With this view he repaired with his microscope to the Island of
Sheppey, and some other parts of the south-eastern shores of England, accompanied by Mr.
Brooking, a distinguished painter of sea-pieces, and by the celebrated botanical painter Ehret.
He had there abundant opportunity of studying a great number of living hydroids, and soon
convinced himself that " these apparent plants were ramified animals in their proper skins or
cases." In this remarkable assertion we have the first philosophic expression of the true nature
of the fixed plant-like hydroids, and thus was finally settled the animality of these organisms.
The results of his observations were published in 1755, in a work" whose beautiful and accurate
figures and admirable descriptions render it at this day indispensable to the student.

Nothing was now wanting to produce general conviction of the animality, not only of the
true corals, but of all those flexible, plant-like productions whose external form had so long
caused their real nature to be overlooked. Even Linnaeus himself was at last convinced by the
discoveries of Ellis, and now declared himself a believer in their genuine animality.

Besides the generally very expressive vernacular names employed by Ellis, his species are,
in accordance with the usual practice of the day, indicated by short Latin descriptions rather than
by systematic designations. Linnasus's grand invention of the binomial nomenclatiu'e was,
however, making its way among systcmatists. The ' Systema Naturas' had already passed
through several editions, and in 1766 Ave find the various species of Hydroida then known
enumerated by Pallas under their binary designations in his admirable ' Elenchus Zoophytorum.'-
In this work the species are characterised by a precision which leaves little to be desired ; a
complete synonomy is prefixed to each, and in their arrangement the celebrated Prussian natu-
ralist affords evidence of an insight into those affinities on which the more natural classifications
of subsequent systematists have been based.

In the tenth fasciculus of his ' Spicilegia Zoologica,'^ published in 1774, Pallas describes and

^ John Ellis, ' Au Essay towards a Natural History of the Corallines and other jMariue pro-
ductions of the like kind commonly found on the coasts of Great Britain and Ireland. To which is
added the description of a large Marine Polype taken iiear the North Pole by the Whale-fishers in the
summer of 1753.' London, 1755.

" Petr. Sim. Pallas, ' Elenchus Zoo[ihytorum sistens generum Adumbratioiies generaliores et
specierum cognitarum succinctas descriptiones enm selectis Auctorum syuonymis.' Hagie-Comit., 1766.

' Petr. Sim. Pallas, ' Spicilegia Zoologica quibus novse imprimis obscures Animalium species
iconibus, descript. atque commentariis illustrantur,' torn, i, fasc. i — x, Berolini, 1767 — 1774; tom. ii,
fasc. xi — xiv, Berolini, 177G — 1780.

HISTORY OF RESEARCH. 7

figures two now hydroids. One oi' tlieiu is a Cori/iie, n genus wliich he adopts from a MS. of
Gaertnei- ; the otlicr lins no gonoiic name assigned to it by Pallas ; it can, however, be easily
recognised as a CIma, a genus founded a few years afterwards by Gmelin for the Hydra
squamnfa of Miiller. Pallas's figures, however, though sufficient for idc^ntification, cannot be
compared, either in beauty of execution or in truthf'uhiess, to those of Trembley, Jussieu, or
Ellis.

A much better figure of a Chwa was given shortly afterwards by the Danish naturalist and
traveller, Forskal, in his ' Icones Rerum Naturalium,'' where the species is named Hydra
multicornis ; and in the same work, besides two other tolerable figures of hydroid trophosomes,
we find some very expressive and, indeed, up to that time, the only really recognisal)le ones of
true hydroid Medusae.

Among the means which tend most powerfully to advance the progress of the natural history
sciences is an accurate and expressive iconography. The beautiful figures of Trembley and of
Ellis hold in this respect the first rank. As we have already seen, Jussieu had given an admirable
figure of Tabularia indivisa, aiul Forskal some very good ones of other hydroids, while some
tolerable figures of a Tabularia and of some Sertularian and Campanularian hydroids had
been published by Raster ;" but hitherto no attempt had been made at the publication of
coloured drawings. Between 1777 and 1780, however, were issued the first two fasciculi of the
' Zoologia Danica' of O. F. Miiller,^ which after Midler's death was continued with additions by
Abildgard. It contains coloured figures of Scandinavian animals, mostly invertebrate, from the
surrounding seas, and amongst them several hydroids. la the accuracy, beauty, and abundance
of the figures, too much praise cannot be given to the ' Zoologia Danica,' which marks out an era
in zoological iconography.

The posthumous work of Ellis and Solander,' published in 1786, contains many hvmdreds
of figures, chiefly of corals, but having also among them several hydroids. Many of the figures
contained in this work are masterpieces of iconography.

Esper also gives us a most lal)orious iconography, partly copied, partly original, consisting of
coloured figures of' corals, sponges, &c., as well as of numerous hydroids.^ Where the hydroid
figures are not copied from Ellis they are vastly inferior to those of the English naturalist.

The naturalists who during the eighteenth century contributed most to advance our
knowledge of the Hydroida close with the name of Cavolini. Cavolini, like Ellis, studied the
Hydroida in a living state. His investigations were made in the Bay of Naples, where he dis-
covered many hydroids previously unknown, and determined many points of interest in their
.structure and physiology. He was the first to observe a Medusiform gonophore in connection

' ' Icones Rerum Naturalium quas in Itinera Oriental! tlepingi curavit Pctrns Forskal.' Copen-
hagen, 1776. The descriptions are contained in a separate volume, published in 1775.

^ Jobi Basteri, ' Opuscula Subseciva.' Harlemi, 1762.
Otho Fredericus Miiller, ' Zoologise Danicae seu Aniraaliuni Danire et Norvegi<e variorum et
minus notorum Icones.' Hafnisc, 1777 — 1780.

* ' The Natural History of many curious and uncommon Zoophytes collected from various parts of
the Globe, by the late John Ellis, F.R.S. Systematically arranged and descril)ed by the late Daniel
Solander, M.D., F.R.S.' London, 1786.

' Die Pflan.senthiere in Abbildungeu uach der Natur mit Farben crleuchten uebst Beschrei-
bungen,' von Eugenius Johann Cristoph Esper. Nuremberg, 1791 — 1797.

8 THE IIYDROIDA IN GENERAL.

with the trophosomc, and has described the radiating canals and the included ova of this body
in his Scrfidaria pennaria (Pennaria disfi/c/ia, Goldfuss), without, however, exactly compre-
hending its true significance or its relations to a free hydroid Medusa. He also insisted on the
vegetality of the proper corallines or nullipores, which, on the estabhshment of the animality of
corals, were carried with these into the animal kingdom. The results of his researches were
published in 1785, in a work' full of valuable information, and illustrated with excellent figures
of living hydroids, corals, and Polyzoa.

In the ' Elenchns Zoophytorinn ' of Pallas, published in 17C6, all the known hydroid
trophosomes were distributed among three genera — Hydra, Tnbnlaria, and Sertidaria. In the
' Spicilegia' Pallas adds the genus Coryne from Gaertner MS. Gmelin, in his edition of the
' Systema Naturae,' 1788, while he overlooks the genus Coryne, adds the new genus Clava.
Besides these different genera of hydroids, all characterised from their tro])hosomes, several true
hydroid Medusae had been at this date known and described; but they were all included along
with steganophthalmic forms, and with Siplionophora and Ctenophora, under the common generic
name of " JNIedusa," given to them by Linnaeus.

The state of the natural history of the Hydroida at the date of the publication of the
thirteenth edition of the ' Systema Naturae' (Gmelin's) may thus be stated in a few words : — The
animality of the Hydroida was fully acknowledged. Such species as were known by their
trophosomes were distributed under five genera — Hydra, Tubularia, Sertularia, Coryne, and
Clava, while such free gonophores as were known were thrown together with all the other free
forms of Hydro~oa under the common name of Mediim.

The natural history of the Hydroida, which during the latter half of the eighteenth century
had been thus steadily advancing in the hands of Trembley, Jussieu, Ellis, Pallas, Forskal, O. P.
Midler, and Cavolini, was, with the commencement of the nineteenth century, destined to receive
a fresh impulse.

The famous voyage of Peron and Lesueur inaugurates the natural history labours of the
nineteenth century. It was commenced in 1800, and in 1S04 the voyagers returned laden with
new and important facts for science. No expedition could have afforded better opportunities of
studying the pelagic forms of invertebrate animals ; and soon after their return Peron and
Lesueur undertook a systematic description of the Medusce which they had obser\'ed in the great
seas which their ships had traversed, as well as of other species which they had studied in expeditions
afterwards made to the coasts of Normandy and to the Mediterranean. In the two memoirs^ in
which they pulilish the results of their researches they propose an entirely new classification of the
Medusa3. The old Linnean genus Medusa is broken up into numerous separate genera, and

' Filipo Cavolini, ' !Mcmorie per scrvir alia storia dc Polypi Marini.' Naples, 1785. Translated
in 1813 into German by Sprengel.

" ' Voyage de Decouvertes aux Terres Australes, fait par ordre du Gouvernement sur les cor-
vettes " le Geographe," " le Naturalists," et la goelette " la Casuarina," pendant les annees 1800 a 1804,
redige par Peron et continue par M. Louis de Freycinet,' 2'' edit, revue, corrigee, et augmentee, par
M. de Freycinet. Paris, 1824, 1825.

^ Peron et Lesueur, " Notions preliminaires sur les Meduses," ' Ann. du Museum,' 1809,
p. 218; and 'Tableau des Caracteres generiqucs et specifiques de toutes les Especes de Meduses
conuues jusqu'a ce jour,' id., p. 325. The plates referred to all through the second memoir have,
unfortunately, never been published.

HISTORY OF RESEARCH. 9

many true hydroid Medusa; arc (Icscribed; but the autliors had as yet failed to recognise the
fundamental differences between the hydroid Medusa^ and the proper Biscophora.

In 1812 Cuvier published a sketch^ of his celebrated arrangement of the animal kingdom,
which he divides into four primary groups. To the last of these he assigns the name of
" Animalia Zoophyta sen Radiata." The group Radiata of Cuvier thus includes all the
Hydroida, but though more precise and definite than the "Vermes" of Linnaeus, it is still a
heterogeneous asscmlilage, and as it fails to recognise the distinction between grade of develop-
ment and morphological plan, it necessarily contains forms which belong to very different types.

The beautiful researches of Savigny on the compound Ascidians were published in 1816,'
and by proving that a large number of organisms which, under the common name of Alcyoniuin,
had been hitherto associated with true coelenterate forms, are in reality Ascidians, these researches
must be regarded as an important step towards the final limitation of the primary groups of tiic
animal kingdom.

For some time past a vast amount of material for the zoology of the invertebrate animals was
being accumulated, and a period had now arrived when a systematic arrangement of the whole was
loudly called for. It was in this state of things that a work destined to exert great influence on
the study of the lower animals made its appearance. The second volume of the ' Histoire Naturelle
des Animaux sans Vertebres' of Lamarck^ was published in 1816. In this celebrated work three
new genera of hydroids are instituted from their trophosomes, namely, Campanuhiria, Antennularia
and Plumularia. Among Medusae, however, Lamarck recognises only a portion of the genera
established by Peron and Lesueur, and in thus attempting to simplify the classifications of his
predecessors he falls behind the famous voyagers in the actual requirements of science.

While Lamarck was engaged in the preparation of his ' Animaux sans Vertiibres,'
Lamouroux was occupied with the study of a set of flexible, plant-like organisms forming a
heterogeneous group, which included not only most of the hydroid trophosomes then known, but
also a large number of Actinozoa and Poli/zoa, and even many indubitable plants ; and in the
same year with the publication of the second volume of Lamarck's work there appeared a natural
history of these organisms by Lamouroux. Lamouroux'* has here defined some good additional
genera of hydroids characterised by their trophosomes ; though some of them are identical with
genera instituted by Lamarck under other names. The names given by Lamarck, however, have
found more general acceptance with subsequent authors ; and whatever doubt may be enter-
tained regarding actual priority of publication, the zoologist of the present day will scarcely
hesitate to give them the precedence, especially when it is remembered that Lamouroux had
complete access to all Lamarck's specimens which had been deposited in the Museum of the
Jardin des Plantes, and which had been already labelled with the names given to them by the
illustrious author of the ' Histoii-e des Animaux sans Vertebres.'

' G. Cuvier, " Sur un nouveau rapprochement & etablir entre les classes qui composent le Regnc
Animal.'' ' Annales du Museum/ 1812.

" Marie Jul. Caesar Lelorgue de Savigny, ' Memoires sur les Animaux sans Vertebres,' part ii,
Paris, 1816.

' J. Bapt. P. Ant. de Monuet Lamarck, ' Histoire Naturelle des Animaux sans Vertebres,'
7 vols., Paris, 1815—1822.

"" J. V. F. Lamouroux, ' Histoire des Polypiers Coralligenes flexible vulgairement nommes
Zoophytes,' Caen, 1816.

•I

10 THE IIYDROIDA IN GENERAL.

Almost precisely at the same time appeared the first edition of the ' Regne Animal' of
Cuvier.^ The general division of the animal kingdom, of which, as we have already seen, a
sketch had previously appeared, is here adopted and carried into detail. In his association of
the free Hydrozoa into a distinct class under the name of Acalepha, Cuvier now takes an
important step, though its value is deteriorated by the admission of Actinia into the same
group. Independently of the advance which the natural history of the Hydroida has thus
directly received, the great influence which Cuvier has exerted on the studies of the zoologist, by
taking anatomical structure rather than external resemblance as the basis of classification, renders
it impossible in the history of any department of zoology not to see in the puljlication of the
' Regne Animal' a well-marked era of development.

The voyage of Perou and Lesueur, which had such valuable results for zoology, was only the
first of a long series of scientific expeditions which, fitted out under the auspices of various
governments, brought back with them rich stores of materials, and mark out the first half of the
present century as eminently the era of the naturalist voyager.

Between the years 181.5 and 18:26 two exploratory expeditions round the world were fitted
out by tlie Russian Government." They were entrusted to the command of Kotzebue, and were
accompanied by Chamisso and Eschscholtz as naturalists. The expeditions afforded fine oppor-
tunities for the observation of pelagic forms of Mo/lttsca and Coelenterata, and are rendered
memorable by Chamisso's famous discovery of the " alternation of generations " in Saljui — a
discovery which was destined to exert great influence on the study of the Hvdroida and the
interpretation of their marvellous life-history.

The study of the ccelenterate animals observed during these voyages was specially under-
taken by Eschscholtz, and after the return of the voyagers from their second expedition we find
this philosophic naturalist publishing a general work on the Medusie and allied forms ' — a work
by far the most important which had as yet appeared upon the animals of which it treats, and
one which, even at the present day, the student is unable to dispense with.

It is here that, for the first time, we find the hydroid Medusae, under the name of " Discophora
cryptocarpa," separated as a distinct and well-defined group from the proper Discophora, to
which Eschscholtz assigns the name of " Discophora phanerocarpa ;" and though the characters
on which this dismemberment was based were but imperfectly understood by Eschscholtz, and
have since undergone considerable modifications, the conception of the hydroid Medusas as a
separate section is a step of primary importance, and could have been entertained only by one
who was able to recognise the fundamental differences and appreciate the true affinities of the
group among which these organisms had been hitherto indefinitely distributed.

Cuvier had already * recognised an essential difference of structure between the actinozoal
and hydrozoal forms included in his group of " Polypes," when he pointed out the presence of a
digestive sac, with differentiated walls in the former, and its absence in the latter ; but the

^ Geo. Leop. Chr. Fred. Dajob. Cuvier, ' Le Regno Animal distiibue d'apres son organisation, pour
servirde base il I'Histoire Naturelle des Animaux at d'introduction a I'Anatomie comparee,' Paris, 1817.

" Otto V. Kotzebue, ' Voyage of Discovery into the South Sea and Behring's Straits, undertaken
in 1815-18, in the ship " Rurick," ' Loudon, 1821 ; and ' New Voyage round the World in 1823-26,'
London, 1830.

' Job. Priedr. Eschscholtz, ' System der Acalephen,' Berlin, 1829.

* 'Le Regne Animal,' 1817, tome iv, p. 79.

HISTORY OF RESEARCH. 11

structure of most of the animals constituting the Jladiatn of Cuvier was still so very imperfectly
known that this important character failed to receive its due weight in classification.

In 1828 it was brought out with greater distinctness and force by M. Milne-Edwards in
an account of his zoological researches in the Chausey Isles' ; but still the fixed plantlike
hydroids continued to be associated with the true corals and PoJijzoa, under the common name
of " Polypi," for, notwithstanding the deep-lying difference which had been indicated by Cuvier,
and more definitely insisted on by Milne-Edwards, between the hydrozoal and actinozoal types,
yet its full value as a classificatory character still continued unappreciated. As Eschscholtz,
however, had seen in the reproductive system of the medusae a ground for their separation into
two primary groups, so in the same year Professor Rapp, of Tubingen, proposed to divide the
polypoid cwlenterata into two sections, also based upon the peculiarities of their reproductive
system." Observing that in the hydroid trophosome the ovigerous buds were produced externally,
while in Actinia and the corals the reproductive organs projected into the interior of the Ijody-
cavity, he assumed this diiference as a basis for the definition of two distinct groups, to which
he gave the names of Endoarii and Exoarii. Though the real nature of the peculiarity on which
this dismemberment of the fixed ccelenterata rests was, like that in accordance with which
Eschscholtz had already based his subdivision of the Medusae, scarcely comprehended by its
author, the proposal of Rapp must, nevertheless, be regarded as an important step towards the
determination of the systematic position of the Hydroiua.

Among the results of Milne-Edwards's investigations carried on in the Chausey Isles, the
most important was his demonstration of a type of structure in the Flustrce, entirely diflerent
from that of the ccelenterate animals with which they had been associated by his predecessors.
A similar conclusion had been just arrived at by Grant,^ but the British zoologist had not
worked out the structure with that completeness which characterised the investigations of
Edwards, who now showed* that the Flustra were constructed on a plan in all essential points
identical with that of the compound Ascidians, with whose organisation Savigny had already
made us acquainted in his masterly memoirs.

Though M. Milne-Edwards had thus proved the existence of a true molluscoid type of
structure in Flustra, no comprehensive name had yet been given to the group so characterised.
While Grant and Edwards, however, were thus engaged in their anatomical examination of
Flustra, J. V. Thompson, then residing on the coast of Cork Harbour, was occupied with a series
of similar investigations, and, quite independently of any knowledge of the labours of Grant and
Edwards, had determined not only tlie molluscoid structure of Flustra, but had in a very complete
way demonstrated an entirely similar structure in other plant-like organisms hitherto associated
with true ccelenterate forms. To the organisms thus characterised by this common type of
structure Thompson, while fully recognising their relation to the Ascidians, gave, in 1830, the
name of Volyzoa:'

^ J. Victor Audouin and H. Milne-Edwards, " Resume des Recherches sur les Animaux sans
Vertebres, faites aux Isles Chausey," ' Ann. des Sci. Nat,,' 1838.

^ Wilhel. Rapp, ' Ueber die Polypeu in AUgemeinen und die Actinien insbesondere,' Weimar, 1829.

■' Robert E. Grant, " Observations on the Structure and Nature of Flustrse," ' Edinb. New Phil.
Journ.,' vol. iii, 1827.

* Op. cit.

' John Vaughan Thompson, 'Zoological Researches and Illustrations,' Cork, 1830. The date is

12 THE HYDROIDA IN GENERAL.

In 1S3P Ehrenberg employs the term Bri/ozoa in a sense exactly equivalent to that of
Thompson's Poli/zoa, and in 1833" we find him proposing a new classitication of the heterogeneous
and unscientific grouj) of the " Polypi," by dividing them into the AntJiozoa, which embraced the
coelenterate forms, and the Bryozoa to which all the moUuscoid forms were referred.

In the same memoir Ehrenberg also forms a separate group for the fixed Hydrozoa, which,
under the name of " Zoocoralia oligactinia," he separates from the Actinozoa. He further gives
a synopsis of such genera and species as were then known, and makes some valuable reforms in
the limitation and arrangement of the genera ; but the most important point in which this memoir
has advanced our knowledge of the Hydroida will be found in the ascription of an independent
zooidal significance to the so-called " egg-capsules" of these animals, and the consequent deter-
mination of a distinct sexuality among the zooids which compose a hydroid colony. This capital
discovery, whose true import, hovvevei', was but partially comprehended by Ehrenberg, has in the
hands of subsequent investigators undergone further development, and must be viewed as the
starting-point for all those more recent researches which have so largely contributed to bring
about the philosophical views now entertained regarding the structure, physiology, and
systematic position of the Hydro in a.

In 1834 De Blainville published his 'Manuel d'Actinologie.'' The important reforms of
Eschscholtz, Rapp, and Ehrenberg are not adopted by De Blainville ; and the ' Manuel
d'Actinologie,' except in so far as it constituted a useful work of reference and description for the
student, cannot be regarded as in any way advancing the systematic arrangement, or aiding in a
philosophic conception of the Hydroida.

The notes and drawings of Medusae made by Mertens during his voyage rovmd the world
as naturalist to the Russian exploring ship " Seniavin," were, after Mertens' death, entrusted
to Brandt, who has given us an account of them in two important memoirs. The first, published
in 1838,* consists in a synopsis of the genera and species observed by Mertens, among which are
several new genera of both hydroid Medusae and Discophora proper ; while in the second* he
gives us a more detailed account of them, and now publishes the numerous and beautiful figures
made by Mertens from the living animals. Brandt does not adopt Eschscholtz's division of the
Medusae, so that the hydroid Medusae are here mixed up with the Discophora proper. He gives
us detailed anatomical descriptions of the Medusae, so far as their structure was at that time
known ; but the chief value of the labours of Mertens and Brandt will be found in their rendering
us acquainted with new forms, and in their giving us the most beautiful and accurate figures of
Medusae which had been up to that time published.

not printed on the title-page, but it will be found on the paper wrapper in which the publication was
originally issued.

' Chr. Gdfr. Ehrenberg, ' Symbols Physicse,' iv, Berolini, 1831.

" Corallenthiere.

^ Henri Marie Ducrotay de Blainville, ' Manuel d'Actinologie ou de Zoophytes,' Paris, 1834.

* J. F. Brandt, " Prodromus descriptionis Animalium ab H. Mertensio in Orbis Terrarum cir-
cumnavigatione obscrvatorum." ' Recueil des Actes de la Seance publique de 1' Academic Imperials
des Sciences de St. Petersbourg,' 1833-31.

* Ibid. " Ausfiihrliche Beschreibung der von C. H. Mertens auf seiner Weltumsegelung beo-
bachteteu Schirmquallen," ' Mem. de I'Acad. Imper. des Sciences de St. Petersbourg,' vi ser., Sci.
Nat., tom, ii, Petersbourg, 1838.

HISTORY OF RESEARCH. 13

In 1838 Dr. Johnston published his 'History of Britisli Zoophytes."' The "zoophytes" of
Johnston inchicle not only the plant-like hydroids, but the Jclif/ozoa and Polpoa. He does not,
however, confound the natural boundaries of these groups, and proposes for the hydroid forms the
name of Hydroida, which thus coincides with the Hydroida of the present Monograph in ail
respects, except in the fact of its not including the hydroid Medusaj, whose relation to tlic plant-
like trophosomes had not yet been definitely recognised.

Without the originality of Ellis's classical ' Essay,' Johnston's ' History of British Zoophytes'
is still a work of great utility for the student. The descriptions of the species are very good,
and are accompanied by a copious and valuable synonomy ; and the figures, though mostly drawn
from the dried hydrosome, and certainly not equal in artistic feeling to those of Ellis, are often
excellent, and always of great use in aiding in the determination of the species. The value of the
' British Zoophytes,' however, lies in its character as a descriptive work, and with its publication
we may date a new impulse to the study of the Hydroida, similar to that which nearly a century
before, Ellis's 'Natural History of Corallines' had exercised in the same direction. In 1847 a
second edition of the ' British Zoophytes ' made its appearance. A great number of additional
species are described in it, and many new plates are added to those which were contained in the first.

The observations of Cavolini" in the last century, and afterwards those of Wagner' and
Loven,* had already made us acquainted with certain facts which show that the hydroid tropho-
some may give rise to buds presenting a close resemblance to Medusae; while the beautiful
researches of Sars had shown that among the Discojjhora phenomena occur which in many points
resemble this budding of Medusae among the Hydroida, and have an intimate relation with it.^
The true significance of these observations, however, was but imperfectly appreciated when
Steenstrup, in 1842, combining them with analogous ones in other groups of animals, correlated
with great skill all the known facts, and generalised the phenomena under the name of " al-
ternation of generations," an expression already employed by Chamisso when describing the
gemmation and generation of Salpa!^

Though in the terms in which Steenstrup enunciates his law of alternation of generations a
false conception of the phenomena may appear to be involved, it is evident that his own view of
them is a correct one, and the modification which Steenstrup's expression of the law has since
undergone can never deprive it of its value in opening up more philosophic views of the
morphology of the invertebrate animals, and marking out a new era in their study.

^ George Johnston, 'A History of the Britisli Zoophytes,' Edinburgh, 1838.

^ Op. cit., 1785.

' Rudolf Wagner, ' Neue im Adriatischen Meere gefundene Art von nacktem Armpolypen,'
Isis, 1833, iii, p. 256.

* S. L. Loven, " Beitrag zur Kenntni.ss der Gatungen Campaiiuhiriu und Syncoryne," ' Muller's
Archiv,' 1837.

" Martin Sars, ' Beskrivelser og Jagttagelser over nogle raoerkelige eller nye i Havet ved den
Bergenske Kyst levende Dyr,' &c., Bergen, 1835; and ' Ueber die Entwickclung der Medusa aurita
und der Cyanea capillata,' in 'Wiegm. Arch.,' 1841.

° Joh. Japetus Steenstrup, ' Ueber den Generationswechsel oder die Fortpflanzung und Eut-
wickelung durch abwechselnde Generationeu, eine eigenthiimliche Form der Brutpflege in den niederu
Thierclassen. Uebers. von Lorenzen,' Kopenh., 1842. Also translated for the Ray Society, by Busk,
Loudon, 1815.

14 THE HYDROIDA IN GENERAL.

Nearly at the same time some beautiful additional observations were made by Dujardin/
who traced various known forms of free hydroid Medusa; to their fixed trophosoraes, and had
seen them produce eggs ; from this epoch we may regard as definitely established the genetic
relation between the free hydroid medusa and the fixed trophosorae.

In 1843 two memoirs on the Campanularian and Tubularian Hydroids were presented by
Van Beneden to the Royal Academy of Brussels." The structure and gemmation of several
sj)ecies are described at length, and the memoirs are accompanied by very beautiful figures.
He has seen and described with much detail the free medusoid gonophores, as well as the fixed
sporosacs, but he mistakes the former for embryos destined by direct metamoqihosis to become
changed into the form of the polypoid trophosome. He founds the new genus, Hi/dractulia, for
the Alcyonium ecUnatmn, of Fleming, and calls attention to the polymorphism of the zooids in
this interesting and remarkable hydroid.

The voyage of the French corvette, " La Coquille," under the command of Duperrey, during
the years 1822 — 1825,^ afforded to Lesson, who, along with Garnot, accompanied the
expedition as naturalist, fine opportunities for the study of the Medusae ; and in 1843 we find
him publishing, as a volume of the " Suites a Buffon," his 'Histoire Naturelle des Acalephes.'*
Lesson's work contains a great mass of information, and shows that its author must have had a
very extensive acquaintance with the Medusae, and yet we cannot say that, beyond the description
of some new forms, our knowledge of the Hydroida has received from him any advance.
Indeed, he does not avail himself as he might of the discoveries of his predecessors, while in his
classification the Hydroid medusas are, as usual, mixed up with the BiscopJiora proper.

Up to this ])oint of our history it is plain that the systematic w^-iters who came after
Eschscholtz have fallen behind him in their appreciation of the Hydroid medusae as a natural
group ; for though Eschscholtz misunderstood the peculiarities of organization on which he
founded his "Discophora cryptocarpge," this group is not the less a natural one, and the
cryptocarjm of Eschscholtz must be recognised in every system which would aim at expressing
the true position and affinities of the various members of that large and heterogeneous assemblage
of organisms which have been included under the common name of Medusae.

It was not, indeed, until 1846 that the Eschscholtzian division of the Medusa? into two
grand groups was distinctly accepted by any other naturalist. In this year, however, a paper
was read before the British Association by Edward Forbes,^ in which the author divides the
Medusae into two groups, corresponding with i\\e phanerocarpcB and cryptocarpm of Eschscholtz.

The erroneousness of Eschscholtz's interpretation of the characters on which he founded his
subdivision had by this time become apparent, and Forbes accordingly, while admitting the
value of the groups, bases them on other characters than those employed by Eschscholtz ; for he
finds in the condition of the marginal bodies and of the gastro-vascular canals points of structure

' Fel. Dujardin, "Observations sur un nouveau genre de Medusaires [Cladonema) proveuaiit de la
Metamorphose des Syncorynes," 'Ann. Sci. Nat.,' 2e ser., 1843.

- P. J. Van Beneden, ' Recherches sur I'Embryogenie des Tubulaires, et I'Histoire Naturelle des
differents genres de eette famille qui habitent la cote d'Ostende,' Bruxelles, 1843.

' 'Voyage autour du Monde sur la Corvette la Coquille, pendant les Anuees 1822, 1823, 1821
et 1825, par M. L. J. Duperrey, capitaine de fregate, commandant I'expedition.'

^ Eene Primevere Lesson, ' Histoire Naturelle des Zoophytes. Acalephes.' Paris, 1843.

•^ Edward Forbes, " On the Pulmograde Medusse of the British Seas," ' Brit. Assoc. Rep.' for 1846.

HISTORY OF RESEARCH. 15

by which the two divisions may be anatomically characterised. In 181S I'orbes's views were
further developed in a beautiful monograph on the Naked-eyed Medusae of the British seas,'
in which every known British species of hydroid Medusa is described and illustrated by
an original figure. In this work he distinguishes the two groups by the names of " Stcganoph-
thalmia," corresponding with the Discophora plianerocarpa of Eschscholtz, and " Gymnoph-
thahnia," corresponding with his Discophora cryptocarpa. AVitli Forbes's monograph we may
date the definite acceptance of the hydroid jMedusae as a well-marked and legitimate group.

The Radiata of Cuvier, which, by the elimination of various groups originally included in it,
had been gradually attaining to a form more in accordance with the requirements of a natural
classification, was in 1847 subjected to an important revision by Leuckart," who insisted on the
necessity of attending to a remarkable type of structure, which was common to certain members
of the Radiata as then accepted by zoologists, namely, the free communication of the general
body-cavity with the external world through the mouth. He saw in this feature a character of
great value by which, after the exclusion of the Polyzoa, the whole of the Cuvierian " polypi "
and " acalepha " would requii'c to be united into a distinct group apart from the EcJiinodermata.
To the group thus constituted Leuckart gave the name of Ccdenterata ; while soon after Huxley
was led to adopt similar views, and, quite independently of Leuckart, proposed the construction
of an exactly equivalent group, under the name of Nematophora, suggested by the universal
presence of thread-cells in the tissues of the animals composing it.^

The relation between the fixed hydroid trophosomes and the free hydroid meduste, which,
as we have seen, had already become apparent, received about this time additional light
from the researches of Dalyell,* an acute, laborious, and conscientious observer, though without
that technical precision in his descriptions which indicates a special zoological training, and
without much acquaintance with what had been already done by others. Dalyell has recorded
many additional instances of the development of Medusae from the hydroid trophosome ; and the
accumulation of such facts now began to exert an influence on the classification.

We accordingly find Carl Vogt in 1851 combining the fixed hydroids with the whole of the
Discophora of Eschscholtz, in order to form a single group, to which he gives the name of
Hydromedma!'

In 1853 we find Kolliker employing the name of Hydromedusida^ but restricting it to a
group composed of the gymnophthalmic Medusse of Forbes, the Siphonophorous genus Velella,
and aU the non-natatory hydi-oids, except Hydra, which he unites with the rest of the Siphono-
phora to form his group Hydroiden, while he confines the name Discophora to the Sleyanoph-
thalmia of Forbes, or Discophora phanerocarpa of Eschscholtz.

These proper Hydrozoal groups are, along with the Ctenophora, united to the Actinozoa, in
order to form his division Radiata molluscoida, which would thus constitute a great natural

' Edward Forbes, ' A Monograph of the British Naked-eyed Medusse,' London, 1848. Published
by the Ray Society.

• Hnr. Frey und Rud. Leuckart, ' Beitrage zur Kenntniss wirbelloser Thiere,' Braunschweig, 184".
' Tb. H. Huxley, " An Account of Researches into the Anatomy of the Hydrostatic Acalepha;,"

' Brit. Assoc. Rep.' for 1851.

* Sir John Graham Dalyell, ' Rare and Remarkable Animals of Scotland,' Loudon, 1847.
" Carl Vogt, 'Zoologiscbe Briefe,' Frankfurt-a-M., 1851.

^ Albert Kolliker, ' Die Schwimmpolypen oder Siphonophoren von Messina,' Leipsig, 1853.

16 THE HYDROIDA IN GENERAL.

group correspoiuling to the Ccelenterata of Leuckart, were it not that KoUiker unfortunately
combines with the coelenterate forms the Poli/zoa, so as to form by the union of the two his
Radiata molluscoidea.

The same year Leuckart' jiroposed to divide the whole of the Ccelenterata into three classes
— the Ctenophora of Eschscholtz, composed of the Beroes and their allies ; the " Medusae "
(Scheibenquallen), and the " Polypes," which last name he confines to the Actinozoal Cwlenterata.
Under the " Medusse " of this classification are comprised three orders — the Dkcophora proper,
tiie llydroida with their Mediisffi, and the SipJionophora.

The union thus nearly simultaneously proposed by Kolliker and by Leuckart of the non-
natatory forms of hydroids with the gymnophthalmic Medusae so as to form a group distinct from
tluit of the steganophthalmic Medusae, is the first expression we have of a natural relation which
ail subsequent research has oidy tended to confirm.

\n 18.56 Leuckart published his Supplement to Van der Hoven's 'Manual of Zoology. ^ In
this valuable little work we find him distributing the gymophthalmic Medusae between two
groups. Those which he believes to undergo a direct metamorphosis without the intervention of
a hydroid trophosome are combined into one order, to which he gives the name of Ceratostera,
from the rigid habit of their marginal tentacles ; while those which are known to be produced as
buds from a trophosome are associated with their trophosomes, and with other fixed hydroids in which
the gonophore never assumes the form of a free medusa, so as to constitute his order Hydroidea.

In the same year Huxley' proposed to divide the group Cwlenterata into two classes, which
he names Hydrozoa and Acfinozoa. Instead of assigning to the Ctenophora the value of a
distinct class, he regards them as Actinozoa, thus bringing them into immediate relation with
Actinia.

In 1857 Gegenbaur published a very valuable paper on the Medusae,* in which he describes
and figures many new forms of Gymnophthalmia, the subordinate groups of which he revises and
limits more in accordance with the natural affinities of the animals than had been hitherto
attempted. He adopts the Eschscholtzian subdivision ; but instead of basing it on the characters
assumed by Eschscholtz, or on those proposed by Forbes, he finds the grounds of the subdivision
in the presence or absence of the membranous diaphragm which extends horizontally inwards
from the margin of the unibclla, and which Forbes had already designated by the name of
"Velum." In accordance with this view Gegenbaur divides the Medusae into the Acraspeda, or
those in wliicli no velum is developed, and which correspond to the Steganoplithalmia of Forbes,
and the Craspedota, or those which are provided with a velum, and which correspond to the
Gymnophthalmia of Forbes.

In 1858 McCrady published an important paper on the " Gymnophthalmic Medusae,"^ in which
many new forms are described and figured. Instead, however, of limiting his group Gymnoph-
thalmia within the bounds of that to which Forbes had already assigned this name, he extends

' Rud. Leuckart, ' Zoologiscbe Untersuchungen.' Erstes Heft. Giessen, 1853.

^ Eud. Leuckart, ' Naclitrage nud Berichtiguugen zu dcm ersten Bande von J. Van der Hoven's
Handbuch der Zoologie,' Leipsig, 1856.

^ Thomas Henry Huxley, " Lectures on General Natural History," in ' Medical Times ' for 1856.

* Carl Gegenbaur, " Versuch eiues Systemes der Medusen," 'Zeitsch. fiir Wisseusch Zoologie,' 1857,

'•" J. McCrady, the Gymnophttialmata of Charleston Harbour, ' Proc. of the Elliott Society of
Natural History,' Charleston, 1858.

HISTORY OF RESEARCH. 17

it by uniting witli tlic mcdusoid form wliicli constitutes the Gymnophthnlmin of Forbes, not only
the polypoid forms of liydroids, but also the Siphonophora. The Gi/mnopJilludmia of iM'Crady
thus correspond to the lljdromedusida and lljjdroidea of KoUiker united.

In a small treatise on the alternation of generations among the liydroids,' Gegenbaur had
already contributed some valuable observations to our knowledge of the generative phenomena
and life-history of the IIydkoida; while, as we have seen, he had also made the so-called naked-
eyed ]\Iedusa3 a subject of careful systematic study. In his ' Outlines of Comparative Anatomy,'
pubhshed in 1859,^ he retains his divisions of Acraapeda and Craspedota, uniting them into
a single order, Medusida, while he combines the various non-natatory hydroid colonies into
another order under the name of Ilydroidu.

In 1859 Huxley gave us a monographic treatise on the Siphonophora observed during the
circumnavigatory voyage of H.M.S. "Rattlesnake."' The "Rattlesnake" was fitted out in
1 846, imder the command of Captain Owen Stanley, for the purpose of surveying the channel lying
within the great barrier reef which extends along the east coast of Australia, and for the explora-
tion of the neighbouring seas ; and the expedition was accompanied by Mr. Huxley as assistant-
sm-geon. Owing to the refusal of the Admiralty, on the return of the expedition, to furnish the
means of publication, the results of Huxley's observations remained unpublished, until, after many
fruitless attempts to obtain the aid of Government, and many years of vexatious delay, the
Ray Society undertook the task of publication. In this valuable work the Siphonophora are
described under the designation of Oceanic Ilydrozoa. The special part of the treatise is preceded
by a general introduction, which abounds in original and philosophic views of the morphology of
the Hydroida and of their relation to the other groups oi Hydrozoa ; and the author proposes
a new and comprehensive terminology, much of which has been adopted in the present
Monograph. In his systematic arrangement of the Hydrozoa; he does not venture to unite
into a single group with the polypoid phases of the Hydroida those hydroid Medusae which
have not been proved to proceed from a polypoid trophosome, but prefers to arrange them
as a distinct order of Hydrozoa under the name of Medusidee, attaching, however, only a
provisional significance to the group thus constituted.

Between 1860 and 1862 there appeared the third and fourth volumes of Agassiz's 'Con-
tributions to the Natural History of the United States.'* These volumes are devoted to the
Hydrozoa, which are treated of under the designation of Acalepha?. We learn from the
preface to the first volume that the author has been assisted by Prof. H. J. Clark, to whom
the microscopical researches which form so valuable a portion of the work are mainly due.
Many new genera and species of liydroids are described, and their tropliosomes as well as
gonosomes represented in elaborate and beautiful figures drawn from the living animal, while
the number and beauty of the drawings expressing anatomical and embryological details give to

^ Carl Gegenbaur, ' Zur Lelire voni Generationswechsel uud dor Fortpflauziing bei Medusen nud
Polypen,' Wurzburg, 1854..

" Carl Gegenbaui', ' Grundziige der Yergleichenden Auatomie,' Leipzig, 1839.

^ Thomas Henry Huxley, 'The Oceanic Hydrozoa; a Description of the Calycoplioridae and Phy-
sophoridK observed during the Voyage of H. M. S. " Rattlesnake," in the years 184G-50.' London,
printed for the Ray Society, IS.'iO.

* Louis Agassiz, ' Contributions to the Natural History of the United States of America,'
Boston, 1857—62.

3

18 HISTORY OF RESEARCH.

tliis work a spccinl value, and place it among the most important contributions we possess to the
natural history of the IItdroida.

Agassiz here divides his " Acalephaj" into three orders, CtenopJiora, Biscophora, and
llydroida, which last embraces the Gijmnnpldhahnia of Eorbes, the admitted polypoid forms of
Hydrozoa, and the Sijjhonopkora and Lucernarim ; and he also, contrary to the universally enter-
tained opinion of previous naturalists, maintains that the corals constituting the still living group
of the Tahulata, and those forming the entirely extinct groups of the Tubulosa and Uugosa, have
a true hydroid structure, and that they must accordingly be removed from the Adinozoa, with
which they had been hitherto associated, and take their place among the genuine Hydroida.

It will be seen that Agassiz here follows Kolliker, Leuckart, and M'Crady in uniting the
polypoid forms of Hydrozoa with the gymnophthalmic Medusae as elements in a group equivalent
in value with that of the Biscophora, a name which he employs in the limited sense adopted by
Kolliker to indicate the Biscophora phanerocarpa of Eschscholtz.

In 1861 Greene published his excellent little Manual of the Coelenterata,^ a work which gives
us in a condensed form a very complete view of the structure, development, and relations of the
various members of this group. He adopts the name of Hydrozoa in the sense in which it
was limited by Huxley, while he combines the gymnophthalmic Medusa? into a distinct order
from which the polypoid forms are excluded, and thereby as, I believe, fails to express the true
relations of these organisms.

Though Greene's Manual lays no claim to originality, it discusses in a philosophic
spirit various questions bearing on the subject with which it is occupied, and constitutes one of
the most valuable aids to the general study of the Cwlenterata which can be placed in the
hands of the student.

In the first volume of the ' Manual of Zoology,' by Peters, Carus, and Gerstaecker," pub-
lished in 1863, J. Victor Carus gives us among other articles an excellent one on the Cwlenterata.
The Hydromediism which form his third order of Ccelenferata are here divided into the Siphono-
phora and Hydroidea, the latter embracing both the free gymnophthalmic Medusae and the
polypoid colonies. The article contains an account of the leading anatomical and embryolo-
gical features of the Hydroida ; and the subordinate groups under which the author believes
that they ought to be distributed are characterised. It also contains a very useful synopsis with
diagnoses of all the genera of hydroids.

Observations on both the hydriform and medusiform elements of the Hydroida had been
thus for several years accumulating, and the time had already come when it seemed possible to
assign to the free hydroid Medusa its proper place in a comprehensive system of the Hydroida.
The importance of uniting the two elements in the definitions of genera had been already recog-
nised by M'Crady, Agassiz, and Victor Carus ; but notwithstanding the prominence which these
authors, and especially Agassiz, had given to the medusiform buds, it did not seem that a
thoroughly natural distribution of the Hydroida under legitimately limited genera had yet been
effected, and this belief led me in 1864 to attempt a revision of the older genera of all tubularian
and campanularian hydroids whose hydriform element was known.'

' Joseph Ileay Greene, 'A Manual of the Sub-kingdom CcBlenterata,' London, 1861.

^ Wilh. C. H. Peters, Jul. Victor Carus, und C. E. Adolph Gerstaecker, ' Handbuch der Zoologie,'
Leipzig, 1863.

' Allman, ' On the Construction and Limitation of Genera among the Hydroida.' ' Ann. and
:\rag. of Nat. Hist.' for May, 1864.

HISTORY OF RESEARCU. 19

In the diagnoses of tlie genera I regarded the reproductive zooid, wlietlier fixed or free, of as
much importance as the nutritive, and the resulting classification, in so far as it applies to the
subject of the present Monograph, is, with such modifications as further investigations rendered
necessary, that which I have here adopted.

At this period the yUffinidis constituted a group of Medusae whose structure was but imper-
fectly understood, and wdiose systematic position and affinities had given rise to nuich discussion.
In 1865, however, Ernst Haeckel published some very valuable observations^ which no longer
leave any doubt that these Medusae possess a true hydroid structure, by which tiiey become asso-
ciated with the ordinary hydroid or gymnophthnluiic Medusje ; while he still further made the
remarkable discovery that Citnina, a typical ^Eginidan, is produced as a bud from Geryonia, a
medusa of an entirely different form, and one whose true hydroid affinities had never been doubted.

In the same year, Alexander Agassiz publislied his illustrated Catalogue of the North
American Hydrozoa contained in the Museum of Comparative Zoology at Harvard College.'^
In this work, besides other hydrozoal groups, a large number of hydroid Medusae, occasion-
ally accompanied by their trophosomes, and including many new forms, are described and illus-
trated by very expressive woodcuts. The views of Prof. Louis Agassiz on almost all that
concerns nomenclature, generic groups, and systematic position and affinities are throughout
adopted by his son ; but whether these views be in all points accepted or not, we cannot but
regard the work of Alexander Agassiz as tending in no small degree to advance to our knowledge
of the zoology of the Hydkoida.

In 1S6G, while the present sheets were passing through the jiress, M. Van Benedcn
published in the ' Memoirs of the Royal Academy of Belgium ' a treatise on the Natural
History of the Ccelciilerata of the Belgian Coast, a group for which he uses the general name of
" polypes."'

The greater part of the work is devoted to the Hydroida ; it is illustrated by many
beautiful plates, and is one of the most elaborate treatises which has been hitherto published
on these animals. M. Van Beneden describes and figures not only a considerable number of the
fixed forms, but also many free medusiform gonophores ; and the work must be regarded as a
valuable contribution to the iconography and descriptive natural history of the Hydroida, though
the author's opinions on many points, more especially such as concern the synonomy and
determination of species, cannot receive our assent, and will be discussed in another part of the
present ^lonograph.

While the natural history of the Hydroida was thus gradually advancing towards perfec-
tion in all that appertains to descriptive Zoology and systematic arrangement, the anatomists were
developing it from a structural and embryological point of view. The introduction of the achro-

^ Ernst Haeckel, ' Ueber eiue iieue Form des Generationswechsels bei den Medusen, und iiber
dieVerwandeschaft der Geryonideu und Aegmiden Monatsbericht der Kouige Akad. der Wiss. zu Berlin/
2 Feb., 1863. Translated in the 'Ann. of Nat. Hist.' for Jan. 1865.

Id., ' Beitrage zur Naturgescliiclite der Hydrottiedusen,' Leipzig, 1805.

■ ' Illustrated Catalogue of the Museum of Comparative Zoology at Harvard College.' No. II.
"North American Acalephee." By Alex. Agassiz. Cambridge, U.S., 1805.

' P. J. Van Beneden, " Recherclies sur la Faune littorale do Belgique. — Polypes." ' Mem. dc
I'Acad. Roy. de Belg.,' tome xxxvi. Published also in a separate form.

20 HISTORY OF RESEARCH.

matic microscope had already placed in their hands a new and powerful instrument of investiga-
tion, and discoveries in the anatomy and physiology of the lower animals had begun to accumu-
late with a rapidity hitherto unprecedented. We have already referred to the important light which
the microscope in the hands of Ehrcnbcrg, Bars, Dalyell, Steenstrup, Dujardin, Leuckart, and
Gegenbaur had thrown upon the mutual relations and life-history of these animals, and we have seen
the influence it has thus had on their classification, while within the last twelve years the micro-
scopical investigation of the Hydroida has been pursued with an assiduity greater than at any
former period, and has yielded the results which might have been expected from the numerous
able observers who have been engaged in it. To give, however, in the present historical sketch
even an imperfect analysis of the various discoveries by which recent anatomical research has
enriched our knowledge of the Hydroida, would be impossible without extending our introductory
pages far beyond the limits within which it is expedient to confine them, and the reader must
accordingly be referred for an account of such discoveries to the parts of this work where they
will be specially considered.

The steps which in the history just sketched have exerted the greatest influence in deter-
mining the boundaries and systematic position of the Hydroida, in the sense in which this group
is limited in the present Monograph, would seem to be the following : —

1 . The discoveries of Trembley in the anatomy and physiology of Ihjdra.

2. The researches of Ellis among the marine forms, which resulted in the complete establish-
ment of the animality of the fixed plant-like Hydroida— residts to which the observations of
Peysonelle and De Jussieu had led the way.

3. The establishment of the Radiata by Cuvier as one of the four primary divisions of the
animal kingdom.

4. The recognition by Eschscholtz of two types of form among the Medusae, and his con-
sequent subdivision of these organisms into the Phanerocarjm and Cri/ptocarpa.

5. The recognition by Cuvier, Milne-Edwards, and Rapp, of the difference of structure
which separates the actinozoal from the hydrozoal forms of the so-called " Polypi."

6. The determination by Ehrenberg of the sexuality of the Hydroida.

7. The discovery of a genetic relation between certain free gymnophthalmic Medusae and
the fixed Hydroids, by Loven, Sars, Dalyell, Dujardin, and others, and the correlation of
this with analogous phenomena by Steenstrup.

8. The gradual rectification of the Radiata of Cuvier by the successive elimination of the
Nullipores, Worms, Infusoria, Rhizopoda, and Polyzoa.

9. The further analysis by Leuckart of the amended Radiata, resulting in his establishment
of the group Ccelenterata.

10. The union by Kolliker and Leuckart of certain free gymnophthalmic Medusae with the
fixed Hydroida, so as to constitute a single group equivalent in value witli that of the Discophora
or steganophthalmic Medusae.

MORPHOLOGY.

21

MORPHOLOGY OP THE HYDROIDA— TERMINOLOGY.

I. The IIydrosoma in general.

1 . Generalised conception of a Hydroid.

We shall best bring this part of our subject before the mind, if we first endeavour to conceive
of hydroid organisation freed from all non-essential comphcation.

Let us, then, imagine an open sac (woodcut,
fig. 1), whose walls consist of a double membrane ^'•'- ^•

{x,y), and whose orifice {h) is surrounded by a
circle of tubular tentacles (c) formed by csecal off-
sets from its cavity. Let us further suppose that
there projects from some part of the walls of this
sac another sac (c/), also composed of a double
membrane. To the morphological conception
thus acquired, let us add the physiological one
which we obtain by regarding the walls as
endowed with irritability ; and as a further phy-
siological element in our ideal picture, let us ima-
gine that while the former sac subserves the func-
tion of a digestive cavity, the latter, which has
been emitted as an external bud from its side, is
destined to give origin to generative elements
which are formed between the two membranes
which constitute its walls ; and we shall then
have as general an idea, morphological and
physiological, as it is possible to form of a
hydroid.

The hydroid organism, however, which we
have thus reduced to its simplest expression,
admits of much and varied complication, while
the actual forms which we meet with in nature

Diagramatic section of a Hydroid.

a. Body cavity ; 5, orifice serving for ingestion and egestion ;

c, tentacle ; x, outer membrane of body walls (ectoderm) ; y, inner

membrane of body-walls (endoderra) ; rf, generative sac, containing

.... eggs. The endoderm is throughout indicated by a darker shading

present a beautifully graduated series, in which than that of the ectoderm,
the law of specialisation is expressed with a

distinctness and significance in the highest degree instructive, and whose study possesses a pecu-
liar interest in leading us to those wider generalisations in which other groups of the liydrozoa
may be included, and by which we are enabled to assert a morphological unity of the whole.

2. General Structure — Ectoderm and Endoderm.

Every hydroid is composed of two membranes, an outer or ectoderm (woodcut, fig. 1, ■?'), and an
inner or endoderm [y], the ectoderm having its free surface in direct relation with the medium in which

22 MORPHOLOGY.

the animal lives, while the free surface of the cndoderm is turned inwards, and forms the boundary
of the gastric cavity and of all its prolongations through the organism. A similar composition may
be demonstrated not only in all the rest of the Ilydrozoa, but in the whole group of the Ccelen-
ierata. For the important generalisation which thus asserts the composition of every coelenterate
animal out of two membranes — a generalisation which forms the basis of the whole morphology
of the Ccelenterata — we are indebted to Prof. Huxley, who first enunciated it as a great
anatomical truth .^

Another character which the Hydroida possess in common with the entire group of Ca;len-
terata is the presence of the peculiar bodies known as thread-cells. These bodies, which will be
afterwards more particularly examined, are developed in the ectoderm, where they are frequently
aggregated in definite groups very characteristic of the species.

3. Composite character of the Hydroida — Trophosome, and Gonosome.

The Hydroida, wherever our knowledge of them is sufficiently complete to justify us in
arriving at any well-founded conclusion regarding the entire life of the individual are all, with
only a single apparent exception, composite animals at some one period of their existence, each con-
sisting then of an assemblage or colony of zooids,^ in organic union with one another (woodcut,
fig. 2). The colony thus formed constitutes the " hydrosoma" of Huxley.

It will be shown in the sequel that, except in the solitary — and, perhaps, after all, only

1 Huxley, " On the Anatomy and Affinities of the Medusae," ' Phil. Trans.,' 1849.

" For the introduction of the very convenient term " zooid" into the language of zoology, we
are indebted to Prof. Huxley, who, in defining tlie " individual" as " the total result of the develop-
ment of a single ovum," proposed to designate by the terra zooid all more or less independent forms
which may be included as elements in this total result. (See Huxley, " Observations on Salpa," &c.,
in ' Phil. Trans.,' 1851 ; " Lecture on Animal Individuality," ' Ann. Nat. Hist.,' June, 1853 ; and his
review of J. Miiller's " Researches on the Development of the Echinodermata," in ' Ann. Nat. Hist.,'
July, 1851. See also Carpenter, 'Principles of Gen. and Comp. Physiology,' 1851, p. 906, and 'Brit,
and For. Med.-Chir. Eev.' for Jan. 1848 and Oct. 1849, where the same idea is clearly supported.
Dr. Carpenter using the expression " a generation" for all that intervenes between one act of true or
sexual generation and another.

The distinction between a " zooid" and an " organ" is not always easy, and may indeed some-
times appear to be arbitrary. I believe, however, that we may define a zooid as a more or less
i)}d/vidualised animal organism, ivhich may or may not be capable of independent existence, and which
constitutes one of a series tvhose members are related to each other by some form of non-sexual reproduc-
tion, and morphologically repeat one another eitlier actually or homologically . In this sense not only
are the free medusiform buds of the Hydroida true zooids, but we must also regard as such the fixed
hydranths and those fixed gonophores which never attain a developed medusiform structure, as well as
the simple generative sacs which are developed on the radiating canals of Obelia, Thaumantias, &c.
(see p. 35).

On the other hand, from the above definition are necessarily excluded all mere organs, however
capable they may be for a longer or shorter period of self-maintenance, such, for example, as the
Hectocotylus of the Hcctocotylus-forraing Cephalopoda.

Many zooids may combine to form the true zoological Individual —the logical element of the species.

THE HYDROSOMA.

23

apparently exceptional — case here referred to, and which will ])C afterwards more particularly
described, those which have been adduced as affording exceptions to this phenomenon are in
every instance based upon incomplete observations ; that they may or may not be confirmed when
opportunities of more extended observation shall have been afforded ; and that accordingly, in the
present state of our knowledge, we are not justified in accepting them.

The associated zooids are always of two kinds. In one {a, b), the zooid is destitute of all
power of true or sexual generation, and has as its proper function the general nutrition of the colony.
The other group of zooids {/,(/) has nothing to do with the general mitrition of the colony ; it has as
its proper function true generation, and the zooids which compose it give origin to the generative
elements— ova or spermatozoa — either directly or after having first developed a special sexual bud.
For the whole assemblage of the former, or nutritive zooids, with their common connecting basis,
I propose the name of trophosome, while I shall designate the entire association of generative zooids
by the name of gonosome. Every hydroid, therefore, with whose life-history we are acquainted
consists essentially — with the solitary exception already alluded to — of a trophosome destined for
the preservation of the individual, and of a gonosome for the perpetuation of the species.

Fig. 2.

Hydrosoma of Campanularia Johnstoni.

a, b, c. d, e. Various portions of the trophosome ; /, /, /, g, of the gonosome, a, Ilydranth expanded ; li, hydranth
contracted; c, empty liydrotheca ; d, free portion of hydrophyton (hydrocaulus) ; e, adherent portion of hydrophyton
(liydrorhiza) ; /,/,/, gonangia, containing generative buds,' which in this genus are blastochemes ; g, a blastocbeme just after
its escape from tlie gonangiuoi.

24 MORPHOLOGY.

4. Orientation.

The attached extremity of the fixed hydrosoma, or its equivalent iu the free one, is
described by Huxley as the proximal end ; that which is placed diametrically opposite to
this is the distal end. In the present Monograph I shall employ the terms proximal and distal
in the sense thus proposed. Beyond these two aspects none other can be definitely distinguished
in the Hydroida. The determination of a right and a left side is impossible, for though an
apparent bilateralism occasionally occurs, as, for example, in the planoblasts of Corymorplia (PI.
XIX), and in those of Gemmellaria (PI. VII) and Bicorjne (PL VIII) ; these conditions must be
regarded rather in the light of an arrest or retardation of the development which in the vast
majority of cases results in the symmetrical disposition of the parts radially round a common
axis.

II. Morphology of the Trophosome.

1 . Ilydrantlis.

Two distinct portions enter into the composition of the trophosome, namely, the hydranths
and the hydrophyton.

The proper nutritive zooids (woodcut, fig. 2, a, h) which constitute the essential part of the
trophosome of the Hydroida have been usually known in common with the zooids of the Actinozoa,
or proper coral-animals, by the name of " polypes." It will be more convenient, however, to restrict
this term to the Actinozoa, to which Reaumur originally applied it, borrowing the word from
Aristotle, who used it, not for coelenterate animals at all, but for the cuttle-fishes ; while for
the alimentary zooids of the Hydroida I shall in the present work employ the term litjdranth}

The hydranth consists essentially of a digestive sac opening at one end by a mouth. Behind
the mouth are situated tubular offsets from the sac ; these are known as tentacula. The mouth
itself is borne on the summit of a more or less distinctly developed proboscis-like extension of the
sac, and to this the name of hypostome may be given.

The hydranth is developed on some part of the hydrophyton, which usually forms for it a
hollow, stem-like support {d), on whose summit or sides it is carried, and with whose cavity its
own is directly continuous.

The form of the hydranth varies. It may be flask-shaped, or fusiform, or nearly cylindrical ;
but is in every case more or less mutable, constantly changing its shape as the result of various
states of contraction. The hypostome may pass imperceptibly into the body of the hydranth
{Hydra, Perigommus), or it may be strongly difierentiated from it, and it then usually presents a
trumpet-like form, though frequently acquiring at the same time increased contractility and

^ Huxley uses in the same sense tlie term polypite, and in former publications of my own I fol-
lowed him in the employment of this word. As, however, a virtually identical term (polypide) is in
^ use for the exsertile and retractile portion of a polyzoou, an objection similar to that which renders
inexpedient the use of the word polype may be urged against the employment of polypite in the
terminology of the Hydroida.

THE TROPHOSOME.

25

mutability of outline {Eudendrium, Cawjmmdana, woodcuts figs. 2 and 13). The tentacles vary
ill form, number, and arrangement in the different species, and thus afford excellent characters
for the purposes of classification. Their axial tube may be continued throughout {Ilydrd),
or it may be more or less obliterated {Coryne, Campanularia). They are usually simple
filiform processes, as in Hydra, Clava, Hydradinia, Campanularia, &c. ; but in some cases their
extremity presents a knob-like enlargement, as in Co/y«(?, when they are said to be "capitate."
Their number may be as low as two {Lar), or it may amount to more than 100 {Corymorplia
nutans). They may be arranged in a single circlet {Campanularia)} or in two {Vorticlava), or
in several (Stauridiim), or they may be scattered over the body of the hydranth {Coryne), or may
be partly verticillate, partly scattered [Pennaria). Finally, they may be all of one kind (filiform,
as in Clava, capitate, as in Coryne), or the same hydranth may carry both filiform and capitate
tentacles {Stauridium) .

In a great number of hydroids the hydranth is destitute of any external protection
{Tubularia, Coryne, &c.). In others, however, there exists a cup-shaped receptacle, within which

Fig. 3.

Portion of the Hydrosoma of Calycella syrinya, showing the operculate hydrotheca;, -Vc.
a a The hydranth extended ; b, b, hvdranth retracted ; c, hydrotheca ; d, operculum open for the exit of the hydranth ;
<f ope'reulum closed after the retraction of the hydranth ; e, e, hydrocaulus ; /, hydrorhiza ; g. gonangium with its contents
previous to the escape of ovum ; g , uonangium after the passage of the ovum into an acrocyst ; h, walls of sporosac ;
r, gubernacular sac ; k, remains of blastostylc; m, ovum still contained within the sporosac ; u. acrocyst ; o, ovum after its
escape into the acrocyst; p, empty sporosac, still enclosing the spadis.

^ A single circlet may be absolutely simple or it may result from the close approximation of tuo
or more verticOlate series of tentacles.

26 MORPHOLOGY.

it may be partially or totally witlulrawn, and from whicli it may again spontaneously extend itself
{Campamdaria, Sertdaria, &c.). To this receptacle, whicli is characteristic of a large section of
the llydroida, the name of hjdrotheca has been given by Huxley.

The hydrotheca is formed of a chitinous membrane continuous below with the perisarc or
common chitinous investment of the hydrophyton, to be presently described. Its margin may be
quite even, or it may be divided into minute teeth (woodcut, fig. 2) ; the cup may be permanently
open (woodcut, fig. 2), or it may be provided with a kind of operculum formed of triangular
segments, united, as by a hinge, to the margin, and capable of completely closing over the cup
when thehydranth is retracted (woodcut, fig. 3). Sometimes the operculum consists of a flexible
membranous continuation of the cup-margin, and then, during the retracted state of the hydranth,
falls together in loose folds.

In a few hydroids {Bimeria, PL XII, fig. 1) the chitinous perisarc is continued from the
hydrophyton for a greater or less distance over the hydranth, which it closely invests without
forming anything like a free hydrotheca. In others (various species of BougainviUia, PI. IX
and X) the perisarc continued over the hydranth invests it so loosely that in extreme con-
traction the hydranth seems to be withdrawn into a cup ; but this apparent cup has nothing of
the permanent form or rigid texture of a true hydrotheca, and is always thrown into more or
less distinctly marked transverse folds or rugse by the contraction of the hydranth.

3. Ilijdroplnjion.

The term lijjdrophjton is used to designate the common basis by which the various zooids of
the hydrosoma, or general colony, are kept in union with one another.

Cmiosarc and Perisarc. — The hydrophyton consists mainly of a fleshy tubular basis, com-
posed, like all the zooids which it supports, of an ectoderm and an endoderm. I shall designate
this fleshy and only essential portion of the hydrophyton by the name of ccenosarc.

In every memljer of the Hydroida, however, with whose trophosome we are acquainted,
excepting only the fresh-water Hydra, and possibly also Nemopsis and Acaidis, whose trophosomes,
like that of Eijdra, are stated — though without sufficient evidence — to be free, the ectoderm
excretes from its outer surface an unorganised pellicle, chemically identical with chitine, and
forming an external tubular investment for the soft organised ectoderm." This unorganised
excretion, which must be placed in a totally different category from that of the ectoderm and
endoderm, I shall designate as the perisarc. In some cases it is confined to the hydrophyton ; in
others it extends, not only over the entire hydrophyton, but is continued for a greater or less extent,
and in a more or less modified form, over the various zooids of the colony. In the Sertularinoi
and Campanularinm it forms the cup-like receptacles or liydrothecce already descrilied, into which
the hydranths are retractile; as well as pecuhar receptacles — \ki& (jonangia (woodcuts, fig. 3,/, and
tig. 3^ fj) — destined for the protection of the sexual buds. It varies greatly in thickness, from a
tough investment, in which numerous layers of deposition can be detected, to a delicate, scarcely
recognisable pellicle, and is invariably absent from those zooids (fig. 3, g) which have detached
themselves from the colony in order to lead an independent life in the open sea. In the adult
Hydractinia and Podocoryne it presents the very exceptional condition of not only investing the

' If future observatious should confirm the view entertained by Agassiz, as to the hydroid nature
of the tabulate and rugose corals, wc shall then have examples of calcareous as well as chitinous
skeletons amone the Hydroida.

THE TROPHOSOME. 27

ectoderm, but being itself overlaid by a soft naked expansion of the cocuosarc, for wliicli it thus
forms an internal framework, recalling tlie sclerobasic corallum of certain Actinozoa.

Ilydrorkiza and Ili/droccmliis. — In almost every case the general colony, or hydrosoma, is
attached to some foreign body, such as rocks, shells of mollusca and Crustacea, sea-weeds, floating
timber, &c., to which it is fixed by some part of its siu'face (woodcuts, fig. 2, e, and fig. 3,/).
In many cases this is effected by a definite organ of attachment, as in the fresh-water
Hydra, where, by means of a disc-like expansion of the end diametrically opposite to the
mouth, the animal can attach itself to the stems and leaves of aquatic plants, from which it
can again spontaneously free itself; or, as in a great number of marine Hydroid.\, in
whose young state a disc occupying a similar position (PI. XIII, figs. 12 — 16) also
becomes an organ of fixation, differing, however, from the corresponding organ in Hydra
by its not admitting of spontaneous detachment, and by its being usually replaced, as
the animal grows older, by adherent tubular offsets, or stolons, given off from the same
part, lor the definite organ of fixation the term hydrorhiza, as suggested by Huxley, may be
employed ; while for the whole of that portion of the hydrophyton which intervenes between the
hydrorhiza and the hydranth (woodcuts, fig. 2, d, and fig. 3, e) it will also be very useful,
especially in descriptive zoology, to have a distinct name, and that of hydrocauliis may therefore
be conveniently used to designate it.

In many cases, however, all trace of a definite hydrorhiza disappears as the animal grows
old, and continues to comphcate itself by the formation of new buds and branches ; and we then
find fixation effected by some part of the general surface of the hydrosoma, as in certain creeping
Campanularians, &c., in which a greater or less extent of the hydrocaulus itself becomes the
medium of attachment. Sometimes it is the hydrocaulus which is suppressed, and the hydranth
will then be sessile on the hydrorhiza, as in Hydradima (PI. XV).

It is occasionally very uncertain whether the part which fixes the hydrosoma ought to be
regarded as a true hydrorhiza or as an adherent hydrocaulus. Most usually, however, some
peculiarity of structure or of form will justify a decision. Thus, in the Campanidaria represented
in woodcut, fig. 2, the tendency of the adherent portion to form a network of inosculating
branches, so very different from anything exhibited by the free stems, will fully entitle us to
regard this adherent network as a true hydrorhiza, and to place it in a category distinct from that
of the free hydrocaulus.

Again, it is by no means always easy to say where the hydranth ends and where the hydro-
caulus begins. In by far the majority of cases the distinction is easy enough, as in the Avhole of
the calyptoblastic hydroids, and in Tahidaria, Corymorpha, and many others among the gymno-
blastic genera, in all of which the line of demarcation is indicated by a marked change of form,
and frequently of structure. In some other cases among the Gymnoblastea, however, the hydranth
passes so imperceptibly into the hydrophyton that it is difBcidt to say how much we ought to give
to the one, and how much to the other.

The limit of the perisarc, or common chitinous investment, will often help us in this.
Thus in Clava (PI. I), where the hydranth possesses a very much elongated form, one might easily
be led to regard as hydrocaulus what is really part of the hydranth. Here, however, if wc
consider the whole of the naked portion of the trophosome as belonging to the hydranths,
we shall have a distinct though rudimental hydrocaulus in the very short, narrow tubes,
invested by a perisarc, which arise from the upper surface of the hydrorhiza, but which have

28 MORPHOLOGY.

usually been overlooked in the description of this genus. It must, however, be admitted that
there are cases among the gymnoblastic hydroids in wliich the boundary can scarcely be re-
garded as otherwise than arbitrary.

Besides the fresh-water Hydra, two cases have been described in which the entire hydro-
soma occm-s in a free state. This very exceptional condition is stated to exist in two North
American genera, Nemopsis and Acaulis, whose trophosomes have been found floating free in the
open sea. Acnulis, however, is described as becoming attached at a later ])eriod of its life.^

In most cases the hydrophyton becomes developed into a ramified, tree-like growth {Euden-
drium ramosum, Houyainvillia ramosa, Laomedea dicJiotoma, &c.). In other cases ic consists of a
creeping adherent, usually ramified stolon, with simple free tubes sent off fi'om it at inter-
vals (certain species of Clava, Campanularla, &c.) ; while sometimes, as in Hydract'mia and
Fodocoryne, it forms a continuous stratum, spreading over the surface of some foreign body which
the hydroid has selected for its al)ode.

Nematop/iores. — In hydroids belonging to the family of Fhnimlnridm certain very remark-
able appendages are developed at definite and constant points from the hydrophyton. They have
been named nematopliores by Busk \" and though the part they perform in the economy of the
hydroid cannot be regarded as strictly determined, they may be here described more appropriately
than anywhere else. They consist each of a cup-like receptacle containing a sarcode mass, which
can extend itself from the cup in the form of simple or branching processes, and again completely
withdraw itself, so as in every respect to resemble the pseudopodial prolongations of a rhizopod.
There is usually a cluster of thread-cells immersed in the sarcode.^

The receptacle is formed of a chitinous membrane like that of the hydrothccac, and nuiy
either consist of a single chamber {AglaopUenia phtma) or its cavity may be divided by a
transverse diaphragm into a distal and a proximal chamber, which freely communicate with one
another through an orifice in the diapjiragm {Antemiularia).

The nematophores may be attached to the hydrosoma by a broad base, or Idc adnate to it for
a greater or less extent {AylaopJienia), or they may taper away to a fine point of attachment at
their proximal ends {Antennularia).

In Aylttophenia jjhima there is a nematophore always situated immediately in front of each
hydrotheca. It is here adnate to the walls of the hydrotheca, with whose cavity that of its own
receptacle communicates by means of a common lateral aperture, through which the sarcode
prolongations of the nematophore can freely pass into the interior of the hydrotheca.

In the singular Campanularian genus OpJiiodes Mr. Ilincks has described certain appendages
which take the place of the nematophores in the nematophore-bearing genera.* They consist of
very extensile tentacula-like bodies, which are carried both upon the hydrocaulus and the
hydrorhiza. They terminate at their distal extremity in a spherical capitulum loaded with thread-

' See M'Crady's account of Nemopsis, in Proc. Elliot, ' Soc. of Nat. Hist.' vol. i; and Stimpson
on Acaulis, in his 'Fauna of Grand INIanoii,' published in the Smithsouiaa Contributions. These
floating tropliosomes, however, arc probably only the detached hydranths of fixed forms, while the state-
ment that Acaulis becomes subsequently attached is almost certainly founded on an error.

" Busk, ' Hunteriau Lectures' (MS.), delivered at the Royal College of Surgeous, London, 1857,
' Figures of nematophores are given below in the section which treats of the physiology of the

HVDROIDA.

■' T. liiucks on Ophiodes, in ' Auu. Nat. Hist.' for Nov., 186G.

THE GONOSOME. 29

cells and arc swnouiidecl at their base by a cliitinous sheath. During tlie life of tlie liydroid tlicy
may be seen to be in a state of great activity, stretching themselves out and twisting about in
every direction.

III. MoRVnOI.OGT OF THE GONOSOME.

1. General view of the Gonosome.

The zooids which compose the gonosome may remain permanently attached to the rest of the
hydrosome, or they may become free and lead henceforth an indc'ijcndent existence in the open
sea. It will be found very convenient to have a common term to express all these free zooids of
the gonosome, and I accordingly propose for them the name of planohlast} With one rare
exception, they are all in the form of the so-called gymnophthalniic medusas.

Under the planoblasts, however, must be included certain comparatively rare instances in
which the medusiform zooid, though having its natatory organ well developed, remains, from
some unknown cause, attached to the trophosome, and attains to sexual maturity without ever
actually becoming free. It is capable, however, when accidentally detached, of swimming by the
systole and diastole of a true natatory umbrella, and cannot, therefore, be placed in a different
category from that of the essentially free planoblast.

The gonophore is the uHimcde (generative zooid, — that on which devolves the duty of giving
immediate origin to the generative elements which are always produced from it between its own
ectoderm and endoderm without the intervention of any other zooid. It is the only essential
part of the gonosome, being never absent. It may remain attached during its whole lifetime, or
it may sooner or later separate itself from the rest of the hydrosome, and thus become a free
generative bud or planoblast. It presents either the form of a " gymnophthalmic medusa" or
else that of a zooid in which the form of the medusa is more or less disguised, or its parts
more or less suppressed, but which can nevertheless be always referred by an easy comparison to
the essential type of the medusa.

The planoblasts may either be the direct producers of the generative elements, and are then
true gonophores, or may never produce the generative elements directly, but only through the
intervention of another bud which is developed from them. For this latter form of planoblast,
which is thus, strictly speaking, non-sexual, and which, notwithstanding its resemblance to a
gonophore, should be carefully distinguished from it, I have given the name of blastocheme ; while
the proper sexual medusiform planoblast may be designated by the term ^onoc/ieme.

Among the fixed zooids of the gonosome there is very frequently one which, like the blasto-
cheme, takes only a subordinate part in the generative function, being, like it, destined for the pro-
duction of other generative buds which are developed from some part of its sm-face, and which
may be either gonophores or blastochemes. It is never medusiform, but may be regarded as a
peculiarly modified hydranth, having its alimentary function suppressed. It is referred to in the
present work under the name of hlastostyle.

The gonophore is always borne as a bud, either directly upon some part of the trophosome,
or upon a blastostyle, or upon a blastocheme. While it is constructed essentially on the plan of
a gymnophthalmic medusa, it varies greatly in the degree of completeness in which this plan is
expressed in it. It may be referred to one or other of two principal types, based respectively on

' " Waudering buds " — TrXavo^ai and /BAaaroc.

30

MORPHOLOGY.

Fig. 4.

the greater or less approach to the completely formed medusa. The peculiar conditioQ by which
one of these types is characterised may be iexmeA. pJuinerocodoHic, while that which distinguishes
the other may be designated as adelocodonic — conditions, how-
ever, which, it must be liorne in mind, jjass into one another by
numerous gradations.

The phanerocodonic condition is found in those gonophores
(PI. X, &c., and woodcut, fig. S) which present an obvious
medusa-form, and which are distinguished by haviqg a well-
developed umbrella, provided with the wide aperture, or codono-
donie, which characterises the complete medusa ; the umbrella,
except in one remarkable form — that presented by Clavatella,
Hincks (woodcut, fig. 5), and EhutUeria, Quatrefages — being
eminently contractile, and fitted for natation. The adelocodonic
condition is found in the bodies to which I have elsewhere given
the name of sporosac ; these bodies (PI. I, &c., and woodcuts, fig. 4<,
c, and fig. 7) have either no lunbrella, or, if this be present, it is in
an incompletely developed state, never provided with a wide, open
Group of zooids from a colony of codonostomc, and qultc incapable of acting as a locomotive organ.

Hudraetinia echinata^ taken from near mi i i • i • i r i_i j_

the margin of the colony. The phauerocodonic gonopliores, in by tar the greater num-

a. alimentary hvdrantli : hh, blasto- , p - . i i. i ^i i r ^i i i iv

style; c, gonophores which h.we been bcr 01 mstances, dctach thcmselves iroiu the hydrosoma after

^nc'if'fined°«tth^™\frf'ipf,i?C: they have attained a certain degree of maturity, and lead

of'tbe' lTon?.*Vhr'a]iL*ntarrTid ^enceforth an independent e.xistence, during which they increase in

spiral hydran'ths are ™n"ected to one J fj dcveloD ucw parts, and sooucr or later give origin to ova

another and to the blastostyle by a ' r I ' o o

common basal expansion or ccenosarc, q« SDCrmatOZOa

In some cases, however, they develope and discharge their reproductive elements while still
attached, and then wither away, without ever becoming free, notwithstanding their well-developed
contractile umbrella apparently fitting them for an independent natatory existence. If an
observation of Agassiz be not really made on two different species instead of one, as he himself
believes, it would seem that this condition is dependent, in one case at least, on the season of the
year ; for he informs us tliat he found the gonophores of Coryne mirabilis, Agass., in the earlier
months of the year, detach themselves from the trophosome and swim away as gymnophthalmic
medusse before the development in them of ova or spermatozoa ; while, somewhat later, he has
seen the gonophores attain to sexual maturity, without ever becoming free.^ It is possible,
however, that the two conditions here described by Agassiz belong to two quite diflerent species.

The free phanerocodonic gonophore is in one rare form (woodcut, fig. 5) ambulatory : in all
others it is natatory ; locomotion being effected by alternate systole and diastole of the umbrella.
In the ambulatory form the umbrella is incapable of evident systolic and diastohc movements, and
locomotion is performed by marginal tentacles peculiarly modified for creeping over solid bodies.
This veiy exceptional form has been met with only in Clavatella, whose trophosome has been

^ Agassiz, ' Contributions to tlie Nat. Hist, of the United States,' vol. iv, p. 1S9.

^ The gonophores of the siphonophorous group Cali/cophoridtB properly come under the designa-
tion of phanerocodonic, though they may never become free, and though we find them departing from
the typical form of the gynmopthalmic medusie by the non-development of the marginal appendages
of the umbrella.

THE GONOSOME.

31

discovered by Hincks, and in the nearly allied Eleutheria of De Qnatrefages, whose trophosome
has not yet been detected.

While all the leading features of a gymnophthalmic medusa are thus at once obvious in the
phanerocodonic gonophore, the adelocodonic gonophores, on the other hand, present the niedusoid
structure only in a disguised or undeveloped condition. They have the form of sacs, and, except
in a single known instance, the whole gonophore remains permanently attached to the hydro-
soma, giving rise within it to the generative elements, which, after attaining a certain degree of
maturity, are ultimately discharged from its cavity. The single exception is afforded by the

Fig. G.

Flo 5

Ambulatory Medusa of Claiaiella pioJ /era

Free locomotive eporosac of Dicoryne. The figure
represents two aspects of the female sporosac when
swimming in the open sea, and viewed in planes at right
angles to one another.

genus Dicoryne, Allm. (woodcut, fig. C), in which, before discharging its generative products, the
gonophore liberates itself from its external investment or ectotheca, and thus, becoming a free
planoblast, swims about actively by the aid of vibratile cilia.

The following tabular view will exhibit at a glance the various parts which may occur in a
hydroid gonosome :

f f- Adelocodonic or

I sacciform
[Sporosacs)

2 -i

Zooids directly sexual
[Gonophores]

Zooids not directly sexual,
but giving origin to
sexual buds.

1 fixed.
I natatory.

"1

L

Phanerocodonic or
medusiform
L (Gonockeines).

f Medusiform

[Blastochemes]

I Styliform

L [Blastosiyles]

natatory,
ambulator}'.

natatory.

fixed.

!■ PlaiKjIjIasts.

Whether the gonosome remains during its whole lifetime connected with the trophosome, or
detaches itself as one or more independent zooids, it is manifest that it constitutes an essential

32

MORPHOLOGY.

element in the character of the species, and tlie study of no one species of hydroid can be regarded
as complete unless it embrace both trophosome and gonosome. Since, however, in many cases we
are acquainted with only the free gonophore, or the free blastocheme, not having yet discovered
the trophosome to which it belongs, while in other cases the trophosome alone is known
to us, we have been in the habit of treating such instances without regard to the missing
zooids, and as if they afforded examples of independent species ; but it must never be forgotten
that the data on which we thus assign to them the rank of determinate species, or genera,
are insufficient for the purposes of a philosophic system : such genera and species must be
regarded as purely provisional ; for the zoologist is no more justified in accepting such incom-
plete characters as sufficient for the accurate determination of his hydroid, than would the
botanist be in regarding the flower alone on the one hand, or the root, stem, and leaves alone, on
the other, as affording characters sufficient for the definition of any flowering plant whose exact
determination he woidd attempt.

The gonosome is that part of the hydroid which presents the most marked variation among
the different members of the group, and it is here that we find most distinctly manifested those
beautiful gradations of form which, by throwing light not only on one another, but on forms
belonging to other groups of the Hi/drozoa, possess for the morphological student a profound
significance. It will therefore be necessary to enter into a more detailed examination of the
gonosome in its various modifications aniono; the Hydroida.

Fig. 7.

Adelocodonic gonophore of llydract-
nia echinata.

a, Spadix surrounded by the ova, the
whole enveloped by the perigonium,
which here consists of J, endotheca, and
c, ectotheca, the mesotheca being absent ;
rf, communication of the cavity of the
spadix with that of the blastostyle.

The Adelocodonic Gonophore {Sporosac).

The adelocodonic gonophore (woodcut, fig. 7) is a sac-
like body, which presents the following parts :

1. An internal hollow process, the spadix [a], which occu-
pies the axis, and whose cavity is in communication with the
general somatic cavity of the hydrosome.

2. The generative elements, ova or spermatozoa, which are
developed round the spadix.

3. A sac [peric/onium) which surrounds the generative
elements, and retains them in their place. This sac may be simple
or its walls may present two or even three layers, in which last
case it will consist of (1) an internal layer [endotheca], which lies
immediately on the generative elements; (2) a middle layer
{mesotheca) ; nwA. {i) a.\\ external layer {ecfotheca). The meso-
theca, moreover, may have a system of canals developed in it.

Of the adelocodonic gonophore we have examples in
such forms as the so-called generative sacs of Clava, Ilydractinia,
Eudendrium, &c.

3. The Phanerocodonic Gonophore {Gonoclieme. Sexual Medusa).
The phanerocodonic gonophore (woodcuts, figs. S and 17) is bell-shaped, and in its most
developed form presents the following parts :

1. A central tubular body, the manubrium [a), which carries a mouth (/) at one extremity.

2. The generative elements [b), which are developed between the inner and outer membranes
(endoderiu and ectoderm), which compose the walls of the manubrium.

THE GONOSOME.

.33

3. A coiitrndilc hell, in/ihrcUa (fig. 8, c), from tlie summit f)f wliosc concavity the maiiulirium is
suspended, and in whose walls is alvvaj'S developed a system of canals (^antrovancidar), consisting of
— 1. A set of c(|uiiiistant longitudinal canals (dd), mostly
four in numhcr, which radiate from the base of the
manubrinni, into which or into a special cavity {atrium,
fig. 17) which often exists at its base — they open at their
origin ; and 2, of a circular canal (e) which surrounds
the codonosloiiie, or orifice of the bell, and receives the
distal extremities of the radiating canals.

4. Contractile tentacida {/), which spring from the
margin of the umbrella, and often carry at their bases
definite accumulations of pigment-granules (fig. S,y, and
fig. 17, c) named ocelli.

5. A membranous extension, velum [It), of the margin
of the umbrella over the codonostome where it forms a
thin, muscular diaphragm, perforated in the centre by a
circular opening of greater or less diameter.

The body composed of the various parts now
enumerated constitutes one form of the so-called " gym-
nophthalmic medusa,"^ and in its young state is usually
invested by an external protective membrane, the ecto-
theca, the homologue of the external layer of the peri-
goniutu in the sporosac.

Of the form of medusa presented by the phanero-
codonic gonophore we have examples in the types
described by authors under the names of Sarsia,
Steenstrupia, Oceania, &c."

4. The BladoityJe

The blastostyle, as has been already said, must be
regarded as a hydranth whose alimentary functions have
become suppressed, and which, though not, properly speak-
ing, sexual itself, is entirely destined for the production of
sexual biuls or gonophores, either directly or through the
medium of a non-sexual bud, the blastochcme. A good
idea of this form of zooid may be obtained from the so-
called "fertile polypes" oi Hydractinia echinata (PI. XV,
and woodcut, fig. 4, (5 1}) . These are bodies of a cylindi-ical
shape, which are scattered among the alimentary hydranths,
and which in all respects they resemble except in the

Oceania coronata, AUm. (provisiomUly),;i medusa
of unknown trophosouie, as an example of a plia-
nerocudouic gonophore.

a, Manubrium ; i, generative elements (ova)
developed between endodenu and ectoJern of lua-
nnbriuni ; c, umbrella ; c', peculiar development
of the solid tissue of the umbrella, which occurs in
the present species ; dd, radiating gastrovascular
canals ; e, circular gastrovascidar canal ; f, mar-
ginal tentacles ; f, rudimental marginal tentacles ;
^, ocellus ; i, mouth surrounded by its four fim-
briated lips.

1 Another form is presented by the blastocheme described below.

- The structure above described is that of the phancrocodonic gonophore in its most completely

5

34 MORPHOLOGY.

fact that the tentacles are entirely suppressed, their place being taken by small clusters of thread-
cells, and that the mouth, if not wholly obliterated, is reduced to a very minute perforation,
which probably never subserves the function of ingestion. Near the distal extremity of these
bodies the gonophores (c) are borne as a dense cluster of buds.

In the whole of the Campanularian and Sertularian hydroids the biastostyle, with its buds,
is enclosed in an external chitinous capsule, the gonangium (woodcut, fig. 2,/), which is never
present in the Tubularince. The gonangium is of very definite form for each species, and affords
good characters for diagnosis.

Though it is necessary to distinguish the blastostyles from the hydranths, it cannot be over-
looked that they may pass into them by certain transitions. Agassiz^ describes a mouth in the
blastostyles of the Hgdractinia polycUna of the North American coast, but as the tentacles are
entirely suppressed, it is doubtful whether the orifice which here exists can be regarded as
destined for the ingestion of nutriment. In certain Eudendria the hydranths which carry the
gonophores grouped round their base present a perfectly developed form while the gonophores are
young; but as these continue to grow, the hydranths which carry them frequently become
atrophied, losing their tentacles and mouth ; and by the time the gonophores have attained to
maturity the hydranths have assumed the condition of blastostyles. Again, among the Sertu-
larince. we find in Halecium kalecinum (woodcut, fig. 29) the female biastostyle developing from
its summit a pair of perfect hydranths with tentacles and mouth, and with their digestive cavity in
communication wuth that of the biastostyle; but I know of no more instructive demonstration of
the relation between biastostyle and hydranth than what is afforded by the female gonangium
and its contents in Sertidaria rosacea, S. tamarisca, and S. falax. In these hydroids
(woodcuts, fig. 23 — 26) the biastostyle develops from its summit a set of peculiarly formed
tentacles, which, after becoming invested with a perisarc, continuous with the chitinous walls
of the gonangium arch, over the summit of the gonangium, so as to form the walls of a special
chamber, which constitutes a marsupial receptacle, iu which the ova, after their discharge from
the gonangium proper, may undergo further development. These, however, are all exceptional
cases, and do not render less valid the association of the biastostyle with the gonosome rather
than with the trophosome, while they are impoi'tant as showing the homological identity between
the hydranth and the biastostyle.

Notwithstanding the transitions which may be thus traced between the hydranth and the
biastostyle, we must carefully avoid the confounding of a true biastostyle, whose characteristic
form and suppression of nutritive function show themselves before the appearance of the generative
buds, and those pseudo-blastostyles which are caused by the exhaustive action of the generative
buds on an ordinary hydranth.

developed form. Such complete differentiation, however, is not always attained even in the Hydroida,
while among the Siphonophora a hydrozoal group possessing the closest relations with the Hydroida,
the margin of the gonocalyx or umbrella of tlie medusiform gonophore in the Cahjcophoridit carries
neither tentacles, ocelli, nor lithocysts, and the manubrium develops, at least usually, no mouth upon
its extremity.

^ Agassiz, ' Contrib. to the Nat. Hist, of the United States,' vol. iv, p. 230.

THE GONOSOME.

35

5. The Blasfocheme {non-sexual Medusa).

It has been already stated that the blastocheme presents, hke the phanerodoconic gonophore,
the form of a completely developed gyranophthalmic medusa. It differs, however, from the gono-
phore, not only in never producing generative elements directly, and in being thus, properly

Fig. 9.

Fig. 10.

Sexual zooid (sporosac budding from a radiating canal
in the blastocheme oC Obelia geniciilnta.

o, Portion of the umbrella, and b. radiating canal of
Jledusa of Campanularia the blastocheme ; c, spadix of sporosac ; d, perigouium

Johjtsioiii, shortly after libera- consisting here of a single membrane ; e, ovum with

tion from the gonangium, il- germinal vesicle and spot ; f^ ovum with numerous

lustrating the peculiarities of germinal spots in the germinal vesicle,

the blastocheme.

a, liithocyst; b, incipient
tentacle; y, incipient sporosac,
formed as a bud upon the ra-
diating canal.

speaking, non-sexual, but also in certain points of structure ; for it is almost universally charac-
terised by the absence of ocelli, and by the presence of peculiar capsules called lit/ioci/sfs, which are
attached to the margin of the umbrella, and enclose one or more transparent refractile corpuscles.'

' An exceptional condition ia this respect is presented by a few medusje referable to the type of
the blastocheme. Thaumantias, as limited by Gegenbaur, has ocelli instead of lithocysts, and the
same is the case, according to Agassiz, in Staurophora, Brandt, and in Laodicea, Lesson ; while in
Melicertum, Oken, there are neither lithocysts nor ocelli.

In Tiaropsis diademata, Agas., a well-defined pigment spot has been described by Agassiz as existing
in the base of the lythocyst, a statement which I can confirm by my own observations on a species of
Tiaropsis captured in the Firth of Forth. As will be afterwards seen, however, I do not regard the
pigment spot of Tiaropsis as representing a true ocellus. According to Strethill Wright, a medusa,
which has been described by him under the name of Goodsirea {' Edin. N. Phil. Journ.,' July, 1859J,
is a true gonocheme, and yet it has lithocysts instead of ocelli. As to Oceania octona, Fleming, and
O. turrita, Forbes — raedusse belonging to the type of the gonocheme — the statement of Forbes,
that thev have a lithocvst imbedded in the base of the tentacle, is founded on an error of interpretation

36 MORPHOLOGY.

AVhilc it is itself noii-scxual, the blastoclicme always gives origin to special sexual biulsj or
gono])hores, which arc borne upon some part of the radiating canals.

As characteristic examples of the blastocheme, we may adduce the planoblasts of Campanularia
■Johnstoni (woodcuts, fig. 2, g, and fig. 9) — which are medusiform zooids referable to the deep-
belled section of Gegenbaur's genus Eucope — and those oi Laomedea dicliotoma and L.genicnlata
(woodcut, fig. 10) of authors — medusaj referable to Peron's type of Ohelia. In none of these
are sexual elements over directly developed, but instead of the direct formation of ova and
spermatozoa, there is produced a new zooid, which no longer presents the complete medusal
type, but is formed upon the plan of the adelocodonic gonophore. This zooid (woodcuts,
fig. 9, (/, and fig. 10) springs as a bud from the radiating canals of the medusa, and is
constructed upon precisely the same plan as that which we meet with in the gonophore of
Clava or Ilydracliivia, except that the perigonium would seem to be simple. It has an axile
spadix (woodcut, fig. 10, c), whose cavity is in direct comnumication with that of the radiating
canal {h) from which it springs. Immediately investing the spadix are the generative elements
(c,/) ova or spermatozoa; while these are themselves surrounded and confined by a true
perigonium [d^, which becomes at last ruptured for the liberation of its contents.

The zooidal nature of these buds is nowhere more distinct than in the genus JgJatira, Per.,
a form not yet traced to a polypoid trophosome. Here the generative elements are produced in
eight sac-like processes which surround the base of the manubrium, which is itself borne on the
extremity of a stalk dependent from the summit of the umbrella. These sacs are undoubtedly
true buds, and are entirely homologous with the gonophores of Clava ; and it is plain that they
are developed from the proximal extremities of the radiating canals, just where these canals pass
off' from the manubrium in order to run along the sides of its stalk before reaching the umbrella.^

In some blastochemes the sexual bud extends over a greater length of the radiating canal,
and presents, in consequence, a less defined and individualised appearance than in the instances
just mentioned, so as to lead one at first to hesitate as to the propriety of regarding it as a true
zooid. Such, for example, is its ciiaracter in the medusaj referable to the types of Thatnnantias,
Tima, and Melicerta, in all of which, while the generative buds are situated as in Obelia, on the
radiating canals, they occupy Mith their extended base so much of the canal as to be readily
mistaken for mere organs — ovaries or spermaries. Notwithstanding this, however, they are con-
structed upon essentially the same i)lan as the others, and offer no exception to the view here
taken.

In Tima, indeed (woodcut, fig. 11), we have an extreme case of this extension of the base of the
generative buds, which here present themselves in the form of four long, flattened, sinuous frill-
like bands, each attached by one edge along the whole length of a radiating canal. When a section
is made from the free to the attached edge of this band (woodcut, fig. 12), the generative elements

vvliile the view of the structure of Slabheria, Forbes, which would make this medusii to be a blasto-
cheme, having ocelli instead of litliocysts, is also, as will be shown below, based on an error. Though
it is thus very rare to find a blastocheme without litliocysts, the absence of ocelli in the gonoclicme
is quite common.

' See Leuckart's description of Atjlaura Peronii (' Wiegmaun's Archiv,' 185G, Erster Band, S.
10). Tjeuckait recognises in the generative sacs oi Aylaura the significance of true zooids, though he
refrains from extending this view to the generative sacs of other medusse.

THE GONOSOME.

37

are seen to be disposed upon each side of a hollow longitudinal septum {c). Tiiis septum consists
of a diverticulum of the cndoderm of the radiating canal ; it admits into its interior the fluid which
circulates in the radiating canal, and is plainly homologous with a laterally extended and flattened
spadix ; while the generative elements are externally confined by an ectodermal covering, which
is in the same way the homologue of the perigonium in an ordinary sporosac, but here flattened
out like the spadix, in accordance with the ribbon-shaped form of the gono[)liore.

Fig. 11.

Fig. 12.

Portion of the ribbon-shaped sporosac, much en-
larged in a female specimen of Tima Baird'd,

a, a, a, radiating canal ; b, b, b, sporosac ; c, ectoderm;
d, endoderm; e, cavity of the spadix ; e', e', distal edge of
the spadix, seen through the ectodermal layer ; /, ova.

Medusa ( Tima Bairdii), a blastocheme of unknown trophosome,
showing the convoluted and ribbon-shaped sporosacs along the
course of the radiating canals.

The blastocheme is thus essentially a free non-sexual medusa, which, like the blastostyle,
gives origin to sexual buds, but wiiich, unlike the blastostyle, is endowed with the locomotive
powers of a medusa, so that it carries those buds from place to place by the contractions of an
umbrella.

In the account here given of the blastocheme, I have confined this term to such medusae as
develop distinct sexual buds upon the gastro-vascular canals. In some of those medusae,
however, in which the sexual elements are produced in the walls of the manubrium, we find the
portions of the walls w-hich give origin to the ova or spermatozoa more or less difl'ercntiated from
the general walls of the manubrium, and presenting a lobulated appearance, which might easily
lead to the belief that the manubrium emitted from its sides true sexual buds. I am not,
however, prepared to place any of these cases in the same category with the blastocheme.

38 MORPHOLOGY.

The condition alluded to is especially well marked in the medusa of Cladonewa and of
Bou(/ainvillia, and in certain forms of Oceania, in the sense in which this group has been
restricted hy Forbes and Gegenbaur, and generally accepted by the German zoologists ; and an
opportunity of studying an undescribed medusa of this type (woodcut, fig. 8), which I obtained
abundantly in the autumn of 1865, on the west coast of Scotland, has plainly shown that the
generative lobes of the manubrium cannot be regarded as true zooids. The generative elements
are here simply produced between the endodcrm and ectodenn of the manubrium, and the lobes
are nothing more than a puckered or sacculated condition of the walls in those parts where the
ova or spermatozoa originate. I do not, however, deny the possibility of the manubrium as well
as of the gastrovascular canals giving origin to true buds, to which the development of the
sexual elements may be confined ; when this has been shown to be the case, the medusa
presenting it must take its place among the blastochemes.

Whether the medusae referable to the type of Geryonia and its allied forms ought to be
regarded as blastochemes is as yet uncertain, though my own opinion is in favour of so viewing
them. The parts which in these medusae give origin to the generative elements have been
described as leaf-like expansions of the radiating canals. Haeckel's observations have led him to
deny to them the significance of true zooids,^ while he sees in them nothing more than mere
lateral expansions of the canal, a portion of whose epithelium becomes here difi'erentiated into
ova or spermatozoa. This question as regards the Geryonida is, as we shall afterwards see, one
of considerable importance. Having had no opportunity of examining for myself specimens of
these medusae, I can bring no direct personal observations to bear upon it ; but the account
given by Haeckel, who has so admirably worked out the structure and history of the Geryonida,
does not appear, as I shall afterwards show, to necessitate the conclusion to which he arrives,
or to be inconsistent with the zooidal nature of the reproductive pouches. Further observations,
however, instituted for the express determination of this point will be needed before we can regard
the question as thoroughly cleared up.

6. Homoloyical parallelism between Sjjorosac and Medusa.

While it will be found very convenient to insist upon the differences pointed out above
between the adelocodonic gonophore on the one hand, and the phanerocodonic gonophore and
blastocheme on the other, it must not be supposed that these two forms are constructed upon
plans widely different from one another. We find, on the contrary, that the most exact
parallelism admits of being demonstrated between them ; for though they may at first sight
appear very different, it can nevertheless be shown that the closed generative sac of a Chiva
or a Hydractinin is an easily understood modification of a medusa."

^ Haeckel, ' Die Famille der Riisselquellen,' Vorwort, viii.

" It is now many years since I endeavoured to demon'strate that the so-called " ovarian vesicles"
of the Tubularince, and the fi.xed sacs contained within the gonangium of the Seriularina and Campanu-
larinm, are in all cases strictly homologous with the free medusa, that they possess a true medusal
structure in a more or less degraded or disguised condition. (" On the Anatomy and Physiology of
CordylopJiora," ' Phil. Trans.,' June, 1853.)

THE GONOSOME.

39

In comparing the two classes of bodies with the view of determining their homological
relations, their composition out of the two membranes ectoderm and endoderm must be carefully
kept in mind.

Fig. 13.

1_...

Diagrammatic sections ui' (ilianerocodonic and adelocodumc gouu[)liores.

A, Phauerocodonic, and B, adelocodonic gonophore.

a, Ectotheca; b, mesotheca or umbrella; c, endotheca, or ectodermal layer of manu-
brum ; d, spadix, or axile tube of manubrium ; e, ova ; f, radiating gastrovascular canals ;
g, circular gastrovascular canal seen in transverse section j h, marginal tentacle ; i, ocellus
in bulbous base of tentacle ; k, velum ; I, peduncle of gonophore ; m, general cavity of
ccenosarc ; «, mouth of medusa.

In both sections the endoderm is distinguished from the ectoderm by giving it a darker
shade.

Commencing with the central parts of a hydroid medusa, and comparing these with the central
parts of a sporosac, we shall find that in the medusa (woodcut, fig. 13, A) we have a manubrium
ill the form of a more or less elongated tubular body occupying its axis. The walls of the
manubrium are composed of two layers, an internal or endodermal layer [d] and an external or
ectodermal (c) ; and in all phauerocodonic gonophores or hydroid medusae of the sexual type, these
two layers become ultimately more or less separated from one another by the development of
the generative elements [e) between them.

In the sporosac or adelocodonic gonophore also (£) we have a double-walled tubular body,
between whose two walls the generative elements are developed exactly as in the medusa ; but
while in the medusa this body is in almost every case perforated by a terminal mouth, in the
sporosac it is completely closed, so that it assumes, by the increasing volume of the generative
elements, the appearance of a sac, filled with ova or spermatozoa, and having a csecal diverticulum
(spadix) plunged into the middle of the mass. This caecal diverticulum [cl) is plainly the
equivalent of the endodermal portion of the manubrium in the medusa, while the wall (perigonium)
of the sac represents more or less completely the structures which in the medusa lie external to
the generative elements.

When the perigonium presents its highest degree of development it consists as we have
already seen of three layers. Of these the inner (c) (endotheca) is the equivalent of the ecto-
dermal layer of the manubrium in the medusa ; the middle layer {b) (mesotheca) corresponds to

40 MORPHOLOGY.

the umbrella, and like it may have a system of canals (/) more or less completely developed in
it, and may even present a rndimental codonostonie, while the most external layer («) (ectotheca)
corresponds to a similar external layer frequently present in the young medusa bud.

It would seem that in no case is a velum or its homologue developed in the adelocodonic
gonophore, even though the representative of the umbrella should possess as in Tuhularia a
rudimental codonostome, while the marginal tentacles of the medusa are, except in the "me-
conidium " (see below, p. 57), also without their representative in this form of gonophore ; for
the tentacula-Iike tubercles which crown the summit of the sporosac of some of the Tuhdarkla
{Tuhularia larynx, for instance, PI. XXI) are of an entirely different significance, being merely
processes of the ectotheca.

7. HoinoJo(/ical parallelism between Ilydra/dh and Medma.

While we have been thus enabled to ti'ace a close parallelism between the medusa and the
sporosac, another comparison of great interest in this inquiry suggests itself, that, namely,
between the medusa and the hydranth. Now there can be little difficulty in finding in the
distal portion of the hydranth the homologue of the manubrium of the medusa ■} but the equivalents
of the umbrella and gastrovascular canals of the medusa are not at first sight so obvious. I
believe, nevertheless, that these are not totally lan-epresented in the hydranth. It will be kept in
mind that the tentacula of the hydranth are merely tubular radiating prolongations of the digestive
cavity, though with the cavity of the tube usually more or less obliterated by the peculiar condition
of the endoderm, and that for some distance from their origin they are necessarily included in the
thickness of the body-walls of the hydranth, where they consist merely of radiating canals extending
through these walls and lined by a layer of endoderm. Now this included portion I regard as the
true representative of the radiating cauals of the medusa ; and if we were to imagine the ectoderm
of the hydranth in a Eudendrium or Campanularia to acquire unusual thickness in a zone cor-
responding in position to the roots of the tentacles, we should have a disc-like extension of the

^ Huxley (' Oceanic Hydrozoa') strongly insists on this relation, and is so impressed with the
closeness of the homology between manubrium and hydrautli, that he uses the same term, " polypite,"
for both.

Agassiz (op. cit., vol. iv, p. 326) has witnessed the very simple adelocodonic gonophore in male
specimens of his Rhizogeton fusiformis, instead of withering away after the discharge of its contents,
elongate itself, develop tentacles, and become transformed into a hydranth. I have myself, on one occa-
sion, seen an analogous phenomenon in the female gonophore of Cordylophora lacustris, in which, after
the discharge of the ova, the spadlx had become elongated through the ruptured chitinous investment of
the original gonophore, had developed an ectoderm, thrown out tentacles from its summit, and become
metamorphosed into an ordinary hydranth. In the case of Cordylophora the transformation is confined
to the spadix, while, according to Agassiz, the entire gonophore of Rhizogeton takes part in the meta-
morphosis.

I believe that in both cases the phenomenon is an abnormal one ; it certainly is so in Cordijlo-
phora, for, in the ordinary conditions to which this hydroid is exposed, no metamorphosis of the kind
takes place.

TIIH GONOSOME.

41

ectoderm traversed in a radiating direction l)y tubular extensions of the endoderin which lines
the body-cavity of the hydranth, and this disc would only need to become still further expanded in
order to show itself as an unmistakable umbrella, with radiating gastrovascular canals, whde the
hypostome or proboscidiform extension of the body, which in these genera advances far in front
of the base of the tentacles, would rescnd)le in all essential points the manubrium of the
medusa.

Now, the commencement of such an expansion is evident in the hydranth of many Campanu-
laridm, while in certain species, as Laomedmflexuosa and Campanidina acuminata, the ectoderm of
the body is actually extended as a thin disc for a considerable distance in the plane of the tentacles,
which acquire in consequence the appearance of being connected at their bases by an intervening
web (woodcut, fig. 14).

Ijongitudinal section through tlie liydrantli and hydrotbeca of Laomedea Jlexuosa, showing the u*eb-like membrane by
which the bases of the tentacles are connected to one another.

a. Body of hydranth; b, hypostome, carrying the mouth on its summit; r, c, intertentacular web; rf, hydrotbeca.

While the portion of the tentacles included in the thickness of the body-wall of the hydranth
will thus be the equivalent of the radiating canals of the medusa, their free portion is plaiidy
homologous with the free tentacles, which in the medusa hang from the margin of the umbrella
at the points corresponding to the entrance of the radiating into the circular canal, and which
must be regarded as strictly the continuation of the radiating canals beyond their apparent

G

42 MORPHOLOGY.

termination in the circular canal. The tentacles, which in many medusae spring from the
intervening spaces upon the margin of the umbrella, and arc therefore not directly continuous
with the radiating canals, make their appearance probably in all cases later than the others, and
arc frequently less developed. These inter-radial tentacles must be placed in the same category
with the lithocysts as simple marginal appendages, to be carefully distinguished from the primary
tentacles, and, like the lithocysts, have no representative in the hydranth.

It cannot be urged, as an argument against this view, that the circular canal of the medusa is
not represented in the hydranth ; for the absence of a developed umbrella in the hydranth neces-
sarily brings with it the absence of this canal ; and it is for the same reason that velum, lithocysts,
and secondary tentacles, are also absent. Neither can it be said that those cases in which
the tentacles of the hydranth are not arranged in a single verticil, but are repeated regularly or
irregularly in different planes upon the body, are inconsistent with the homological relations here
insisted on ; for such cases can be regarded only as special modifications of the more typical plan
which has directly suggested our comparison.

Huxley, believing the difference in structure and development between the locomotive disc
of the gymnophthalmic and that of the steganophthalmic medusae to be so great as to place them
in different categories, would confine the term " umbrella" to the disc of the stef/a/iophfJMlmafa,
and would designate that of the (/i/mjiophfhalwafa by the terms " nectocalyx" and " gonocalyx." I
was at first disj)osed to adopt the same view ; but an investigation of the mode in which this part
makes its appearance in the gymnophthalmic forms has convinced me that the development is
essentially the same in both cases, and that, notwithstanding some marked structural differences,
there is sufficient unity between the two to render it more convenient to speak of them under the
same term as strictly homologous organs. In both cases they are formed by an outgrowth of the
walls of the polypoid manubrium, and the fact that the steganophthalmic medusa is produced by
successive transverse divisions of a " scyphostoma," while the gynmophthalmic medusa is formed
as a lateral bud from a hydrosome, is no valid argument against this approximation ; for
every segment of the " scyphostoma" is strictly comparable to the bud of the hydroid, and
developes its umbrella by an outgrowth from its sides in quite the same way.

A very instructive example, which strikingly bears out the comparison I have here attempted
to make between the hydranth and the medusa, is afforded by the remarkable locomotive zooid
which forms the gonophore of Dicomjne. This little zooid (woodcut, fig. G) is essentially a free
medusa, reduced to the condition of an ova-bearing or spermatozoa-bearing manubrium, from whose
base two free tentacnla in form and relations like those of a hydranth are developed. Now, there
is here no umbrella, locomotion being affected by the action of cilia ; but it is evident that we have
only to imagine the ectoderm of the manubrium projected as a disc, in the way already supposed,
in the horizontal plane passing through the base of the two tentacles so as to include the basal
portion of these tentacles in its thickness, in order to have an umbrella with two radiating canals
added to the manubrium.

But development, as we shall afterwards sec, entirely coincides with anatomy in pointing to
the same conclusion ; and it is only necessary to trace the formation of the umbrella and radiating
canals in the budding medusa, in order to become convinced that their origin is essentially that
here insisted on) ; while the interesting observations of Johannes Midler on the development
oi yEginopsis, and of M'Crady on that of Cimina — observations which will be specially referred to
below — show tliat in these genera the umbrella grows out as a horizontal disc from the walls

THE GONOSOME

43

of a free polypoid manubrium, which bears a close resemblance to the generative zooid of
Dkoryne. It would appear, however, from the observations of Fritz lAlullcr,' and of llaeckcl,Hhat
in certain geryonidan medusa; the umbrella is formed by the excavation of a solid spherical embryo ;
but it must be noted that neither of these observers had seen their embryo mcdusa> at a period
anterior to the commencement of the developing umbrella.

The parallelism which T have thus endeavoured to demonstrate may be expressed in the
following scheme :

Phanerocodonic Gonophore and

Adelocouoxic

Blastocheme.

GoNOPUORE.

IIydrantu.

(Medusa.)

(Sporosac.)

Ectotheca ......

Ectotheca .

0

Umbrella ......

Mesotheca .

Web-like membrane unit-
ing the bases of the
tentacles in Laoniedea
flextwsa, &c.

Gastrovascular canals ....

Canals of Mesotheca .

Base of tentacles extend-
ing through the thick-
ness of the body walls,
and through the web-
like membrane.

Ectoderm of manubrium

Endotheca

Ectoderm of hypostome.

Endoderm of manubrium

Walls of spadix .

Endoderm of hypostome.

Manubrium .....

S padix + endotheca

Hypostome.

Primary or radial marginal tuntacula

Primary tentacula in the

Free portion of tenta-

mcconidium.

cula. [

Secondary or interradial marginal teuta-

Secondary tentacula in

0

cula.

the meconidiura.

Ocelli and lithocysts ....

0

0

Velum ......

0

0

Generative elements in gonoclieme .

Generative elements

0

8. Further modifications of the Gono)<oine.

Besides the great leading differences already described, many others of a more subordinate
kind are met with. The adelocodonic gonophore in particular exhibits many special modifications,
and presents us with a regular series of gradations in complexity, which throw much light on its
morphology.

The simplest form is jjrobably that which we meet with in the female gonophores of the
freshwater llifdra. Here there would seem to be no difl'ercutialion of an ectotheca. while the

Wiegra. Archiv,' 1859, p. 310.

Die Familie der Russelquallcn (Gcryonidfe),' 18GfJ.

44

MORPHOLOGY.

spadix itself remains in a rudimcntal condition, being scarcely elevated above the base of tlie
gonophore, whose whole cavity becomes at an early period occupied by the ovarian mass.

An advance over this condition is seen in the sexual bud which is borne by that form of
medusa described above, under the name of " blastocheme." Here we have the ultimate sexual
bud f[uite destitute of ectotheca, and reduced to the condition of spadix and endotheca separated
from one another by the intervening generative elements (fig. 10, p. .35).

In Clava, Hydractinia, &c., we have a still further advance in complexity. The gonophore
has here the form of a simple closed sac, whose axis is occupied by a cylindrical or club-shaped
spadix, round which the generative elements are clustered (woodcut, fig. 1 .5, A). Careful examina-
tion, however, will show that the walls of the sac consist of two membranes, an outer or ectotheca
and an inner or endotheca. The mesotheca is entirely absent.

In Garveia nutans, I have found a mesotheca to be distinctly demonstrable ; but it is closed
at the summit, and destitute of circular canal, while four short radiating canals may be seen in
its walls extending from the base of the spadix for about a third of the height of the sac (wood-
cut, fig. 15, £).

Fig. 15.

J b

Types of Gonophores.

A, Hi/dractinia ec/miala. B, Garveia nutans. C, Tulularia indhha. D, Si/ncori/ne eximia.

o, ectotheca; 6, mesotheca ; c, endotheca; rf, spadix; rf', manubrium ; c, radiating gastrovascular canals ;/, circular
gastrovascular canal; g, marginal tentacles; h, ocelli; o, ova; p, ovarian plasma in Tubularia.

Ill Tuhularia indivim the mesotheca presents the highest degree of development which
it attains in any adelocodonic gonophore, if we except the peculiar body described below under

THE GONOSOxME. IT)

the name of " mccoiiiclium." It is perforated at its summit, and tlic perforation is surrounded
by a distinct circular canal, which receives four radiating canals, which open into it by small
bulbous expanions (woodcut, fig. 15, C). We thus find almost entirely the conditions of a medusa
— a medusa, however, which never divests itself of its ectotheca, and accordingly never becomes
free, while the spadix remains as a simple caecal diverticulum, and the codonostome is reduced to
a mere perforation of the mcsotheca, this last exhibiting but the faintest traces of contractility,
and being quite incapable of acting as a locomotive umbrella.

From the sporosac of Tuhulnria hidivisa it is thus but a single step to the true phaneroco-
donic gonophore, such as we find in Corymorplia nidans, or Syncoryne cximia, where the
mcsotheca assumes the condition of a contractile locomotive umbrella, with a well-developed
codonostome and velum, and, the manubrium now becoming perforated by a mouth, the gono-
phore is no longer dependent on the trophosome for its nutrition, but can l)ccoine free and lead
an independent life in the open sea (woodcut, fig. 15, D).

The typical and ordinary condition of the spadix is that of a hollow cylindrical or clavate
body, occupying the axis of the adelocodonic gonophore. Occasionally, however, it departs from
this condition and becomes more or less branched, as in Cordylnphora lacustris (PI. Ill),
Plumidaria jnnnata, Laomedea caliculafa, &c.

The gastrovascular canals in the adelocodonic gonophore may, as we have already seen, be
either entirely suppressed or present the condition of simple, short, blind tubes, radiating from
the base of the gonophore, or be continued from this point as fully developed radiating canals to
the distal extremity of the gonophore, where they become united by a circular canal. In certain
free medusa3 [WUlia, Chidoncma — PI. XVII) the radiating canals subdivide before reaching the
circular canal.

The usual condition of the adelocodonic gonophore is that of a simple, more or less spherical
or oval sac. In Eudendrium, however, the male gonophores present the form of a simple sac
only at first ; for by the time that their contents have approached maturity, new spermatogenous
tissue becomes apparent between the endoderra and ectoderm of their supporting peduncles, and
these two membranes thus become separated from one another so as to form a second sac inune-
diately below the first, while a third may in the same way be formed below the second, the gono-
phore thus acquiring the peculiar moniliform or polythalamic conformation characteristic of this
genus (woodcut, fig. 16). It will beat once apparent that the separate chambers presented by this
peculiar form must not be regarded as so many distinct gonophores ; the whole moniliform series
ought rather to be viewed as a simple adelocodonic gonophore, in which the perigonium is not
uniformly separated from the spadix by the intervention of the spermatogenous tissue, but
remains at intervals permanently adherent to it. Among the planoblasts an entirely analogous
})henomenon occurs in a Sarsia-X^Q medusa of unknown trophosome, which I captured in the
towing-net on the south-west coast of Ireland (woodcut, fig. 17). In this the manubrium, which is
extraordinarily extensile, and can be projected to a great length beyond the umbrella, was enlarged
at distinct intervals by the development of the generative elements between its ectoderm and
endoderm. The specimen captured was a male, and the manubrium, when extended, presented,
by the mode in which the spermatogenous tissue was developed in its walls, five elongated
cylindrical enlargements, separated from one another by long thin intervening portions, in which
the ectoderm and endoderm of the manubrium continued in direct contact with one another, no
generative elements being there developed. The spermatogenous mass which occupied the free

46

MORPHOLOGY.

end of the manubrium was divided into two by a shallow strangulation. The peculiar mode in
which the generative elements are developed in the manubrium of Dipurena, a nearly allied "-enus
described by M'Crady," would seem to afford an example of an analogous phenomenon.

Fig. 16.

Male gonophore olSuden.
drhtm, showing the develop,
ment of the spermatogenous
tis3ue at intervals between
the ectoderm and endoderm.

Fig. 17.

Medusa {^Sa/i'sia strangulata, AUm., provisionally) of unknown
tropbosome.

a, very extensile manubrium opening at its proximal end into
a special cavity, tbe atrium ; b, h, b, male elements developed
at intervals between tbe ectoderm aud endoderm of tbe manu-
brium ; c, ocellus.

The gonophore may be borne upon a distinct peduncle, Avhich may be simple {Stjucoryne
eximia, &c.) or branched {Tubidaria iiidivisa, Cori/moqjha nutans, &c.), each branch then bearing
a gonophore on its summit; or the peduncle may be obsolete, and the gonophore become
sessile {Laomedea Jlexuosa, &c.)

The gonophores, whether plianerocodonic or adelocodonic, may be destitute of any farther

^ MX'rady, op. cit., p. 13u

THE GONOSOME. 47

covering, and will then, while still forming part of the hydrosoma, have their surface in immediate
contact with the surrounding water {Si/ncorj/nc, Clava, Ilijdractinia, &c.)

In other cases the blastostylc, with the gonophorcs which bud from it, may be surrounded
by a close case or capsule, formed by a layer of ectoderm with an external chitinous investment
{Campanularia unA Sertularia) (woodcut, fig. 2,/). I have elsewhere designated this capsule by
the name of " ffonanyium.'"^ The blastostylc extends through the axis of the gonangium as a
cylindrical column, bearing the gonophores as buds upon its sides, and generally expanded at
its summit into a conical plug or disc, by which the gonangium is here closed.

In some cases the contents of the gonangium escape, when mature, by the simple rupture of
the summit (Phtmiilaria, &c.). In others, hoAvever, the summit is separated as a distinct lid,
which is then either cast oS'at once [Sertularia pumila, &c.), or it remains movably attached by
one spot of its edge, as by a hinge, to the margin of the aperture thus formed in the summit of
the gonangium {Serhdaria opercidata, Antennularla anfennina).

In every instance where a gonangium exists the hydranths also are protected by a hydrotheca,
while the absence of a gonangium is always associated with the absence of a hydrotheca. The
diCFerence thus involved in the presence or absence of these parts corresponds to two primary
sections of the Hydroida, and I have distinguished all hydroids which possess a gonangium and
hydrotheca by the name of caltptoblastic, while the gymnoblastic hydroids are those which —
with the exception of the freshwater Hydras which constitute a separate section — are destitute of
these protective coverings.

In by far the greater number of cases the blastostyle in the calyptoblastic genera carries
numerous buds, which are either sporosacs or blastocheraes, and which always increase in
maturity as they recede from the base and approach the summit of the gonangium (woodcuts,
fig. 2,y, and figs. 18 and 19). In some cases, however, the blastostyle bears but a single bud;
and then it often happens that this enlarges to such an extent as to fill nearly the entire cavity
of the gonangium, the blastostyle being pushed aside out of the axis, and becoming often
partially absorbed, so as to render it difBcult to demonstrate its existence^ (woodcut, fig. 3, k).

It usually happens that a fleshy membrane abounding in thread-cells may be seen passing
over the whole of the generative buds while still attached to the blastostyle within the gonangium
(woodcuts, figs. 18,6? and 19, J). It closely confines them as in a common sac, and is probably an
internal layer which has separated from the original formative ectoderm on whose outer surface the
gonangium had been excreted. It performs an important office in the economy of the hydroid by
confirming the generative buds and guiding them or their contents towards the orifice of the
gonangium. I shall refer to it under the name ol guhernaadum.

Sometimes the blastostyle, though in the very young state quite simple, soon breaks up,
from a comuion point near the base, into several distinct tubes, which again luiite in the

^ " On the Structure and Terminology of the Reproductive System ia the Corynidse and Sertu-
lariadfe," 'Ann. Nat. Hist.,' July, 1860.

" The difference presented by the gonangia, according as they coutaiu numerous gonophores or
only a single one, is regarded by Gegenbaur (' Generatiousweclisel,' p. 38) as of sutiicieut importance to
induce him to distinguish the gonangia into "polymeric" and " monomeric." I am not disposed
liowever, to give much weight to this difference, the number of gonophores which a gonangium may
contain varying with particular conditions of development.

48

MORPHOLOGY

comuioTi cavity of the plug-like summit. This has been shown by Agassiz to be the case in
his Laomodra {C/i/lia) polerium, and I have myself seen it in a nearly allied species from the cast

Fig. 19.

Fig. 18.

"\\-

Gonangiiim, with its contents from Laomedea Jlexiiosa.
a, Goniingium ; i, blastostyle, carrying gonosacs, wliicli in-
crease in maturing from below upwards ; each containing a
single ovum in which the germinal vesicle and germinal spot
are conspicuous; c, opercular summit of blastostyle; d, common
investing meuibrane of the contents of the gonangiutn.

Gona

■ith its contents, from Obr-lia geiiiculata.

a, Cionangium; I, blastostvle loaded with medusa; (blasto-
chemes), which increase in maturing from below upwards;
c, opercular summit of blastostvle ; d, d, common membrane,
investing and confining the medusae

coast of Scotland (woodcut, fig. 20). It is entirely analogous to the breaking up of the Cffinosare
which will be described below as occurring in the stem of Antennularia.

In most cases the gonophore discharges its generative elements directly into the surrounding
water. In the females of some calyptoblastic species, however, the ova, instead of escaping directlv
from the gonoi)hore into the water or the cavity of the gonaiigium, are retained for some time in a
peculiar receptacle, where they undergo further development, and which is supported upon the
summit of the gonangium, and lies entirely external to its cavity (woodcuts, figs. 3, n, 21 and 22).
It will be found convenient to employ a special term for this receptacle, which confers upon the
gonosomes in which it occurs a very characteristic feature. I have already designated it by the
name oS. " acroci/d}" It may be seen in Serftdaria pumila, S. cupressina, S. polyzonias, Calycella
gi/riiif/a, Calycella laceraia, &c., and would seem to be in every instance confined to the female.

^ 'Pioc. Roy. Soc. Edin./ 1858.

THE GONOSOME.

49

There is some difficulty in (Ictcrmiiiiiig the exact morphology of the acrocyst. In its usual
form it seems to consist of a sim-
ple extension of the endothcca Fig- 20.
of the gonophorc, protruded as
a hernia-like sac through tiie
summit of the gonangium, when
the whole becomes surrounded
by a thick gelatinous-looking
envelope, which is excreted from
the outer surface of the sac, and
which shows no appearance of
true structure, though distinct
zones of deposition may occa-
sionally be observed in it.

In CalyceUa lacerata the
spadix itself, as was first shown
by Dr. S. Wright, is, with the
surrounding endotheca and ova,
carried upwards u[)on the blasto-
style, by whose elongation these
various parts ai'e protruded from
the summit of the gonangium,
and the endotheca thus becomes
a true acrocyst and excretes from
its outer side the usual thick
gelatinous investment (woodcut,
fig. 22). The peculiarity of the
acrocyst in this case is found
in the presence within it of the
spadix, which, however, is de-
pressed by the enlarging ova,
and forced back into the bottom
of the sac.

In the interior of the acro-
cyst the ova pass through certain
stages of their development, and
ultimately escape as free ciliated
embryos by the rupture of its
walls.

From the observations of Strethill Wright it would appear that in IJ'rijIilia {AlrudyJis) arenosa,
a gymnoblastic species, the ova escape, by the rupture of the sporosac, into a gelatinous mass which
is a secretion from the outer surface of the summit of the sporosac, to which it remains adherent,
and that in this gelatinous nidus they become converted into planulaj, and then escape into the
surrounding water. It is plain, however, that the simple gelatinous secretion which thus affords

7

Gonangium, with Its contents, from Laomedea repens, Alliu., showing the
compound blastostyle.

A, General viewj B, plan of transverse section, a, gonangium ; b, lower portion
of blastostyle continuing simple; c, opercular summit of blastostyle; d, separate
tubular branches into which the blastostyle has become divided ; «, the endotheca;
/, ramifications of spadix; (^, an ectodermal membrane, connecting the divisions of
the blastostyle with one another.

50

MORPHOLOGY.

a jjrotective nidus for the developing ova cannot be confounded with a true acrocyst, which is
found only in the calyptoblastic hydroids.

In the cases above described the acrocyst is destitute of any further covering, and has its
walls with their gelatinous investment freely exposed to the surrounding water. In Sertularia
rosacea, S. fallax and S. tamarisca, however, an additional covering is provided for the acrocyst,
and there is thus formed a curious and coniiilicatud receptacle, in which the ova, as in a sort of

Fig. 22.

Fig. 21.

Female gonangiuin, with acrocyst
of Seriularia pumila.

a, gonangivmi; i, blastostyle ; c,
opercular summit of blastostyle ;
d d, cffical offsets from the summit
of the blastotyle; e, gonophore
after having discharged its contents
into the acrocyst ; J", spadix ; ^,
proper sac of acrocyst; h, external
gelatinous investment of acrocyst;
i, ova contained in acrocyst ; Ic,
young ova in blastostyle.

Female gouangiuin with acrocyst of Cali/cella lacerata.

A, Young gonangiura, with its contents before the formation of
the acrocyst. B, older gonangiura, bearing an acrocyst.

a, gonangium ; b, blastostyle ; c, opei'cular summit of blasto-
style ; d, membrane passing over the gonosac while still confined
within the gonangium ; e, gonosac still within the gonangium, and
containing ova which surround a lobed spadix; y, proper membrane
of acrocyst j its gelatinous investment. Within the acrocyst are
seen ova already somewhat advanced in development, and the
spadix depressed towards the bottom of the sac by their enlarge-
ment.

iiiarsupium, pass through certain early stages of their development, previously to being discharged
into the surrounding water.

The nature and morphology of this receptacle in Sertularia rosacea (woodcut, fig. 23) will be
best understood by tracing its development. The young female gonangium (woodcut, fig. 25)
is a conical body, with eight slightly projecting longitudinal ridges, and with tlie broad end of
the cone constituting the distal end or summit of the gonangium, A blastostyle occupies its axis,
having upon its sides, one over the other, the young budding gonophores, and expanding at

THE GONOSOME. r,l

its summit into a broatl thick disc, which closes, as with a plug, the free end of the gonangiinii.
Upon the outer side of this disc a thin chitinous investment is excreted, becoming continuous ui

Fig. 23.

Mature female gonangium with marsupial
eliamber in Sertularia rosacea.

a, a, Cavity of the proper gonangium ;
c, c, blastostyle ; d, opercular Bumniit of blas-
tostyle; e, one of the eye's radiating arms which
are given oft" from the summit of the blasto-
style, marked by a chitinous sheath /, which is
prolonged from the walls of the gonangium ;
the arms are seen surrounding the marsupial
chamber, which contains ova advancing to-
wards the condition of planula?; g, the six
shorter spines which ci'owu the marsupium ;
h, the two longer lateral spines ; ?, i, the two
lateral ridges of the gonangium, wliich are
continued into the longer spines; k, k, k,
gonophores in various stages of development,
all surrounded by the gubern.icular membrane.

Fig. 24.

Young female gonangium of Sertularia
rosacea, showing the formation of the
marsupial chamber.

a. Cavity of the proper gonangium ;
c, blastostyle ; c, radiating arms prolonged
from the opercular summit of the blasto-
style ; y» chitinous sheath of arms ; i, i,
lateral ridge of gonangium ; k, gonophore
surrounded by its guberuacular membrane.

the edge of the disc with the chitinous walls of the gonangium, while in the centre of the disc
the chitinous investment is deficient, leaving here the summit of the blastostyle naked.

The edge of the disc soon becomes produced into eight thick symmetrically radiating lobes {(%t')
which gradually elongate themselves, carrying with them a continuation of the chitinous excretion,
which forms a wide tube around each ; and now bending upwards, in the form of eight arms
with enlarged extremities, they remind one of the disposition of the petals in a flower, and

52

MORPHOLOGY.

Fig.

present altogether an appearance of great elegance (woodcut, fig. 24, e). These eight radiating
arms are composed of ectoderm and endoderm, and have their axis occupied by a tubular cavity,
whicli communicates witii that of the blastostyle. As the arms continue to elongate, we find
tliem next witli their free extremities bending towards one another, until finally (woodcut, fig.
2.3), they completely enclose a space, which becomes entirely shut in by the lateral coalescence of
tiie wide chitinous tubes with which the radiating processes are each invested.

In the mean time the eight longitudinal ribs of the
gonangium continue themselves upon the radiating arms,
and ultimately extend beyond their extremities as free
pointed processes (y, h). Two of them, however, situated
opposite to one another, greatly surpass the others in size,
and mainly contribute to the peculiar and characteristic
form of the gonangiimi. Into the marsupial chamber thus
formed the ova make their way, enclosed apparently in a
proper acrocyst.

A very similar condition is presented by Scrtularia
fallax. Here, however, the summit of the blastostyle is
prolonged into four instead of eight lobes ; but the subse-
quent history of these is in all essential points the same as
in Serttilaria rosacea.

If we compare the structures now described with an
ordinary hyrandth, we shall have no difficulty in recog-
nising an exact parallelism ; for the tubular processes
which are developed from the summit of the blastostyle
may be regarded as homologous with the tentacles of a
hydrantli. They have, however, undergone a special modi-
fication, by which they become subservient to an entirely
diff'erent function from that of the tentacles of the hydranth ;
for, no mouth being developed on the blastostyle, they are
no longer prehensile organs administering to the alimenta-
tion of the colony, Imt, like the blastostyle itself, have
assigned to them functions appertaining to reproduction rather than nutrition, and are destined
to circumscribe a cavity for the retention and development of the ova.

The ova would seem to continue in the marsupial cavity until they have acquired the
condition of ciliated embryos.

The modification of marsupial receptacle which occurs in Serttilaria tamarisca is also very
interesting. The female gonangia (woodcut, fig. 26 b) are here of an oval form for about the
proximal half of their length, and then become trihedral, with the sides diverging upwards,
while the whole is terminated by a three-sided pyramid. The sides of the pyramid are cut
into two or three short teeth along their edges, and each of their basal angles is prolonged into a
short spine.

The trihedral portion, with its pyramidal simnuit, is formed of three leaflets (y,^',y") which
merely touch one another by their edges, without adhering, so that they may be easily pushed
aside from one another by the dissecting needle, or by the embryo during its escape.

Very yonng female gonangium of

Sertularia rosacea,
tf. Cavity of the gonangium ; c, blasto-
style; d, opercular summit of blastostyle;
e, radiating lobes from summit of blasto-
style, about to be prolonged into arms ;
k, young gonopbore.

THE GONOSOME.

53

On laying open the gonan
a blastostylc [d), which gives
origin to one or more gono-
phores. It terminates upwards
by closing round the distal ex-
tremity of the blastostyle, where
it forms a ring (/) with tooth-
like processes, l)y which the
extremity of the blastostyle is
encircled. This oval portion
constitutes the gonangium pro-
per, and is the only part deve-
loped in the male (fig. 20 a).

From the summit of the
blastostyle three tentacula-like
processes {//) arc given off.
They constitute ramified cffical
offsets of the cavity of the
blastostyle, and are composed
of an ectodermal and an endo-
dermal layer. Immediately
after their origin from the
blastostyle they enter the ter-
minal leaflets, and now lie
between the two lamiua3 of
which these leaflets are com-
posed.

The leaflets, with their
contained csecal oSsets from
the blastostyle, surround a
membranous sac (0 which is
borne on the summit of the oval
portion, or proper gonangium,
and contains one or two ova,
which are usually in an ad-
vanced stage of development.
Each ovum is immediately in-
vested by a delicate structure-
less sac (/•), which is continued
by a neck-like prolongation to
the summit of the blastostyle.'

The homological relations

gium the oval or proximal portion of it is seen to be occupicxl by
Fig. 2(5.

Male and female gonangia of Sertularia tamarisca.
A, Male gonangium ; B, female, with its marsupial chamber.

a. Proper gonangium ; b, blastostyle with c, its opercular summit; d, gubernacu-
lum ; e, gonophores, having the spadix surrounded by the spermatozoa in the male
and by the ova in the female ; fy chitinous roof of the gonangium proper in the
female, perforated by a central denticulated aperture ; g^ g\ g'\ the three leaflets,
forming the walls of the marsupial chamber, and each containing within it a rami-
tied ca?cal process, h, from the summit of the blastostyle; g'' is cut across and de-
", acrocystftl sac ; A", ovum, included in a special sac.

The account here given of the marsupial receptacle of Sertularia tamarisca differs in some

54

MORPHOLOGY.

between the marsupial receptacles of Sertularia rosacea and Sertularia fallaj; on the one hanrl,
and Sertularia tamarisca on the other, are at once apparent, and are very interesting. The
ramified tubes of S. tamarisca are manifestly the representatives of the simple tubes in S. rosacea
and S.fallax; while the three broad chitinous leaflets within which the ramified tubes are con-
tained are homologous with the hollow chitinous processes of the gonangium, which in S. rosacea
and S.fallax enclose the simple tuljes, and subsequently coalesce to form with their prominent
ridges and spines an external capsule-like covering for the sac, into which, as in S. tamarisca, the
ova are expelled from the gonangium proper.

The structures just described in Sertularia rosacea, S.fallax, and S. tamarisca will, I think,
enable us to explain a peculiar feature observed in aS*. pumila, and probably some other species. In
S. pumila the blastostyle of both male and female gonangia gives off from its enlarged opercular
summit several more or less ramified esecal tubular processes (woodcut, fig. 21, d, d), which,
instead of developing themselves externally, are found entirely within the gonangium, where they
hang freely from the summit of the blastostyle. Their walls are composed both of endoderm and
ectoderm, and their cavity communicates with that of the blastostyle, so that the peculiar coloured
corpuscles which circulate within the cavity of the blastostyle are freely admitted into the esecal
tubes, where they may occasionally be seen in active motion. The tubes can be most satisfactorily
examined in the younger gonangia. In the older ones they will frequently be found to have
contracted adhesions to the gonangium, to have become atrophied, and, finally, even to have

disappeared.

I believe that these tubes are the exact equivalents

of those which in Sertularia tamarisca and S. rosacea are
given off from the same part of the blastostyle, but where,
instead of growing into the cavity of the gonangium, they
are developed in an outward direction, and assist in the
formation of the peculiar receptacle which surrounds the
acrocyst in those species.'

Among the most remarkable modifications of the tro-
phosome is that of Cojjpinia arcta (woodcut, fig. 27). In this
singular hydroid the hydrothec;e and gonangia springdirectly
from a creeping retiform hydrorhiza, while the gonangia,
which are very numerous, become closely adherent to one
another by their sides, so as to form with the jjroximal por-
tion of the hydrothecac and with the hydrorhiza a continuous
encrusting basis spreading over the surface to which the

Fig. 27.

A portion of the hydrosoma of Cophinia arcla.

a, a, a, HydrothecEE ; b, a hydranth retracted
and destitute of tentacles; c, a hydranth re-
tracted and with its tentacles present; d, d,
basal eucrnsting portion formed by the juxta-
position and adhesion of tubular gonangia ;
e, sporosac, enclosing a solitary ovum ; f,f,
acrocysts, enclosing the ovum in a more ad-
vanced stage of development; g, retiform
stolon, forming the hydrorhiza.

respects from my original description of the same part {' Brit.
Assoc. Report on Hydroid.\'), subsequent more favorable ex-
aminations havinar caused certain modifications of mv former

' It is evidently the tubes here described to which Agassiz
(' Nat. Hist. U. S.,' vol. iv, p. 329, pi. xxxii, figs. 10, 10") refers
as occurring in a North American hydroid which he regards as
identical with the Sertularia pumila of the European coasts.
He views them, however, as simply representing the fleshy
bauds which may frequently be seen iu the trophosome of the Hydroida, exteudiug from the outer

THE GONOSOME. 55

Iiydi-oid had attached itself. Each gonangiura in the female contains a single sporosac with
a single ovum ; and this ovum, after a time, becomes extra-capsular in order to undergo within
an acrocyst some of the earlier stages of its development.'

But tiiere is, perhaps, no modification of the gonosome more interesting than that which we
meet with in Gonothijraa Loveni, Allm., and at least one other allied species. In this calypto-
blastic hydroid there are borne upon the summit of the gonangium, and altogether external to
its cavity, certain very peculiar gonophores, which convey the impression of small, fixed,
imperfectly dcvcloiiutl medusc^ (woodcut, fig. 2S). It was to these extracapsular gonophores

surface of the coenosarc to the inner surface of the chitinous pciidcrm, and which these tubes
certainly resemble when they become more or less atrophied and adherent to the walls of the
gonangium. They are also described and figured by Lindstrom in a paper " On the Development of
Sertulariapumi/a" (' Oefversigt af Kongl. Vetenskap-Akademiens FiJrhandlingar,' 1855).

' As the nature of Coppinia arcta has been hitherto very imperfectly understood a more detailed
description of it may here be given.

The hydrosoma forms small sponge-like masses on the stems of the larger hydroids, and is espe-
cially abundant on Seriularia abietina and Pimmlaria falcata from deep water.

Even on a superficial inspection it may be seen to consist of two distinct portions. Of these, one
constitutes a continuous encrusting base, and the other consists of curved cylindrical tubes which
project from the free surface of the base. In each of these tubes is contained an essertile and retractile
hydrauth. The tubes are thus true hydrothec*. The hydranths are conspicuous by their fine lemon-
yellow colour, and are furnished with a verticil of filiform tentacles disposed round the base of a
short conical hypostome. They are, however, often imperfect and apparently destitute of mouth and
tentacles.

The encrusting base which forms the most remarkable part of the hydroid has never yet been
correctly described.

The liydrothccal tubes can be traced through it to its attached surface, while vertical and
transverse sections show that the rest of the crust is mainly composed of vertical chitinous
tubes rendered polygonal by mutual pressure. They adhere to one another by their sides, and
each, as had been long ago shown by Dalycll, opens on the free surface of the crust by a small circular
aperture.

These tubes are true gonangia ; within each is a solitary sporosac which seems to have originally
budded from a blastostyle, which soon, however, becomes suppressed by the growing sporosac. A
sufficiently obvious spadix may be recognised in the sporosac which contains a single large lemon-
yellow ovum, in whose earlier stages there may be seen a distinct germinal vesicle, while the place of
the germinal spot is taken by numerous clear spherical bodies which disappear in a few seconds after
the ovum is pressed out of the sporosac and exposed to the action of the surrounding water.

Segmentation commences while the ovum is still within the gonangium, and the ovum becomes
thereby converted into a mass so plastic that it allows of its being forced through the small aperture
in the summit of the gonangium. In its exit it carries out with it a hernial extension of the attenu-
ated walls of the sporosac, which thus form for it an acrocyst in which it is still for some time confined.
It ultimately, by the rupture of the acrocyst, escapes as a planula into the surrounding water. The
planula and its development into a hydranth enclosed in a chitinous tube have already been observed
by Daly ell.

Both hydrotheca; and gonangia spring from an adherent retiform hydrorhiza without the inter-
vention of a hydrocaulus.

A knowledge of the structure of Coppinia will enable us to give a more correct generic

56 MORPHOLOGY.

that Loveii long ago^ called attention when he supported and developed the doctrine, just then
announced by Ehrenberg, of the sexuality of the Hydroida — a doctrine which, though in its
mode of statement not aljsolutely correct, was yet fidl of significance.

The bodies {(],(/') in question are nearly spherical sacs, and occur in both the male and female
colonies. In their walls may be demonstrated an ectotheca, mesotheca, and endotheca. The
generative elements are formed within the endotheca, and surround a well-developed spadix.
The endotheca, however, is generally of short duration, becoming absorbed or ruptured under the
increasing volume of its contents. In the female four very distinct radiating canals may fre-
quently be seen ; these spring from the base of the spadix, and thence run in the walls of the
mesotheca towards the opposite end of the sac. In many cases, however, I was unable to
detect any trace of these canals, and could never find them in the male. We should, however,
be scarcely justified in asserting that in such cases they are altogether absent ; for it is quite
possible tliat emptiness or some other peculiar condition at the time of observation may have
caused them to escape detection — a supposition which receives confirmation from the fact that,
even in cases where they are obvious, they become obliterated under slight pressure.

At the summit of the sac the mesotheca is perforated by a circular aperture, round which
its walls appear to be thickened, and seem to contain here a circular canal in which the radiating
canals terminate ; at least, the presence of a line of coloured granules at this place affords an
indication of the existence of such a canal. The ectotheca, which is loaded with thread-cells, is also
perforated by an aperture corresponding to that of the mesotheca; and the gonophore is crowned
by a circle of short tentacles, which seem to originate from the thickened margin of the perfo-
ration in the sunmiit of the mesotheca.

These tentacles possess, like the marginal tentacles of a true medusa, considerable con-
tractility. They may frequently be seen of very different lengths in different gonophores of the
same colony ; and this, which is really the result of different degrees of contraction, may be
easily taken for different degrees of development, the tentacles being especially sluggish in the
acts of extension and contraction. Their length, when fully extended in the female gonophore,
will equal about half the diameter of the gonophore. While under external irritation, they will
slowly contract to a third of their original length, and will then show themselves as a little
stellate crown on the summit of the gonophore, reminding us of the sessile stigma in the pistil

diagnosis than was possible so long as we were ignorant of the true nature of this curious hydroid, and
the following may be taken as expressing the essential characters of the genus :

CoppiNiA, Hassall.

Trophosome. — Hydrocaulus absent ; hydrothecoe tubiform, springing from an adherent retiforra
hydrorhiza.

Hydranths with a single verticil of filiform tentacles.

Gonosome. — Gouosphores adelocodonic ; gonangia tubiform, sessile on the hydrorhiza, aud forming,
by the close approximation of their sides, a continuous incrusting mass surrounding the bases of the
hydrothecte which project from it at intervals.

^ Loven, " Beitrage zur Kenntniss der Gattungen Campanuluria und Sijncoryne ,''' 'Wiegm. Arch.,
1837. Loven names the hydroid in which he witnessed the extracapsular gonophores Campamdaria
geniculala, which is certainly a wrong determination of the species. See " Notes on the Hydroida,"
'Ann. Nat. Hist.' for August, 1859.

THE GONOSOME.

Fig. 2S.

of a poppy. They vary in number ; I have counted in the female from eight to sixteen or twenty.
They are composed of ectoderm and endoderm, the ectoderm containing thread-cells, and the
endoderm presenting the usual sep-
tate appearance. They are less
numerous and less developed in the
male than in the female.

The contents of the sacs are,
as in an ordinary gonophore, either
ova or spermatozoa, and the sexes
are invariably found separated on
distinct colonies. The ova, while
contained within them, pass through
the various stages of development
up to that of ciliated embryos, in
which state, as has been already
shown by Loven, they are discharged
into the surrounding water through
the orifice in the summit.

If we follow the development
of these extracapsular gonophores,
we shall find, as, indeed, Loven had
already pointed out, that they are
originally produced within the go-
nangium where they originate, ex-
actly like intracapsular gonophores,
as buds from the blastostyle. By
the growth of the blastostyle the
gonophores are carried upwards with
it, in the order of their maturity —
the oldest ones, while within the
gonangiinn, being always nearest the
summit of the blastostyle ; but in-
stead of discharging their contents
and then withering away on their
arrival at the orifice of the gonan-
gium, as in ordinary adelocodonic
forms, they are here carried out
through the orifice, become truly
extracapsular, and in this state un-
dergo, with their contents, further
development, while the growing

blastostyle alwaVS keeps its extremity "^^ ™°'''^ advanced meconidium ; /, radiating canals, united by the c
"^ "^ n ; m, ovum with germinal vesicle and spot; o, o, embryos, escapin?

truncated on a level with the summit of ciliated piauuia;.

of the gonangium, whose orifice it continues to close by a plug-like expansion, which at the same

Gonangium with Meconidin of OonothyrtBa Loveni (female).
a. Cavity of gonangium; b, b, blastostyle, with c, its opercular summit;
d, d', d'\ gonophores in various stages of maturity still within the gonangium,
all surrounded by the gubernacular membrane; in d" the tentacles are seen
already formed and lying back on the walls of the gonophore; g^g't meconidia,
g' being more advanced tlian g ; e, mesotheca, external to which may be seen the
ectotheca loaded with thread-cells ; Jc, endolheca ; k\ remains of endotheca in

rcular canal,
in the form
of ciliated plauulEe.

58

MORPHOLOGY.

Fig. 29.

time affords a support for the gonophores after they have become extracapsular. Two or tliree of
these extracapsular gonophores, in different stages of development, may Ije usually seen, borne
each by a short peduncle upon the opercular summit of the blastostyle, with whose cavity that
of their spadix freely communicates through the tul)ular axis of the peduncle.

While the gonophore is still contained within the gonangium, the mesotheca has become
developed in it, and in the more advanced ones {d") the rudimental tentacles may be seen thrown
back in their walls in the form of a little star.

That the bodies now described belong to the class of adelocodonic rather than to that of phane-
rocodonic gonophores must, I think, be admitted. In all essential points, except in the presence of
entacles developed from the mesotheca, they agree with the gonophores of Tuhidaria iudivisa, which
must certainly be classed among the adelocodonic forms, notwithstanding their possession of a
well-developed mesotheca and gastrovascular canals. In both the
aperture of the mesotheca is reduced to a mere perforation, and in
neither is the mesotheca ever developed as a locomotive organ.

It nuist also be borne in mind that, when planoblasts are
produced in the Campmudarince, they are in almost every instance
blastochemes ; in other words, they belong to the type in which
the generative elements are produced, not dii'ectly, as in Gono-
tliyraa, between the ectoderm and eudoderui of the manubrium,
but are formed in special zooids developed from some pai'ts of the
gastrovascular system ; Leptoscyplms tenuis, Allm.,^ affording the
only known exception to this rule.

The extracapsular gonophores of Gonotliyraa Love id are thus
of no little interest in the morphology of the Hydroida, and it will
be found convenient to speak of them under a special name. Their
resemblance to a pomegranate, or perhaps still more obviously to
a poppy-capsule, with its sessile stellate stigma, will instantly strike
us ; and it is this comparison which has suggested to me the name
of meconidium ^ by which I have elsewhere found it useful to
designate them.

A very remarkable feature, which one is at first sight tempted
to place in the same category with the formation of meconidia,
but which is in reality of an entirely different significance, is pre-
sented by Halecium halecinum. In this hydroid there is borne
upon the summit of the female gonangium, in a situation precisely
similar to that of the meconidia of Gonothyroea Love/ii, a pair of
hydriform bodies (woodcut, fig. 2Q d). These bodies present no
appreciable difference by which they may be distinguished from the
ordinary hydranths of the trophosome. They are of an elongated
oval form, with the mouth situated on the summit of a short conical
hypostome, which is siUTounded by a circle of filiform tentacles. They are always two in

^ "Notes on the Hydroid Zoophytes," 'Ann. Nat. Hist.,' Nov., 1859.

^ A diminutive noun, formed from /jh'ikuiv, a poppy. "Notes on the Hydroid Zoophytes,'"' 'Ann,
Nat. Hist.,' August, 1859.

Gonangium with gonangial
hydranths in Halecium hale'
cinum,

a, a, Gonangium ; h, h, blas-
tostyle ; r, gubernacular mem-
brane, still confining the ova, e, e,
which are here in an advanced
' stage of development, the proper
gonophore having become ab-
sorbed after discharging its con-
tents ; d, gonangial hydranths.

THE GONOSOME. 59

number, and diverge from a common point of attachment, wliilc thcii- wide gastric cavities, after
contracting below, communicate here with one another and with the tuljular cavity of the
blastostyle.

I have never been able to discover any direct relation between these gonangial liydranths
and the generative functions of the hydroid. The ova, so far as I can determine, seem to be
produced in the usual way in a sporosac which springs from the blastostyle, and are tiicn dis-
charged into a chamber formed by the gubernacular sac (c), the sporosac itself entirely disappearing
after the loss of its contents. In the sort of internal marsupium thus formed the ova pass
through certain stages of their development before their ultimate liberation as planulae through
the summit of the gonangium, which takes place, probably, after the disappearance of the
gonangial hydranths.

I may here mention a very singular body, whose exact significance I have never been able
satisfactorily to determine, and which may he seen in the female gonangium of Aiifennularia
antennina, where it is of frequent occurrence. It is always found floating free in the cavity of
the gonangium, along with the ova which had escaped from the ruptured gonopiiores, and
resembles an imperfectly developed medusa, with a large and apparently imperforate manubrium,
but with its umbrella closed, and without any trace of gastrovascular canals, the walls of the
umbrella being separated from the manubrium by a considerable space, which is filled with
a clear fluid. It may be compared to a free sporosac ; but it is much smaller than the ordinary
sporosacs of the AniennuJaria ; and I have never observed in it any trace of generative elements.
It is possibly an undeveloped sporosac, produced, like the perfect sporosacs, as a bud from the
blastostyle, and becoming separated at an early stage ; but I can offer no decided opinion either
as to its origin or its ultimate destination. It may be a parasite, though it is not easy to reconcile
its peculiar structure with this view.

In almost every case the gonangium, when present in the Hydroida, is destitute of any
fiuther covering. In certain hydroids, however, belonging to the family of the Plumiduriclce,
the gonangia are developed in groups, and each group is contained in a common receptacle,
which confers upon the hydroid in which it exists a very striking and characteristic feature.
This receptacle must be carefully distinguished from a proper gonangium, with which, indeed, it
has been confounded in various descriptive works on the Hydroida. It will therefore be very
convenient to give it a special name, and I have already proposed for it the term corbula,
suggested by its basket-like form.^

In Aglaophenia phima the corbulas (woodcut, fig. 30, D) may be plainly seen to be meta-
morphosed ramuli. The peculiar metamorphosis of a ramulus, which results in the formation
of a corbula, consists here in the suppression of the hydrothecse, accompanied by the development
on each side of the ramulus of numerous oval, hollow, alternately placed leaflets ; each leaflet
consisting of a diverticulum from the coenosarc of the ramulus, invested by a continuation of
the general perisarc.

In the earliest stages of these leaflets (A, a) their edges are entire, but they soon become deeply
serrated by the formation of hollow tooth-like processes, more especially upon the edge which is
turned towards the distal extremity of the ramulus. Upon the proximal edge of the leaflet these
processes usually remaii in an imperfectly developed state, though they are occasionally equally

' ' Proc. Roy. Soc. Ediu.,' 1858.

60

MORPHOLOGY.

well developed on both edges. The processes which are thus developed on the edges of the
leaflet are in all respects similar to the lateral nematophorcs of tlie trophosome (see p. 28).
They are filled, like these, with soft granular protoplasm, in which is immersed a cluster of fusiform
thread-cells, and which is in direct communication with the coenosarc filling the cavity of the
leaflet. They are also, like these, perforated at their extremity by an oblique aperture ; but I
have never seen the nematophorcs of the corbulas emit, like those of the trophosome, pseudo-
podial prolongations of their contents.

Fig. 30.

Development of the Corbula iu Aglaoplien'ia pluma.

A, Very young corbula; B, corbula more advanced; C, corbula in a still more advanced stage; D, the mature
corbula; E, transverse section of mature corbula, sbowing two gonangia, each containing a single gonophore. a, a,
leaflets of corbula; h, h, gonangia; c, ramulus supporting the leaflets; d, a hydrotheca.

F, Separate leaflet from mature corbula. a, continuation of the somatic cavity into the leaflet, where it divides into
two branches, & J ; c, nematophores forming tootb-like processes on the distal edge of the leaflet ; d, imperfectly deve-
loped tooth-like processes on the proximal edge; e, septum, dividing the cavity of the leaflet,

G, Gonangium from mature corbula. a, continuation of somatic cavity into gonangium ; 5, blastostyle, partially
suppressed by the enlarging gonophore ; c, gonophore ; d, spadix ; f, ovum ; g, wall of gonangium.

As the young leaflet continues to grow, its cavity becomes partially divided by a cliitinous
septum (F(?) which stretches across from the outer to the inner side, parallel to the axis of the
leaflet, but always nearer to the proximal edge. At the free end of the leaflet the septum is

THE GONOSOME. / Gl

incomplete ; so that here the contents of the cavity at one side of the septum communicate with
those upon the other side, both sides communicating at the base of the leaflet with the common
cavity of the coenosarc.

The leaflets, as they increase in size (B, C), direct themselves vertically from the upper surface
of the ramulus, and those of one side arch over so as to apjjroach those of the opposite. They are
at first free, but they afterwards become intimately united at their edges, the nematophores con-
tinuing to project as tooth-like processes, and forming an elegant serrated ridge between every
two leaflets. Ultimately the leaflets of one side coalesce with those of the other by their summits,
and thus form a completely closed chamber (D).^

In the receptacle thus formed the gonangia are produced. They spring from the upper side
of the metamorphosed ramulus, near the point where the leaflet leaves it, and represent the
hydrothecae which e.xist on an ordinary ramulus, and whose place they here take. They begin
to be produced at an early stage of the corbula, and may be easily examined in the young
corbula before it has become closed (B i^, Ci). The metamorphosed ramulus generally remains
unchanged for a short distance from its origin, and may be here seen bearing one or two ordinary
hydrothecae.

About twelve gonangia are usually contained in each corbula. They are of a
very simple type (G), of a regular oviform figure, and with their chitinous walls thin and
delicate. Each gonangium seems to contain but a single sporosac, which soon comes to
occupy almost its whole cavity. A long, nearly cylindrical spadix extends from the base to
the summit of the sporosac, passing in the male through the axis of the mass of sperma-
togenous tissue, but in the female pushed to one side by the development of the large single
ovum, which here occupies almost the whole remaining portion of the cavity of the sporosac.

There may appear some difliculty in deciding as to whether the corbula ought to be
regarded as properly belonging to the trophosome or to the gonosome. The truth is, that it
holds a place exactly intermediate between the two, and may in this respect be compared to the
bracts in plants; for these are in the same way intermediate between the ordinary leaves
and the proper floral verticils. As the bracts, however, are usually treated of in connection with
the inflorescence, whose limitation they frequently determine, we shall, perhaps, here also find it con-
venient to speak of the corbula in connection with the gonosome rather than with the trophosome.-

^ In some other species (Jfflaophenia myriophyllum) the leaflets never coalesce, and the corbula
remains permanent!}' open.

2 In a very ingenious paper, " On the Morphology of the Reproductive System in tlie Sertularian
Zoophytes," by Professor E. Forbes ('Ann. of Nat. Hist.,' 1844, vol. xiv, p. 385), the autlior recognises
in the corbulse oi Aglaojihenia plmna, and some other allied species, their true significance as meta-
morphosed branches. He mistakes, however, the nature of the metamorphosis, while, in accordance
with the prevailing view, he sees in the receptacles in question bodies in all respects corresponding to
the proper gonangia of the other hydroids.

Forbes, moreover, extends his generalisation, applying it to the gonangia of the other Sertularians,
which he believes must be all regarded as peculiarly metamorphosed branches, with metamorphosed and
confluent hydrothecse, exactly in tlie same way that the floral verticils in plants may be referred to
verticillate, metamorphosed, and variously combined leaves. " The vesicle," he says, " is formed from
a branch or pinna through an arrest of individual development, by shortening of the spiral axis, and, by
a transformation of the stomachs (individuals) into an ovigerous placenta, the dermato-skeletous (or

62 MORPHOLOGY.

As a general rule, there is no perceptible difference between the male and female colonies of the
same species of hydroid, either in the trophosomc or the gonosome, beyond what is, of course, pre-
sented by the generative elements themselves. In some cases, however, the difference is sufficiently
well marked. Thus in Sertularia famaiisca the male and female gonangia (woodcut, fig. 26) differ
strikingly from one another ; for the male gonangia are compressed, somewhat obcordate recep-
tacles, with a short terminal tubular aperture ; while the female are oval for about the proximal
half of their height, and then become trihedral with the sides diverging upwards, the whole being
terminated by a three-sided pyramid whose edges are cut into two or three short teeth, and the
basal angles prolonged into a short spine.'

So also in Serf/ikiria romcca a well-marked difference may be seen. The male gonangia
are here of a conical form, curved near the apex, which is their point of attachment, and provided
with six longitudinal ridges in the form of thin projecting lamellas, each of which terminates at
the distal extremity in a free-pointed process which arches over the summit of the gonangium.
In the female gonangium (woodcut, fig. 23) the longitudinal ridges are eight in number, while
two opposite ones being greatly more developed than the others give to the gonangium the very
elegant and striking form which caused Ellis to compare it to a " lily or pomegranate-flower just
opening." A very similar difference exists between the male and female gonangia of Sertularia
falla.r, and generally in the group which under the name of Biphasia, Agassiz has separated from
Sertularia. In all these cases the difference depends on the formation in the female of the remark-
able marsupial chamber whose structure has been already described (see p. 52).

It will also be borne in mind that, in those species which develop an acrocyst on the
summit of the gonangium, this body is formed only in the female ; while it is on the female
gonangium alone of Halecium halecinum that the little geminate hydranth already described
(p. 58) is produced ; and to these cases we may also add the difference presented by the male
and female meconidia in Gonothyrea Loveni (see p. 5G).

Among the gymnoblastic hydroids, also, certain differences may be occasionally observed
between the male and female. Thus, the tentaculoid tubercles which, in certain Tulularice, crown
the gonophore are in some species more fully developed in the female than in the male ; but the most
striking difference is fouiul in the genus Eudendriuin, whose male gonophores are situated in a
verticil on the body of the hydranth, and present the remarkable polythalamic condition already
described, while the female gonophores originate irregularly for some distance backwards on the
branch, and are always monothalamie (see Pis. XIII and XIV). This difference between the
male and female gonophores in Eudendrium struck Cavolini long before the presence of a male
element in the Hydroida was suspected, and led him to suppose that Eudendrium reproduced

cells) unitiug to form a protecting capsule or germeu ; whicli metamorpliosis is exactly comparable with
that which occurs in the reproductive organs of flowering plants, in which the floral bud (normally a
branch clothed with spirally arranged leaves) is constituted through the contraction of the a.\is and the
whoiling of the (individual) appendages borne on that axis, and by their transformation into the several
parts of the flower (reproductive organs)."

The theory, however, involved in the above statement, attractive though it be, is contradicted by
the actual development of the parts in question. When Forbes wrote, so little was known of the struc-
ture and development of the Hydroida, that this accomplished and lamented naturalist may well be
excused if some parts of his very suggestive paper have refused to stand the test of subsequent research.

^ It is apparently the m^le gonangia which Ellis has figured iu his description of this species.

THE GONOSOME. 63

itself by two different kinds of eggs. In accordance with this view, he called the female gono-
phores in his Sertularia {Eudciuhium) raceiiiosa "nova a racemo," and the male gonophores
" nova a corimbo."^

The differences above described between the male and female are all confined to the
gonosome ; the trophosome, however, does not appear to be always exempt from a participation
in sexual difference, for in Hydradinla polyclina, Agass., the hydranths of the male colony
are described by Agassiz as differing from those of the female colony by their more elongated
proboscis.^

We may now consider how the principal modifications which we have described as presented
by the gonosome are distributed among the leading groups of the Hydroida.

There is no fully established instance of the same species of hjdroid producing both
phanerocodonic and adclocodonic gonophores, either simultaneously or consecutively ; and Sars'
is certainly in error when he includes under his Podocoryne carnea, two forms of hydroids,
one with developed medusae, and the other with sporosacs. Neither is there any known instance
of a species with blastochemes producing gonophores in any other way than through the medium
of the blastocheme, and there can be little doubt that Van Beneden* has made some confusion
between two distinct species when he figures a portion of a hydroid colony, which he names
Campanularia (/enicuJata, with two kinds of gonangia, one containing medusae and the other
sporosacs.

Among the gymnoblastic hydroids the gonophores may be borne either l)y the trophosome
directly or by blastostyles, but they are never included within a gonangium. We have here
some species with phanerocodonic and others with adclocodonic gonophores, and the two forms
would seem to be pretty ecpially distributed through the group. Unless Nemojysis should prove an
exception, there is no known example of the occurrence of a blastocheme among the Gymnoblastea.
It is, however, by no means impossible that the sexual lobes of Nemopsis whose bases extend
over portions of both the manubrium and radiating canals, ought to be regarded as true zooids.
If this be so, then the Nemopsis medusa must be regarded as a blastocheme, though M'Crady
has shown that its trophosome is that of a true tubularian.

Among the CampamdarincE we meet with medusiform planoblasts, as well as with fixed
sporosacs, both forms being produced in nearly equal proportion. The planolilasts, however, belong,
with only a single known exception, — that, namely, which is afforded l)y Leptoscijpjlms tenuis — •
to the type of the blastocheme. Both planoblasts and sporosacs are in the CampanalariucB
always developed upon blastostyles within a gonangium.

The Geri/onidcB, a group composed of medusae which have not yet been traced to a hydraform
trophosome, must probably, as we shall see below, be regarded as true blastochemes.

Finally, among the Serfularince we know as yet of no instance of a planoblast, the gene-
rative elements being among these hydroids always produced in fixed sporosacs, which, as in the
Campanidarina:, are invariably borne on the blastostyle of a gonangium.

' Cavoliui, ' Mem. Polypi Marini,' 1785.

^ Agassiz, ' Nat. Hist. United States/ vol. iv, p. 228.

' Sars, ' Fauna lit. Norv.,' p. 7, pi. ii, fig. 5.

* Van Beueden, 'Mein. sur les Canipanulaires,' pi. iii, figs. 1 — G.

MORPHOLOGY.

9. The Generailve Mements.

I have thus far endeavoured to give a complete account of the morphology of those parts
which are destined for the origination and protection of the generative elements. These elements
themselves maj' next be examined.

The existence of generative elements — ova and spermatozoa — has now been fully determined
in every important group of the Hydroida.

Ova. — The hydroid ovum (woodcut, fig. 40 A), in all those cases where its structure has been
satisfactorily seen, consists of a granular vitellus enveloping (except in the genus TubuJaria and
probably the other ac//;2«/a-producing hydroids) a distinct, more or less excentric, germinal vesicle,
in which one or more germinal spots may be almost always demonstrated, and occasionally with one
or more puncta or nucleoli in the interior of the germinal spot. The whole is invested by an ex-
ceedingly delicate vitellary membrane, which, though it sometimes escapes detection, is probably
always present, at least during some period in the existence of the ovum. In the genus T/ihuIaria
the most careful investigation has as yet failed in detecting any trace of germinal vesicle or
spot.

From some observations which I have been enabled to make on certain very earl)' stages of
the ovum, it would seem that the germinal vesicle shows itself before any distinctly differentiated
vitellus has begun to envelope it, and that the vitellus afterwards accumulates round the germinal
vesicle as round a separate centre of differentiation. (See below, where this process is more fully
described in the physiological section.)

In Conjne inisilla and many other species, the ova, when escaping from the gonophore under
the pressure of the compressorium, present a peculiar ajipearance. They are then seen to be
each invested by a special membrane of great delicacy, which is continued backwards by a
narrow neck-like prolongation ; so that in this state the whole ovum presents a pyriform shape.
This membrane is probably nothing more than the vitellary membrane of the ovum, which, from
the mode in which the pressure is applied, assumes the form described.

In no hydroid ovum have I found any evidence of a raicropyle.

Spermatozoa. — The spermatozoa possess the form which so generally characterises those bodies
throughout the animal kingdom, being here in all cases active caudate corpuscles (woodcut, fig. 31
D d). The caudal filament is sometimes of such extreme tenuity as to render it very difficult of
detection, while the head varies in form, being usually conical — and then with the filament
attached to the wide end of the cone — but sometimes spherical, or cylindrical, or " guitar-shaped,"'
according to the species. In Eudendrium ramosum a very minute granule may always be seen
attached to one side of the head of the spermatozoon, where it looks like a parietal nucleus.
(PI. XIII, fig. 17.)

The spermatozoa seem to be always developed in true sperm-cells, which are themselves

^ The spermatozoa of Eudendrium dispar, Agass., and some other species, are so described by
Agassiz. • Nat. Hist. United States,' vol. iv.

THE GONOSOME.

G5

Fig. 31.

frequently contained as a brood in the interior of mother-cells, as may he very well seen in
Sertidariapoli/zoiiias, where the cells which give immediate origin to the spermatozoa form groups
of from two to four enclosed within a common mother-cell. The spermatozoon itself seems due
to the metamorphosis of the nucleus of the sperm-cell.

In Laomedea Jlexuosa I have carefully followed the progress of the spermatogenous tissue from
a very early period to the formation of the mature spermatozoon. In the very young gonophore
of this hydroid (woodcut, fig. 31 A) the spadix may be seen . surrounded by a nearly transparent
mass, which is destined to become
developed into spermatozoa, but which
presents as yet no obvious structure
beyond a minutely granular condition,
which under the action of acetic acid
becomes more distinct.

In a stage a little further advanced
(B) the gonophore has increased in size,
and the spermatogenous mass has
become more voluminous and has ac-
quired a manifest structure, being now
plainly formed of a peculiar tissue
which, when liberated from the con-
finement of the gonophore and spread
over the field of the microscope, is
seen to consist of a multitude of bodies
of a rather irregularly pyriforui or
conical shape, and about j^th of an
inch in diameter (D a). These bodies,
when set free, present for the most
part an evident vibratory movement,
which seems distinct from mere mole-
cular motion, though as yet no filament

or other source of the motion can be detected. When treated with acetic acid, they assume a
regularly spherical form, and have then all the appearance of thick-walled cells with an undoubted
nucleus in their interior (D b).

In a more advanced stage the contents of the gonophore have still further increased in
opacity, and are now seen to be entirely composed of very minute spherical corpuscles (D c) about
go^th of an inch in diameter, and presenting a close resemblance to the nuclei of the cells
composing the spermatogenous tissue in the stage last described. They exhibit distinct but
not active motion under the microscope, though no filament can as yet be demonstrated in
them.

In the next stage (C) the gonophore has attained maturity, and the spermatogenous mass
has become still more opaque than in the preceding stage, and presents a peculiar striated
appearance, the striae radiating from the sides of the spadix to the walls of the gonophore. Soon
after the gonophore has attained this condition it bursts, and allows its contents to escape into
the surrounding water as mature active spermatozoa (D d). These spermatozoa have an ovo-

9

6'0

Development of tlie Spermatozoa iu Laomedea Jie.rv.osa.

A, Very young male gonopliore, showing the spermatogenous plasma
interposed between the ectoderm and the endoderm.

B, Gonophore further advanced.

C, Mature gonophore.

D, Structure of spermatic tissue at various stages ; a, spermatic tissue
from B ; b, the same after treatment with acetic acid ; c, spermatic tissue,
from a gonophore somewhat further advanced tlian B; rf, mature sperma-
tozoa from C.

66 MORPHOLOGY.

conical licad, witli a caudal filament of extreme tenuity ; the head is about jjr,r,th of an inch in
its longer diameter, and about -j^jth in its shorter. The tail is attached to the wide
end.

In attempting an interpretation of the above appearances, we must, I think, regard the
nucleated cells which constitute the contents of the gonopliore in the second of the stages (wood-
cut, fig. 31 B) just described as spermatic cells which in the next stage have set free their nuclei ;
these nuclei, after liberation from the cells, acquiring a more elongated form, developing a
filament, and becoming converted into true spermatozoa.

Allusion has been just made to the peculiar striated appearance presented by the mature
spermatic mass while yet contained within the gonophore. This appearance, wliich is very
common in the mature male gonophores of the Hydroida, suggests to us the idea that the
corpuscles composing the mass are confined in an exceedingly fine tubular tissue. I have,
however, in vain sought for any indubitable evidence of tubes, and I believe that the appearance
in question is the result of a mere arrangement of the corpuscles — a condition induced in the
plastic mass by the pressure exerted on it by the resisting walls of the gonophore as the mass
within increases in volume ; for the component corpuscles have now become changed from the
spherical form of the previous stage to a more oval form, and their axes are compelled by the
surrounding pressure to take a definite direction. It is a phenomenon which in this view would
be purely physical, and which we cannot avoid comparing to that of slaty cleavage, though
occurrin<: in an organized and living mass.

IV. Development.

There can be no doubt that the phenomena of development, involving as these do the
changes of form undergone by the organism in successive periods of time, constitute a department
of Morphology, and should, when possible, be treated in connection with other morphological
phenomena instead of being included under the head of Physiology, as is the usual, and perhaps
in some cases the more convenient, practice.

In the account already given of various parts of hydroid organisation, it has been found
necessary to dwell with more or less detail on their development. A more systematic treatment
of hydroid development may now be attempted. This may be best considered under two heads,
the one treating of the development of the bud, the other of that of the embryo.

These two kinds of development, notwithstanding certain differences which necessitate their
separate consideration, possess close analogies with one another. The progress of the bud in its
development, like that of the embryo, is always from the general to the special; and just as it is
impossible at first to point out any difference between embryos which are destined to branch off'
into widely separated types, so in their early stages it is impossible to distinguish from one
another buds which are destined to become developed into very different forms. Thus, no
difference whatever can be detected on their first appearance between three buds, one of which is
destined to become a hydranth, another a sporosac, and another a medusa ; and the analogy will

DEVELOPMENT. 67

appear still closer when we bear in mind that buds formed by entirely difTcrcnt bud-ljcaring types
may in their early stages be undistinguishable from one another.

While the hydroid embryo, however, continues to develop as an independent organism, the
bud remains for a longer or shorter time dependent on the pai'ent stock ; but it will yet be seen
that the stage of differentiation on which the bud stands at the earliest period at wliich any
differentiation is perce])tible corresponds to that of the embryo at the moment when the ectoderm
and endoderm of the planula become differentiated as distinct structures, and the further
development of the planula and of the hydranth-bud present a close parallelism with one another.

1. Development of the Bud. — Zooidal Development.

Reproduction by budding or gemmation is the phenomenon which, of all others, most vividly
impresses us in our study of the Hydroida, and is that which confers upon this remarkaljle group
of organisms its peculiar and characteristic physiognomy. It struck with all its force the
earlier observers, and united with the flower-like form of the hydranth in suggesting the term
" zoophyte," by which the wonderful budding and blossoming plant-like animals which adorn our
rocks at low water, and are dredged up at various depths from the bottom of the sea, have long
been known to the naturalist.

Gemmation in the Hydroida has for its object, 1, the extension of the trophosome; 2, the
origination and extension of the gonosome.

The primordial trophosome (PI. I, fig. 11 ; PI. Ill, fig. S; PL XIII, fig. IG) is quite simple ;
but it soon begins to complicate itself by budding, and this complication is frequently carried to
a great extent, the primary buds giving rise to secondary buds, and these again to tertiary, while
buds of a fourth, fifth, or even higher order, may continue to be produced in succession ; and as
every bud may develop itself into a branch, the result will be the production of those complicated
dendritic groups (Pis. Ill, IV, IX, XIII, &c.) which attain to such perfection in numerous species
among the Tatjidarian, Campamdarian, and Sertidarian hydroids.

The complex trophosomes which thus result from successive buddings may present sym-
metrical and asymmetrical forms. Symmetrical forms are, as a general rule, presented throughout
the Sertularians ; the hydranths, with their hydrothecai, being in these hydroids developed upon
points which are symmetrically disposed in relation to a common axis or a common plane ; while
the ramification of the trophosome is here also usually symmetrical — distichous in most species,
verticillate in others. The Campanularians, on the other hand, and especially the Tubidarians,
present in most cases an asymmetrical disposition of their branches. The genus Pennaria among
the Tuhularians aff'ords a remarkable exception in this respect, its gemmation being so singularly
symmetrical as to give to the entire trophosome a close resemblance to that of a Plumularia — so
close, indeed, as to have led the earlier systematists to place it in that genus.

Under the general head of Gemmation, we may here consider the development of the
hydranth, the development of the blastostyle, the development of the sporosac, and the develop-
ment of the medusa.

68 MORPHOLOGY.

a. Development of the HydranUi.

It is exceedingly rare to find the trophosome retaining through life the simple condition
which it presents during its primordial state. Cases, however, of permanently simple tropho-
somes occur. We meet with them, for example, in Corpnorpka (PI. XIX) and certain aUied
forms. The curious free trophosome of Nemopsis as described by M'Crady, and of Acaulis, as
described by Stimpson, are probably only the detached hydranths of some fixed Tubularians
which may possess the habit of throwing off" their hydranths, as we know to be the case in
certain European species of Tubularia.

Ihjdranili-hud in the Gymnohlastea and Eleiitlieroblastea. — When a hydranth-bud is about to
become developed from any part of the coenosarc in the gymnoblastic hydroids, the two layers of
the coenosarc are seen at this spot to be pushed outwards as if by an incipient hernia, and the
little hollow tubercle thus produced forces before it the investing perisarc, which is first extended
over the advancing bud, and — except in the very young parts, where it is still in the condition of
a mere film — is at last absorbed or ruptured.

The little bud, however, has been in the mean time clothing itself with a new perisarc,
which, now that it has escaped from the confinement of the old one, is seen to cover it with a very
delicate, transparent, structureless pellicle. The bud continues to increase in size, becoming
longer and thicker, with its endoderm and ectoderm very distinct, and with its cavity opening
freely into that of the branch from which it springs, and admitting into its interior the fluid with
the floating granules which fill the general cavity of the coenosarc, and which are kept in a state
of active rotation within the bud. It continues to enlarge, but has its distal extremity still
closed, while the entire bud is still invested by its delicate perisarc (PI. II, fig. 5, «fec.).

We next find that the little bud has acquired a somewhat clavate form by the enlargement
of its distal extremity. While the perisarc which clothes the growing bud continues, by means
of new layers deposited upon its inner surface, to increase in thickness over the proximal part
of the bud, these new layers cease, in almost every case, at a very early period to be excreted from
the free extremity of the bud, and the perisarc here accordingly remains in the condition of a
transparent structureless pellicle of extreme tenuity, which at last, in most cases, entirely dis-
appears. We now find tentacles begin to grow out from the enlarged extremity of the bud, and
a terminal mouth to become developed ; the form is thus gradually assumed which is to charac-
terise the adult hydranth.

In some cases, however {Coryne voginata, Hincks (PI. IV, fig. 8), and Eudendrium vat/inatum,
Allm. (PL XIV, fig. 7) ), the perisarc which clothes the free extremity of the growing branch
attains considerable thickness, and does not disappear until a later period ; but it ceases in such
Gases to be in close contact with the ectoderm, and forms an outer chitinous capsule, within
which the hydranth continues to become developed ; and this development proceeds to the
formation of tentacles and the assumption, more or less, of the adult form of the hydranth-bud,
before the rupture of the enclosing capsule places the young hydranth in direct relation with the
surrounding water.

DEVELOPMENT.

GO

-our only rcpvcsentativc of the Eleulhero-

Fig. 32.

The development of the hyclranth-ljiul in Ui/di
blastea — seems to be in all essential
points the same as in the Gymnoblasiea,
the most important dififerences being
those which depend on the absence of
a perisarc in Hydra. The ultimate
destination of the bud, however, is very
different in the two cases ; for while in
the Gymnoblasiea it remains fixed as a
permanent part of the hydrosoma, it is
in Hydra destined to become detached
and enjoy henceforth an independent
existence.^

Sponianeous Decapitation and
Tie-formation of successive Hydranths.
— Our account of the development of
the hydranth-bud in the Gymnoblastea
would be incomplete without some re-
ference to a very remarkable pheno-
menon presented by certain species
of Tubularia, namely, the periodical
shedding and renewal of the hydranths.
This phenomenon was several years
ago observed by Dalyell," and described
with all his usual accuracy by this
excellent observer. I cannot find,
however, that any author has followed
the process with that exactness which
is necessary to enable us to form a
correct idea of its nature. My own
observations have been principally
made on Tubularia indivisa, where I
have bestowed upon the process in
question a very careful examina-
tion.

When the hydranth of this species, with its clusters of gonophores, has acquired full
maturity, the time is come when it is to be cast off (woodcut, fig. 32 A), and its place taken by
a successor. A breach of continuity now occurs in the endoderm of the stem at a short distance
behind the hydranth ; while the ectoderm (c) having already become detached from the endoderm [h)

Shedding and Renewal of the Hydranth in Tubularia indivisa.

A, Part of a stem in which a breach of continuity has just taken place
below the hydranth ; a, the perisarc, closely investing the proximal seg-
ment of the stem j 6, the endodermal tube of the distal segment, which
has become separated and retracted from the proximal segment and from
its own ectoderm ; c, the detached ectoderm of the distal segment ; d, the
perisarc, where it forms a loose membranous sheath, just below the old
hydranth.

B, Early stage of the new hydranth, with the commencement of the
proximal zone of tentacles; a, the perisarc, where it invests the stem just
below the new hydranth ; c, remains of the ectoderm, which had belonged
to the detached part of the stem ; d, remains of the perisarc, where it had
formed a loose, thin membranous sheath, below the old hydranth ; e, com-
mencement of new proximal tentacular zone ;_/, distal constriction, from
which the distal circlet of tentacles is to become developed ; g, the trun-
cated extremity of the proximal segment, now closed over, and containing
a cavity formed by the coalesced canals of the endoderm .

C, More advanced stage of the new hydranth, showing the formation of
the second or distal circlet of tentacles ; a, perisarc, closely investing
the stem just below the new hydranth ; rf, remains of the distal part of the
perisarc, where it had formed a loose membranous sheath ; e, proximal zone
of tentacles, now much elongated j/*, distal zone of tentacles; g, terminal
part of the proximal segment, forming the hypostome of the new
hydranth.

' The curious free reproductive bodies which occur in Corymorpha nutans, and which probably
originate as gemmae, though their exact significance has not yet been determined, will be described
below.

° ' Rare and Remarkable Animals,' 1847, vol. i, p. ■4.

70 MORPHOLOGY.

in the space between this lireach and the base of the hydranth, the npper end of the stem shps out
of its ectoderm, carrying the hydranth with it, and leaving behind it the empty ectoderm as a
thin, collapsed, membranous sheath, surrounded by the perisarc, which here exists as a delicate
loose pellicle (c).

The hydranth thus detached falls to the ground, where it retains for some time its vitality,
the gonophores which still hang from it discharging such of their contents as had not escaped
before the decapitation.

In the mean time the wound which had been formed in the ccenosarc by the detach-
ment of the hydranth heals over, and tlie truncated end of the ccenosarc becomes
closed.

Two slight constrictions, one a little behind the other, are next seen (B) to take place in the
ccenosarc at a short distance from the decapitated extremity, while the peculiar tubular lacunae
which exist in the ccenosarc of the Tuhidaria, and which had hitherto extended as separate canals
through the whole ccenosarc of the stem, now coalesce in front of the anterior constriction (/), where
they form a single cavity by the breaking down of the partitions of endoderm which had upon
this time separated them from one another.

A girdle of minute tubercles (<?) may next be seen budding forth from the co-nosarc, at the
site of the posterior constriction. These soon become extended into tentacles (C e), which embrace
the distal part of the ccenosarc.

In the next stage a similar zone of tubercles (/), becoming, like the others, elon-
gated into tentacles, shows itself close below the anterior constriction ; and there are
thus established the two sets of tentacles, the proximal and distal ones of the new
hydranth.

By the elongation of the ccenosarc from below, the new hydranth is gradually lifted up out
of the tube of the perisarc, when the tentacles, having room to expand, immediately fall into
their normal position, while the rudimental clusters of gonophores may be seen as minute
lobulated elevations between the anterior and posterior series of tentacles, and the free extremity
of the hydranth has by this time become perforated by a mouth.

The hydranth now increases in size, raised higher and higher on the elongating ccenosarc,
which clothes itself with a perisarc as it lengthens, and the hydranth with its clusters of gono-
phores, having finally attained complete maturity, is then in its turn cast off, to be succeeded in an
entirely similar way by a new one.

The formation of successive hydrauths is always accompanied by a periodical elongation of
the stem, and this is indicated by annular markings of the periderm separated by rather wide
intervals, each interval corresponding to a single decapitation and renewal.

From the above description it will be seen that the formation of successive polypites is
not so much a process of ordinary budding, as a true metamorphosis of the decapitated extremity
of the ccenosarc.

Polarity of the Hydroid. — In connection with the phenomena now described, those which
accompany the artificial section of the stem deserve special notice. When the stem is cut across,
the ccenosarc of the upper segment soon heals over at the place of section, the tubular lacunae
become again closed, and the ccenosarc now begins to grow downwards through the cut extremity
of the periderm, presenting the same lacunar structure as in the older portions, and excreting
upon its surface a very delicate perisarc. The well-known cyclotic currents may generally be seen

DEVELOPMENT. 71

with great distinctness in tlic fluid wliicli fills tlie tubular lacunas of the young elongated
coenosarc.

Tiie lower segment, on the other hand, instead of pushing forth from the cut extremity a
simple continuation of the coenosarc, develops from this extremity a liydranth.' There is thus
manifested in the formative force of the Tuhularia-stem a well-marked polaritij, which is rendered
very apparent if a segment be cut out from the centre of the stem. In this case, no nuvttcr in
what position the segment may lie, that cud of it which was directed downwards or proxiuially
while it formed a part of the uimiutilatcd hydroid will never develop a hydranth, b>it will extend
itself as a simple cylindrical prolongation of the coenosarc ; while the upper or distal end, instead
of becoming simply elongated, will shape itself into a true hydranth ; and all this though, of
course, not the least difference in structure or form, can be detected between the two extremities
at the time of section.

It is further manifest from these facts that, when the hydroid is placed under conditions
which allow of perfect freedom of growth, there is no such thing as a stationary extremity,
both ends being really growing ends, while there exists in every segment a neutral plane midway
between the two ends.

Hydranth-hnd in the Calyptohlastea. — In the development of the bud, the Campanularian
and Sertularian hydroids differ in some important features from those which characterise the
process just described. The development may be easily watched in many species, as, for example,
in Laomedea flexmsa. We may here (woodcut, fig. 33) see it proceed, in the first place, to the
formation of a hollow cylindrical branch («), whose cavity is in free communication with that of
the coenosarc, and whose distal extremity ends in a cul-de-sac invested, like the rest of the young
branch, by the chitinous perisarc. Up to this point the phenomena are precisely similar to what
we have just seen in the Tuhularians ; but now the distal extremity of the branch begins to enlarge,
and at the same time continues to coat itself with a chitinous perisarc in the form of a capsule,
which acquires increased thickness by successive deposits of new matter to its inner surface,
thus contrasting with the much thinner pellicle which forms the temporary capsule in certain
Tiibidarians.

The extremity of the branch [Ij] now presents the shape of an inverted cone, plainly recognisable
as the body of the budding hydranth, invested with a strong chitinous covering, which is closely
applied over its whole surface, and is continuous below with the perisarc covering the rest of the
branch. The interior of the young hydranth is hollowed out into a wide cavity lined by a
layer of loose cells — the most internal cells of the endoderm — which are filled with a granular
pigment.

The conical enlargement at the extremity of the branch continues to increase in size (c), and
we soon see the soft parts within become contracted towards the proximal end of the cone, where
they withdraw themselves from contact with the walls of the chitinous capsule, which had up to
this time closely embraced them. At the wide or distal end of the cone they still remain adhe-
rent to the capsule for some distance downwards, while at the proximal end itself there is also
a distinct but narrow zone of contact and adhesion maintained between the internal soft parts

' The first observations on this subject are those of Dalyell, who has made numerous experiments
on the section of the stem in Tu/mlaiia indivisa, with results similar to those here recorded. (See
Dalyell, ' Rare and Remarkable Animals,' vol. i, p. 23.)

72

MORPHOLOGY.

and the external cliitinous capsule. In the cavity which occupies the interior of the soft contents
of the capsule very distinct rotating currents may be now seen, excited, doubtless, by the action of
vibratile cilia, though a direct view of these cilia cannot be obtained through the thickness of the
walls.

Fig. 33.

Development of Hydranth and Hydrotbeca in Laomedea flexuosa.

a. Very early condition, in which the bud forms a simple cylindrical csecal oflset from the
coenosome.

J. The distal extremity of the bud has become enlarged, so as to present the form of an
inverted cone.

c. The cone has increased in size, and the soft parts towards its proximal end have become
retracted from the external chitinous walls.

d. The internal structures have still further withdrawn themselves from the chitinous walls,
with which they are uow in contact only by a narrow proximal and a wider distal zone, between
which they present the form of a tubular cylindrical column.

e. The distal zone of contact has become retracted from the summit of the cup-like_ envelope
of chitiue, tentacles have begun to sprout from its circumference, and a hypostome has risen from
its centre. The leading features of the completely formed hydranth are thus established, and its
chitinous envelope has become the hydrotbeca.

The arrows in the figures indicate currents in the somatic fluid.

Between the proximal and distal zones of contact the internal structures become more and
more withdrawn from the walls of the capsule, while the whole body continues to elongate {d) ; and
this may now be seen in the form of a cylindrical column occupying the axis of a conical cup of
chitine, and expanded below into a narrow ring, which at this point connects it with the walls of
the cup, while above it expands into a broad disc, which fills up the distal extremity of the cup
like a lid or plug. The axis of the column is permeated by a tubular cavity in continuation
below with the cavity of the branch, and expanding above into a wide chamber, which occupies
the interior of the plug-like enlargement of its distal end. It is now plain that, while
the soft contents of the cup are the developing hydranth, the cup itself is to become the
hydrotbeca.

The excreting of the chitine and the shaping of the hydrotbeca would seem to devolve on
the terminal plug-like disc alone, from the time that the lower parts of the nascent hydranth liad
withdrawn themselves from contact with the walls of the external capsule ; and as the hydranth
continues to elongate itself, the surrounding cup is extended at the same rate, by addition
to its wider end from the sides of the disc, while the lower parts of the cup undergo little or no
change.

DEVELOPMENT. 73

The upper surface of the disc has been all along covered with a thin layer of chitiuc, whose
periphery is continuous with the chitinons walls of the cup, but which does not interfere with the
growth of the young hydranth ; for as the latter continues to extend itself, the layer of chitine on
the upper surface of the disc is carried onwards before it, without becoming thereby detached
fi'om the side of the cup — a fact which we can scarcely explain otherwise than by supposing con-
siderable extensibility in the recently deposited chitine of the cup. At hist the hydrotheca has
attained its complete size and shape, and now the young hydranth becomes more or less retracted
within it, the terminal plug-like disc withdrawing itself from the layer of chitine which it had
excreted on its upper surface, and which is now left behind as a roof closing over the mouth of
the cup.

Tiie whole circumference of the retracted disc now begins to develop a circle of minute
tubercles (e), which gradually elongate themselves into short thick tentacles, while the central part
becomes elevated into a blunt conical proboscis (hypostome), and the cylindrical tubular column
which occupies the axis of the hydrotheca has become dilated into a more oval-shaped body, with
a wide internal cavity — the stomach of the developing hydranth.

The young hydranth, still included within a completely closed cup, presents greater and
greater contractility, now withdrawing itself towards the bottom, and now extending itself through
the entire height of the surrounding cup. The tentacles in the mean time have become longer,
the extremity of the terminal cone has become perforated by a mouth, and at last the hydranth
pushes off the chitinous roof of its hydrotheca, and emerges into free contact with the surrounding
water.

h. Bevehpment of the BlastostyJc.

In the gymnoblastlc genera the development of the blastostyle is essentially similar to that of
the first stages of the hydranth. Instead, however, of proceeding to the development of prehensile
tentacles, an arrest takes place, sporosacs bud from its sides, and the nutrition of the colony, to
which the hydranth is destined, becomes replaced by the duty of supporting the sexual buds. In
the calyptoblastic genera the development of the blastostyle is accompanied by some additional
features which render necessary a more detailed description.

Laomedea Jlewuosa wiU afford here too a very convenient subject for tracing the process of
development. The blastostyles of this hydroid arise close to the axillae of the branches, and
present the form of a long cylindrical column, expanded at its summit into a disc, occupying the
axis of a spacious gonangium, and carrying along its whole length adelocodonic gonophores, which
increase in maturity as they approach the summit of the column. The whole is elevated on a
short annulated peduncle.

The blastostyle here originates in a bud precisely in the same way as a hydranth ; and up to
the stage to which we have already followed the development of the hydranth and hydrotheca,
when these parts present the condition of a conical enlargement of the extremity of the branch,
there cannot be found any difference between the hydranth-bud and the blastostyle-bud. It would
seem, however, that at this stage the soft parts, instead of absolutely withdrawing themselves
from contact with the external chitinous capsule, present in their ectodermal layer a number of

10

74 MORPHOLOGY.

lacunte, whicli, increasing in size, become confluent with each other, and the ectoderm thus
becomes split into two layers by a true ckorization. Tlie external layer remains in contact with
the chitinous capsule, while the internal layer, remaining adherent to the endodenn, becomes
more and more withdrawn towards the axis of the bud, where it now constitutes the external
or ectodermal layer of an axile column or blastostyle. The capsule thus becomes lined with a
thin layer of ectoderm, which is continuous with the ectoderm of the blastostyle only at its distal
and proximal extremities, these two membranes being in the whole of the intermediate region
separated from one another by a wide interval. This interval, which constitutes the cavity of the
developing gonangium, is thus nothing more than a large lacuna ; and it is in this lacuna that
the sporosac or blastocheme now begins to bud forth from the axile column. The excreting and
modelling of the chitinous gonangium would seem to devolve for some time still on the ectodermal
lining instead of being, as in the hydranth-bud, transferred at a very early period exclusively to the
disc-like summit of the axis. After a time, however, the lining membrane entirely disappears, and
henceforth the excreting and modelling of the gonangium seems to devolve on the terminal disc
of the blastostyle. While the gonangium is yet young, numerous irregular fleshy bands may be
seen stretching across the cavity from the blastostyle to the external wall. These bands are the
remains of the original union between the two layers into which the ectoderm has split. They
are generally torn, and disappear as the gonangium, increasing in size, has its walls more and
more widely separated from the blastostyle ; but they are also occasionally more or less visible
in the full-grown gonangium.

A compai'ison between the developing hydranth and its liydrotheca, on the one hand, and
the developing blastostyle and its gonangium on the other, aflbrds a most instructive parallelism,
showing the close connection between the hydranth and the blastostyle. If in the hydranth-bud
the development were arrested at the point to which it arrives just before the terminal disc has
withdrawn itself from the roof of the young hydrotheca (woodcut, fig. 33 d), in order to develop
its tentacles, we should have in almost every particular a gonangium with its blastostyle (see
woodcut, fig. 18). The development of a mouth and tentacles, however, points towards a different
destination ; and now, instead of producing zooids destined for generation, the hydranth applies
itself solely to the nutrition of the colony.

The gonangium does not always present the simple form which we find in Laomedea Jlexuosa,
and we have already seen the remarkable modification which it undergoes in the female colonies
of Sertularia rosacea, S./allax, and S. tamarisca, by the formation of a marsupial chamber for the
protection of an extra-capsular sac, in which the ova are retained during the earlier periods of
their development.

c. Development of the Sjjorosac.

The development of the sporosac or adelocodonic gonophore, in its simplest form, may be
easily studied in Hydractinia ecMnafa. In this hydroid the gonophores are borne on a blastostyle
(PL XV, and woodcut, fig. 4 bb, c), which here, just as in the blastostyle of the Sertidarians and
Camjianularians, is morphologically nothing more nor less than an arrested hydrantii, but in
Hydractinia never developing a gonangium.

DEVELOPMENT. 75

In their earlier stages the sporosacs may be seen as minute hollow tubercles, projecting
from the sides of the blastostyle. They are composed of two layers, cndodcrm and ectoderm,
directly continuous with the corresponding layers of the blastostyle, with whose ca\'ity that of
the young bud is in free communication. At first we can detect no change beyond a simple
increase in size ; but wc soon find the ectoderm separated from the cndodcrm by the interposition
of a minutely granular mass between them. This mass constitutes the basis of the generative
elements, and is afterwards to become ova or spermatozoa. In the mean time the ectoderm has
itself become difierentiated into two layers ; and we have thus laid down the foundation of all
the parts which we meet with in the full-grown gonophore. The wall of endoderm which
surrounds the central cavity of the developing gonophore, and is itself immediately surrounded by
the generative elements, is the spadix ; the more internal of the two layers into which the
ectoderm has divided is the endotheca, the more external the ectotheca.

The sporosac now becomes more and more distended by the increasing volume of the
generative mass, while the spadix at the same time continues to grow, and now constitutes a
club-shaped hollow organ, extending through the axis of the mass, while fioating particles from
the cavity of the blastostyle are freely admitted into its interior, where they may be seen per-
forming active rotatory movements.

The sex of the gonophore becomes evident at an early period, by the appearance of ova with
their germinal vesicle and spot in the generative plasma of the female, while in the male the
interval between the spadix and endotheca continues still to be occupied by a uniform grumous
plasma, in which, at a somewhat later period, spherical cells and ultimately free-moving
spermatozoa may be detected.

The gonophore of Ilijdractima ecMnata does not pass to any higher grade of development
than that here described ; but in some other forms of adelocodonic gonophore a further differ-
entiation takes place by the development of an additional membranous sac or mesotheca, with
gastrovascular canals, between the endotheca and ectotheca {Tubularia indivisa — PI. XX, fig. 3,
and PI. XXIII, figs. 8 and 11). I have never succeeded in following the development of the
mesotheca, and cannot say under what condition it begins, or how it proceeds, the membrane
appearing always fully formed from the moment it is recognisable.

It will be seen that in the above account I differ in some important points from the inter-
pretation given by Agassiz to the appearances which present themselves in the development of
the adelocodonic gonophore. In his account of this process in his Clava leptodyla, Agassiz'
regards the perigonium or walls of the gonophore as simple, and as homologous with the umbrella
of a medusa. In Clava nwUicoriiis, however, the existence of two membranes may with care be
demonstrated in these walls, though I admit that I have frequently failed in detecting more than
a single one. In no case, however, can the walls of the gonophore in Clava be regarded as the
homologue of an umbrella. When two membranes can be demonstrated in them, these will be
an endotheca and ectotheca ; if only a single membrane be present, as Agassiz l)elieves to be the
case in his Clava leplostyla, this wOl be an endotheca, while the part which would really
represent an umbrella, namely, a mesotheca, is not developed."

> Op. cit. vol. iv, p. 221.

' In my earlier researches into the anatomy of the reproductive system iu the Hydroida {" On
the Anatomy aud Physiology of Cordylophora," Phil. Trans. 1853), I entertained the view here advo-

76 MORPHOLOGY.

Again, in the gonopliores of Ihjdractimn jmJ^cHna, Agass., TuhiJaria (Paryplia) crocea,
Agass., and Tnhularia {Thamnocnidia) sjiedabUis, Agass., Agassiz correctly figures the two
membranes which enter mto their walls ; but he assuredly assigns an incorrect origin to the more
internal of these membranes when he describes it as rising, subsequently to the formation of the
generative mass, from the proximal end of the gonophore in the manner of a cup closely pressed
against the outer wall, and, at least in Hydractinia and Tubularia spectabiJis, ultimately closing
over the contained structures so as to form a continuous internal wall.

Now, the internal wall in the gonophore of Hydractinia is undoubtedly formed, not after,
but simultaneously with the appearance of the generative mass, and is nothing more than the
internal of the two layers into which the ectoderm of the primary bud has become divided
simultaneously with its separation from the endoderm by the interposition of the generative
elements ; it is thus the endotheca of the sporosac, while the more external layer is the
ectotheca.

Having had no opportunity of examining the development of the gonophores in either of the
two Tubularice cited above, I am unable to bring any direct observation into opposition with the
views of Agassiz as to the gonophores of these hydroids ; but the analogy of Hydractinia and of
other hydroids, whose adult gonophores correspond in all essential points with those of the
American forms, leads me to believe that the process is in all the same as in Hydractinia.

It is only in those cases where a mesotheca becomes developed, as in Tubularia indivisa,
that the adelocodonic gonophore presents any true representative of the umbrella of a medusa,
the mesotheca being properly the homologue of this part. Agassiz, in his account of Tubularia
Couthouyi, Agass., ignores the existence of any membrane between the well-developed mesotheca
of this species and the generative mass which surrounds the spadix. In Tubularia indivisa,
however, this membrane cannot be overlooked, especially in the male, though in the female it
would seem to disappear at an early period, and may thus escape detection.

d. Development of the Medusa.

The medusa, whether gonocheme or blastocheme, shows itself at first in every case as a
minute hernia (woodcut, fig. 34 A), consisting of endoderm and ectoderm, and having its
cavity in free communication with that of the blastostyle, or of the trophosome from which it
springs ; thus in no respect differing at this period from the corresponding stage in the develop-
ment of the adelocodonic gonophore, or, indeed, in that of a hydranth branch.

It is very difficult to follow satisfactorily the several steps by which this primordial tubercle
becomes ultimately converted into a medusa. I have bestowed great attention on it in different
species of Hydroida, and have more recently subjected the development of the medusa bud

cated by Agassiz, as to the homology of the parts in question. Subsequent more extended observa-
tions, however, have induced me to modify in some respects the views then expressed, and to adopt
those which are advocated in the present INIonograph. (See my paper " On the Reproductive Organs of
Serlularia tamarisca,'' in the Report of the British Association foji- the Advancement of Science,
1858.)

DEVELOPMENT.

77

in Corijmorpha nutans to a laborious examination, wliich has led me to adopt the process
now about to be described, as the true interpretation of the phenomena presented in this
hydroid.

We first find that four equidistant processes (woodcut, fig. 31 B c), consisting of cndodcrm
and ectoderm, with an included cavity, which is a continuation of that of the hernia-like tubercle just
mentioned, have begun to grow upwards from a circle round the summit of this primordial bud.
These, however, do not show themselves as free processes ; for simultaneously with their appear-
ance the ectoderm of the summit of the bud becomes split into two layers {«', a"), which become
more and more widely separated from one another as the processes continue to elongate, the
outer layer arching over the space which is surrounded by the four processes. During this
elongation the ectodern which occupies the four intervals between the roots of the processes is
carried upwards as a continuous membrane, stretching across from one process to another in the
manner of a web.

EiG. 34.

Development of the Planoblast in Con/morfha nutans.

A, Very early stage of the medusa-bud when it presents the form of a simple hernia-like tubercle, whose cavity is in com-
munication with the somatic cavity of the hydroid; a, ectoderm; I, endoderm.

B, More advanced stage, a, ectoderm ; b, endoderm ; «', the more external of the two layers, into which the estoderm
of the bud has split; a", the more internal of these two layers; c, commencement of radiating canals.

C, Stage still more advanced, a, ectoderm ; b, endoderm ; c, radiating canals ; d, manubrium.

D, Transverse section of C. a, ectoderm ; b, endoderm ; c, radiating canals ; d, manubrium.

E, A stage still more advanced than C. The distal ends of the radiating canals have become dilated into bulb-like expan-
sions, one of which has begun to extend itself as a marginal tentacle.

F, Transverse section of E.

G, A stage somewhat more advanced than E. The bulbous terminations of the radiating canals have coalesced, and one
of them has become projected into a thick tentacle.

H, Medusa, just after liberation from the trophosome.

The result of this is, that we have now the distal portion of the bud in the form of a deep
cup, closed over by a layer of ectoderm, and having its walls traversed by four equidistant csecal

78 MORPHOLOGY.

tubes, whose Ccavity is continuous with the original cavity of the bud, and which are lined by a
continuation of the endodcrm of the bud. There is no difficulty in recognising in these tubes the
radiating canals of the future medusa, and in the web of ectoderm which unites them the
umbrella.

From the central point of the area included between the bases of the fom- canals another
hernial process (C d and D d) has already begun to make its appearance, composed of ectoderm
and endoderm, and containing a prolongation of the original bud cavity. It advances as a thick
process in the axis of the cup, and is at once recognisable as the future manubrium.

The four peripheral processes continue to elongate, and are soon seen to be dilated into
bulb-like expansions at their extremities (E, F). The bulbs increase in size, and come in
contact by their sides ; while one of them, enlarging much more rapidly than the three others,
gives a marked preponderance to its side of the bud, and makes the distal end of the bud appear
as if obliquely truncated. It then begins to extend itself beyond this distal end into a thick,
hollow tentacle.

In the mean time the four bulbs which had come in contact have coalesced, and their
cavities now communicate with oue another (G) ; but, by the gradual enlargement of the
distal end of the bud, the bulbous ends of the radiating canals are again drawn away from one
another. The communication, however, between their cavities is not thereby interrupted, but
continues to be maintained by a tubular elongation of their original points of union ; and in this
tube we now recognise the circular canal of the medusa.

The cavity of the umbrella is still closed by the more external of the two laminas into which
the ectoderm had originally split at the distal end of the bud. In the final stage this lamina is
either perforated in the centre, in order to form the velum, or, what I now believe to be more
probable, it entirely disappears, and the velum is formed by a centripetal extension of the
ectoderm on a plane with the bulbous extremities of the radiating canals, at the time when
these bulbs are withdrawn from contact with one another, in order to form the circular
canal.

The manubrium, previously imperforate, has now acquired a mouth at its extremity. The
solitary tentacle, too, has now become elongated, and presents its characteristic moniliform struc-
ture, the umbrella rapidly contracts and expands with vigorous systole and diastole, and the
medusa at last hangs on its stalk, a true Steenstriij)ia, ready to break away from the restraint of
its fostering hydranth, and enter upon an independent existence (H, and PI. XIX, fig. 5).

From the above account of the development of the medusa bud, it will be seen that here
also I am not entirely in accordance with the views expressed by Agassiz on this subject. The
distinguished American naturalist gives a very detailed account of the process as he has inter-
preted it in the development of the medusa-bud springing from his Syncoryne mirabilis, and in
which he describes this development as starting with the separation of the endoderm from the
ectoderm in the primordial tubercle, and the inversion of the endoderm into itself, so as to form
the cup of the future umbrella. " In doubhng on itself the retreating fold does not press
closely on all points upon the stationary one, but leaves four equidistant spaces into which the
chymiferous fluid penetrates."^ These four spaces are the foundation of the four radiating canals,
which would thus originate in an entirely different way, and have a significance entirely different

' ' Natural History of the United States,' vol. iv, p. 192, &c.

DEVELOPMENT.

79

from what my observation of the process in Coryinorpha and other genera has led me to regard
as the true one.

M'Crady^ believes that those medusae which occur among the gymnoblastic hydroids, where,
as we have already seen, they belong to the type of the gonocheme, arc developed in a different
way from those which " we find among the Campanularians, and where they present the type of
the blastocheme. He describes the umbrella in the former as produced by an excavation of

Fig. 35.

Segment of a young JEquorea captured off the west coast of Scotland, showing the developmeut
of the radiating canals.

a. Mouth J 6, frill-like Hp ; c, floor of the greatly expanded base of the manubrium, from
whose circumference the radiating canals are emitted; rf, d, primary radiating canals; e, e, canals
developed later, and already united with the circular canal ; e', e', canals which have not yet reached
the circular cunal ; f,f,f,f,ca.m\\3 still less developed, on their way to the circular canal;
g,ff, developed marginal tentacles; h, h, h, rudiments of marginal tentacles; a, z, lithocyets ;
k, k, margin of velum.

the substance of tlie young hud, forming thus a completely closed cavity in which the manubrium
is included, and which only at a subsequent period becomes perforated at its summit to form the

Op. cit. p. 110.

80 MORPHOLOGY.

orifice of the umbrella. In the CamjmnuJarians, on the other hand, he believes that the umbrella
grows up from below as a ring round the manuljrium, which is thus never included in a closed
cavity, but is from the first directly exposed to the surrounding medium. In accordance with
these views, M'Crady divides the gymnophthalmatous or hydroid medusae into the " endostomata"
and the "exostomata." My own observations, however, will not allow me to adopt this division
of M'Crady. In the medusa of Campanularia at all events the development is essentially the
same as that just described in the medusa of Corymorpha.

The medusa has not necessarily attained its complete development at the time when it has
become fitted for an independent existence, and has detached itself from the trophosome in order
to spend its future life in the open sea. It is very common to find both tentacles and lithocysts
less numerous at the time of liberation than at a more advanced period ; while in some cases
[^quoria) the radiating canals continue to increase in number with the age of the free
medusa > (woodcut, fig. 35).

In every case in which I have had an opportunity of observing the formation of new
radiating canals, these have been developed in a centrifugal direction. They commence as ofisets
from the base of the manubrium (woodcut, fig. 35,/), or from the previously existing canals,
and then becoming elongated in the gelatinous substance of the umbrella, they direct themselves
towards the umbrella margin until they meet the circular canal with which they inosculate. This
penetration of previously formed tissue by the nascent canals, their invariable maintenance in it
of a definite direction, and their inosculation with a canal already completed, are phenomena not
without their general significance in the formative forces of living beings.

In some cases still more striking transformations have been witnessed in the free medusa.
Thus Gegenbaur observed that the Tracliynema ciliatam, Gegenb., a medusa not yet traced to
a polypoid trophosome, is in its young state a free-swimming flask-shaped body, with three or
four minute tentacles in a circle round the base of its contracted neck-like portion, and with a
clothing of vibratile cilia over its whole surface. It subsequently developes an umbrella
and gastrovascular canals, and becomes provided with numerous imperfectly contractile
tentacles."

It is, however, in the family of the Geryonidm to which the JEginida, as follows from
Haeckel's observations, must now, notwithstanding their very diSerent form, be united, that we
meet with medusae which, during their free state in the open sea, undergo the most striking
change, passing through a series of metamorphoses which consist, not only in the development
of new parts, but in the loss of organs which, being destined to enjoy only a transitory existence,
disappear, as is described below, to make way for permanent ones of an entirely different form.
It is true that none of these medusae have as yet been traced to a hydraform trophosome ;
but they are not on that account of less importance in the general history of hydroid
development.

^ Alexander Agassiz has shown that in their order of succession the marginal tentacles of the
Hydroid Medusse obey a law very similar to that which Milne-Edwards and Jules Haimes have
shownjto regulate the formation of the successive septa in the Actinozoa. A. Agassiz in ' Proc. Bost.
Soc. Nat. Hist.,' vol. IX, Aug. 1862.

" ' Generationswechsel,' p. 51.

DEVELOPMENT. SI

An exceedingly interesting case of metamorphosis in an yEginidian medusa has been
described by jM'Crady.^ He observed, lying free in the umbrella-cavity of an Oceanidan medusa,
to which he gives the name of Tarrifopsk mitricula multitudes of little organisms, presenting various
forms, from that of a minute club-shaped hydroid to that of a well-developed medusa belonging
to the type of the ^Ej/inida, and all undoubtedly connected with one another as stages of a simple
developmental process.

Though he at first believed these to be the proper offspring of tlie Turrilopsis in which they
occurred, he afterwards rejected this notion, and recognised in them the young of a species of
Cuiiina {Cuniiia odonaria, M'C), which had selected the umbrella-cavity of tlie Oceanidan in
order to spend there as parasites the early stages of their existence.

The untentaculated, club-shaped larva (the earliest stage observed) was followed by a
bitentacular hydroid form with long imperforate proboscis and distinct internal digestive cavity,
and he noticed the interesting fact that this bitentacular stage freely repeats itself by budding.
Next, two other tentacles make their appearance synuuetrically between those first formed, while
the extremity of tlie proboscis seems now to be perforated by a mouth. The umbrella next begins
to make its appearance by an annular extension of the circumference of the body close to the oral
side of the roots of the tentacles ; and four new tentacles begin to sprout between those already
formed, while lithocysts become developed on the margin of the incipient umbrella. After
this the larva assumes the form of an adult Cunina in all essential points, except in the pos-
session of a long proboscis, like that of a Geryonia, in which stage it leaves the umbrella-cavity
of the Turritopsis to spend a free life in the surrounding water. It is only after it has quitted the
medusa on which it had been hitherto living as a parasite that it loses its proboscis, and that the
digestive cavity thereby assumes the form characteristic of the family of the ^!Erjinidce. M'Crady
views this case as presenting an instance of direct development from the ovum, believing that the
Cunina originally gained access to the umbrella of the Turritopsis in the condition of a free-
swimming planula.

Fritz Miiller- has given an excellent account of the metamorphosis of a Geryonidan medusa,
Liriope cathariensis, Fr. Miiller; and his observations have been confirmed and extended by
Haeckel,^ who has described similar metamorphoses in two other Geryonidans, Glossocodon {Liriope)
euryhia and Carmarina [Geryonia) hasfata. In all these cases the medusa in its earliest observed
condition was found swimming free in the open sea. The youngest medusa noticed presented
the form of a minute hyaline gelatinous sphere ; on one point of whose surface was a small
pit-like depression closed over by a perforated diaphragm ; and the most striking feature in the
subsequent metamorphosis consisted in the development of two sets of peculiar larval tentacles,
of which one or botli sets were destined after a period to disappear, their place being supplied by
an entu'ely different set, which remained as permanent organs during the life of the medusa.
The larval tentacles are solid and rigid, and have no connexion with the gastrovascular system ;
while the permanent tentacles are hollow offshoots from the circular canal, and are eminently
flexible and extensile.

While we are entirely ignorant of the origin of the free-swimming bodies which have been

' Proc. Elliott, ' Soc. Nat. Hist. Charleston, 1856,' p. 55.

= ' Wiegraann's Archiv,' 1859, p. 310.

' Haeckel, ' Die Farnilie der Riisselquallen.'

11

82

MORPHOLOGY.

Fig. 36.

Medusa, probably young Lizzia, captured in the open
3a, with luedusa-buds springing from the base of the
lanubiiura.

thus traced through a series of metamorphoses into adult Gcrvonidans, it has been proved that

certaiu other free meduscE of the /Eginidan type
have originated as buds from achdt forms. This,
however, leads us to consider the formation of hiids
by the medusa.

Formation of Buch hi/ the Medum. — The
plienomenon of medusa-budding does not neces-
sarily find its extreme term in the formation of the
medusa itself. Many free-svviniming medasoe, some
of which are known to have originated in liydroid
trophosomes, complicate themselves by gemmation,
which manifests itself in the production of other
medusa-buds upon various parts of their bodies.
A fine example of this phenomenon is afforded
by the medusa of the tubularian hydroid, Ili/bu-
codon prolifer, Agass. In this beautiful animal,
Agassiz^ describes the base of the solitary tentacle which is continued from the distal extremity
of one of the radiating canals of the medusa as itself producing a cluster of medusa-buds,
which in time assume the form of the primary medusa, and may themselves repeat the same
process, through the production of successive broods of similar buds, before they become detached
as free natatory medusae. Steenstrup- has observed buds developed from the base of the tentacles
in a medusa which he believes to have originated in a Coryne-like trophosome, which he names
Coryne fritillaria ; Greene has described the production of buds, not only from the bulbous base
of the tentacles, but also along the course of the tentacles themselves in a nearly allied medusa,
Biplura, Greene ;' while the emission of buds by medusse has also been described by Forbes,'
Sars,^ Busch,^ and others.

I have several times witnessed this phenomenon in medusae captui'cd while swimming in the
open sea. In some of these cases the buds were borne on the base of the manubrium (woodcut,
fig. 36), in others, on an elongated tubular peduncular extension of the manubrium (woodcut,
fig. 37), and in others upon the bulbous bases of the marginal tentacles (woodcut, fig. 38). The
singular ambulatory medusa of Clavatella also multiplies itself by budding from the interten-
tacular spaces on the umbrella-margin (see PL XVIII, fig. 5). In these various cases the buds
seem destined to assume the form of the medusa which save origin to them, l)ut observations on

' Op. cit. vol. iv, p. 24"), pi. 24.

" J. J. Steeustrup, ' Alternation of Generations,' p. 26 ; Roy. Society's Translation,
1845.

^ J. R. Greene, in 'Nat. Hist. Rev.,' 1857, vol. iv. The Medusae is there named Diplonema, but
from this name having been already given to a genus of plants, it was subsequently changed by
Greene to Biplura. See my paper " Ou the Genera of the Hydroida," in 'Ann. Nat. Hist.' for INIay,
1801.

* ' British Medusse.'

^ ' Fauna lit. Norveg., erste Lieferung.'

^ ' Beobachtungen ueber Wirbcllos. Seethieren.'

DEVELOPMENT.

S3

their history after detachment from the parent-medusa are still wanting, in order to enable us to
speak positively on this point.

In certain medusae belonging to the family of the Gcryonida it would seem, from the
observations of Gegenbaur,' Kollikcr," Krohn,'' Fr. Miiller,'' Keferstein and Ehlcrs,'^ and, above all,
from the remarkable researches of Haeckel," that the formation of buds within the cavity of the
stomach is a constant and normal phenomenon. It would further appear that these buds, for the
most part, detach themselves while still in a very immature state, and that after becoming free
they undergo a metamorphosis before arriving at their adult condition ; and, still further, it has
been shown that in at least some of these cases there is a hcteromorphisni, the b\ids becoming
developed into a form very different from that of the medusa which gave rise to them.

Fig. 37.

Fig. 38.

Sarsia, captured in tlie open sea, with
medusa-buds borne by tbe manu-
brium.

a, Wide oral extremity of the manu-
brium; b, b, attenuated proximal por-
tion of the manubrium, carrying the
buds in various stages of development.

Medusa, probably the planoblast of a Sifncorynet captured
in tbe open sea, and bearing clusters of medusa-buds on
the bulbous b;ises of the marginal tentacles.

This last phenomenon has been witnessed in a case recorded by Fritz Miillcr,' who describes
the formation of cihated buds from the internal surface of the stomach in an eight-tentacled
Cunina, which he names C. Kollikcri. He traced these buds through various stages until he

' ' Generationswechsel/ p. 56. " ' Zeit. f. wisscn. Zool.' 1843, p. 327.

^ 'Archiv, fur Naturgesch.' 18G1, p. 1G8. * Thid., p. 51.

^ ' Zoologische Beitrage/ 1861.

^ 'Die Familie der lliisselquallen (Geryonida)/ 1865, p. 115, &c.

^ ' Wiegmann's Archiv/ 1861.

84 MORPHOLOGY.

saw them detach themselves and swim free in the cavity of the stomach. Here tliey
underwent further development, which he continued to observe until he saw them trans-
formed into true Cunina, differing, however, from the parent by the fact of their having
twelve tentacles and twelve stomach-pouches, instead of eight, the number characterising
the medusa which gave origin to them. Beyond this point Miiller lost sight of them, and we are
accordingly ignorant of their further changes and destination.

But in no recorded case of the production of one medusa from another by budding is the
heteromorphism between the budding medusa and the buds produced by it so striking, and
nowhere has it been so fully traced as in the observations of Haeckel, described in his remarkable
memoir on the Gerjjonidm, in which he has shown that a sexually mature Geryonidan medusa
{Carmarina liastata, Haeckel), having its segments a multiple of six, produces upon the tongue-like
process which in this genus projects from the fundus of the stomach into its cavity a multitude
of buds which become developed, not into a six-rayed form resembling that of the Geryonidan
which gave rise to them, but into true Ciaiince, ^ginidan medusae having eight instead of six
elementary body-segments, and like all the yEginidan medusae belonging to a type which had
been previously regarded as possessing no relation whatever with the Geryonida.

It is not alone in the fact that the buds belong to a very different medusa-type from the
budder that the phenomena described by Fritz Miiller and by Haeckel present features peculiar
and anomalous ; for the situation of the buds within the stomach cavity of the bud-producing
medusae is without parallel in any other group of Hydroida. In every case where medusa-buds
have been observed among other famihes of the Hydroida, the somatic cavity of the bud has
been in communication with some part of the somatic cavity of the hydroid which produces
it ; while here such a communication is impossible before the development of the mouth in the
bud shall enable the young ^ginidan to receive nutriment through this orifice from the stomach
cavity of the supporting medusa.

Two other cases, however, both among the ^Eginidan type of the Geryonida — namely, that
of Cunina proli/era, described by Gegenbaur,^ and that of jEyineta gemmifvra, described by
Kcferstein and Ehlers' — have also been recorded, in which the young luedusse are formed as buds
within the cavity of the stomach, in both of these instances the buds having been developed from
the internal surface of the stomach walls. In all these cases the buds must have been formed in
a very different way from that which takes place in the ordinary cases of budding medusae — so
different, indeed, that were it not for the competency of the observers who have described
them as cases of true budding, we should be disposed to regard them as suggesting parasitism,
rather than gemmation.

It is not, however, only in the Gcroiiykhe that we meet with cases of heteromorphic budding
from the medusa ; for the blastocheme, as we have already seen, is constructed on the plan of a
fully developed hydroid medusa ; while its sexual buds are simple sporosacs.

' ' Generatioiiswcclisel,' p. 56. " 'Zeit. fur wisseu. Zool.,' 1853, p. 352.

DEVELOPMENT. 85

2. Development of the Ovum. — Embri/onal Developmenf.

The general form and strnctnrc of the ovum has already been considered ; the phenomena
presented by the development of the embryo now remain for discussion.

Devehjmient of the Embryo from the commencement of the segmentation of the Vitellus to the
attainment of the free locomotive stage. — I shall here describe this process as 1 have observed it in
Laomedea fexiiom, which may be regarded as affording a type of cmliryonal development through-
out the IIydroida. In this species the gonophores, which belong to the adelocodonic class, are
included within a gonangium, where they are borne along the whole length of a blastostyle,
regularly increasing in maturity as they recede from the base towards the summit of their sup-
porting column (woodcut, fig. 18). Each gonophore in the female colony contains but a single
ovum — a fact which facilitates the observation of the development.

The mature-ovum (woodcut, fig. 39 A), previous to the commencement of segmentation, is
about 001 inch in diameter; it is of a granular structure, and contains a very distinct cldar
germinal vesicle about 0002 of an inch in diameter, situated very excentrically, and easily
separated from the surrounding vitellus, when it may be isolated as a perfectly spherical vesicle
upon the stage of the microscope. There is occasionally a single germinal spot, but its place is
usually taken by several (2 to 10) minute more or less spherical or oval bodies, which lie in the
perfectly transparent and colourless fluid contents of the germinal vesicle. When the germinal
vesicle is freed from the surrounding vitellus, and floated in sea-water on the stage of the
microscope, these bodies almost instantly disappear without leaving a trace behind, being
apparently dissolved by water absorbed from without through the walls of the vesicle. If, how-
ever, a little tincture of iodine be previously added to the water, they continue visible, and are
now plainly seen to be themselves vesicles, containing within them a few minute granides which
have been rendered obvious by the action of the iodine.

The vitellus is entirely composed of minute spherical corpuscles of apparently homogeneous
structure, about 00002 of an inch in diameter, along with granules so small as not to admit of
measurement. There is no obvious vitellary membrane in the mature ovum, but I have satisfied
myself of its presence while the ovum is still in a very young state. In other species, Ilydractinia
echinata for example, this membrane is very obvious in the ovum just before segmentation.
There is no trace of a micropyle in the ovum of this or of any other hydroid which I
have examined.

As already said, there is never more than a single ovum in each gonophore of Laomedea
flexuosa ; and as this ovum continues to enlarge, it presses back the spadix until the latter is
reduced to a small hollow projection in the bottom of the gonophore.

Up to this time the germinal vesicle continues quite distinct, but it now entirely disappears
(B). The disappearance of the germinal vesicle is unaccompanied by any apparent change
in the structure of the ovum, which retains the same peculiar composition of spherical cor-
puscles and granules. I have little doubt that the vesicle now ceases to exist, and that its
disappearance is not due to its being merely concealed in the mass of the vitellus. It has
probably burst, and in so doing must have liberated its peculiar contents, which will then, of
course, be no longer visible in the vitellus. The disappearance of the germinal vesicle is

86

MORPHOLOGY.

probably the iminediate result of impregnation ; for I have seen active spermatozoa a little before
this time in the cavity of the female gonophore.

Until, however, we have farther evidence of what really becomes of the germinal vesicle.

Fin. :V.).

Develoiiment of the ovum in Laomedeaflexuosa.

A, Youn^ ovum iu the gonophore previously to the disappearance of the germinal vesicle; the gerniiual vesicle is here
seen to contain several germinal spots. B, The germinal vesicle and spots have disappeared. C, The vitellus has hecome
cleft into two segmentation-spheres. D, The ovum after the second cleavage. E, The segmentation-spheres have become
numerous, and many of them now show a distinct nucleus. F, The segmentation-spheres have greatly increased in
numbei-, and a nucleus may now be detected in each of them. G, The segmentation-spheres have still further increased
in number, while the most superficial have become arranged into a stratum distinguishable from the deeper portion of
the ovum. H, The superficial stratum has become more distinct, and is now seen to he composed of long prismatic cells.
I, The ovum has begun to elongate itself, and one end has become folded on the remainder. K, The embryo just after
its escape in the form of a ciliated planula.

it is nseless to speculate upon the influence which the supposed liberation of its contents
may exert in exciting the new series of phenomena which arc now about to take place in
the ovum ; at all events, shortly after the disappearance of the germinal vesicle, the process of
segmentation sets in. This process is certainly not preceded by the visible occurrence of a new

DEVELOPMENT. 87

nucleus destined to take the place of the germinal vesicle. It is quite possible, however, that
such a nucleus may exist, though, fi'om its small size, and from being so deeply imbedded in the
mass of the vitcllus, it may have eluded our attempts to discover it.

The first step observable in the segmentation-process is the cleavage of the yolk into two
segments (C), immediately followed by the cleavage of these into other two, so that the vitelhis
is now composed of four cleavage-spheres (U). In none of them, however, can a nucleus be as
yet demonstrated.

The segmentation would now appear to proceed very rapidly, but not always with
absolute regularity ; for it would seem occasionally to advance more rapidly in some of the
previously formed spheres than in others. By the time that the vitellus presents about thirty-
six or more cleavage-spheres (E) we begin to recognise in some of these spheres a distinct
nucleus ; while, as the spheres become smaller and more numerous, the nuclei become more and
more ap]3arent, until at last there may be seen in every minute sphere, of which the segmented
yolk is composed (F), a brilliant nucleus, visible not only in the superficial spheres, but also in
the deeper ones which come into view when the ovum is broken down under the compressor.
It is, therefore, highly probable that in the earlier stages also a nucleus exists in every
cleavage-sphere, but that in consequence of the thickness and opacity of the enveloping vitellus
it is withdrawn from observation. The cleavage-spheres at this stage present the same peculiar
structure which we find in the yolk just before the commencement of segmentation, consisting, as
they do, of minute spherical corpuscles, with still more minute granules. When the vitellus has
thus become broken up into _a great number of minute spheres, it is evident that the most
superficial of these spheres have arranged themselves into a distinct stratum, consisting of a
single layer of spheres, and completely enveloping the more internal parts (G).

We next find that the spheres composing this layer have increased in number, while at the
same time they have become longer in the direction of the radius of the ovum, and now form a
rather thick layer of undoubted cells, arranged with their long axes perpendicular to the surface
of the ovum, having their sides in close contact and investing, as with a continuous wall, the whole
interior of the mass (H).

It is impossible not to see in the entire process here described an exact parallelism with the
early stages in the development of the holoblastic vertebrate ovum, while the superficial layer of
cells, to the formation of which we have just arrived, must — though as yet showing no obvious
tendency to a splitting into distinct lamintc — be at once recognised as the representative of the
vertebrate blastoderm.

The nuclei, which were previously visible in the cleavage-spheres, have now ceased to be
distinguishable, while these spheres at the same time show a distinct investing membrane. In
fact, on now carefully breaking down the ovum under the microscope, its interior is found to
consist entirely of loosely aggregated cells, some spherical, some more elongated, and all with a
more or less copious endogenous brood of secondary cells within them.

The external enveloping layer having thus attained a considerable thickness, and a well-
defined differentiation between it and the more internal parts having been established, the ovum
begins to elongate itself, and at the same time the interior has undergone a further change; for
we no longer find in it the large mother-cells with their- endogenous brood, but a multitude of
small, free, clear vesicles, of various sizes, mingled with minute granules, similar to those which
had all along formed a part of the constituents of the segment spheres.

88 MORPHOLOGY.

At this point we may conveniently, though somewhat arbitrarily, designate tlie developing
body as the " embryo." We next find that one end of the oval embryo begins to be prolonged
beyond the rest, upon which it becomes bent back as it continues to elongate itself (I). By this
time the embryo has become endowed with evident contractility, as manifested by sluggish
changes of contour.

Shortly after this the embryo escapes from its confinement by the rupture of the walls of the
gonophore, when it speedily straightens itself, and then, in the form of an elongated, nearly
cylindrical body, slightly tapering towards one end, is discharged through the summit of the
gonangium into the surrounding water (K).

We now find that its whole surface is clothed with vibratile cilia, by whose aid it moves
slowly along the bottom of the vessel, while the cells and granules which occupied its deepest
parts seem to have undegone a kind of liquefaction, resulting in the formation of an elongated
cavity in the axis of the embryo which is thus, at this period, a nearly cylindrical sac, without,
as yet, any appearance of a mouth, but with an endoderm and ectoderm already differentiated,
while multitudes of very minute elongated oval bodies, with a high refractive power, soon make
their appearance in the ectoderm ; these are most pi'obably thread-cells, though no sign of a
filament can as yet be discovered in them.

I am unable to form any decided conclusion as to whether the endoderm, which thus about this
period becomes demonstrable, is to be regarded as the remains of the more central cells of the
segmented ovum, or as an inner lamina formed by a differential splitting of the peripheral layer
or blastoderm. Tlie appearances are rather in favour of the former view ; but if the latter be the
true interpretation, the analogy up to a certain point with the development of the vertebrate ovum
will be the closer. At all events there can be little doubt that the two membranes which now
make their appearance and continue as the endoderm and ectoderm of the developing hydroid
are fimctionally equivalent, the endoderm to the internal or vegetative layer, and the ectoderm
to the external or animal layer of the vertebrate blastoderm. It will be seen in the sequel that
the parts which are concerned in digestion and in generation have their seat in the endoderm,
while those which are destined for the fimctions of sensation, locomotion, and protection, originate
in the ectoderm — a state of things which has its exact parallel in the two laminae of the vertebrate
blastoderm.^

We have thus arrived at the ciliated and locomotive stage of the embryo. To this stage Sir
.John G. Dalyell has given the name of " planula" — a name, however, suggested by a mistaken
view of its form, which he compares to a Phinaria. In this comparison he has probably
been led astray liy the imperfection of his microscope ; for the locomotive embryo has no
tendency whatever to a flattened shape, as indicated by the name of " planula," but is always
conical or cylindrical. Instead of " planula," therefore, one is strongly tempted to employ for

^ The comparison of the structure of the Hydrozoa to the early stages in the development of
the highest animals has been very distinctly made by Professor Huxley. " The outer and inner
membranes appear to bear the same physiological relation to one another as do the serous and mucous
layers of the germ ; tlie outer becoming developed into the muscular system and giving rise to the
organs of offence and defence ; the inner, on the other hand, appearing to be more closely subservient
to the purposes of nutrition and generation." — " On the Anatomy and Affinities of the Medusce."
'Phil. Trans.,' 1849, p. 426.

DEVELOPMENT. 89

this form of embryo some term wliicli shall not tend to convey a false impression of its figure.
The term " planula," however, has passed into such general use, and lias, moreover, become so
intimately associated with the memory of one to whose admirable and conscientious observations
our knowledge of the Hydroida owes so much, that the defects of the term will hardly justify
our suppression of it.

Further BeveJopment from lite Planula to the attainment of the Adult Form. — The further
progress of the animal, up to that stage in which it has acquired all the essential features of the
adult, admits of being easily traced in many different species. I shall take as a good type of the
changes which the ciliated embryo undergoes in this progress the development of Eudendrium
ramosum (PL XIII), in which I have satisfactorily followed the various steps.

After the embryo (PI. XIII, fig. 10) has enjoyed for a period (which probably extends over
two or three days) its locomotive existence, it loses its cilia, and with them all power of active
locomotion, though still apparently retaining the power of slowly creeping from place to place by
the contractility of its body. It may now be occasionally seen with one end dilated, so as to
assume a flask-shaped form (fig. 11).

We next find that the animal has attached itself to some fixed object by the enlarged extremity
of its body, which becomes flattened over the surface to which it thus adheres (fig. 12). From
the centre of this enlarged base the rest of the embryo rises perpendicularly as a little cylindrical
or somewhat clavate hollow column. The base now expands laterally, while, at the same time, it
becomes compressed vertically, so as to acquire the condition of a little circular disc of adhesion ;
and simultaneously with these changes the embryo becomes enlarged a little behind its distal or
free extremity by the formation of a slightly prominent circular ridge, while an exceedingly delicate
perisarc has been excreted as a scarcely perceptible film over its whole surface (fig. 13).

It will next be seen that a remarkable change has taken place in the disc of attachment by
the division of this part into lobes separated from one another by radiating fissures, which com-
mence as shallow notches at the circumference, and thence gradually increase in depth until they
nearly reach the central vertical column (fig. 14). These lobes, hke the rest of the young
hydroid, consist of a layer of endoderm enveloped by one of ectoderm, while each con-
tains a prolongation from the cavity of the column, and is invested by a delicate perisarc,
which may be traced into the bottom of the dividing fissures. The lobes of the disc increase
in number by successive dichotomous division, though absolute regularity is not usually
maintained.

In the mean time the young Eudendrium has increased in size, and the circular ridge has
become more pronounced, while the part at the distal side of this ridge has in the same proportion
become more decidedly marked off from the rest of the body, and the perisarc has here become
more distinct by the partial withdrawal from it of the included structures.

Soon after this the circumference of the ridge will be found to have extended itself as a
circle of about ten short, thick tentacles, while at some distance behind these the body is seen to
be narrowed into a short, nearly cylindrical stem, springing directly from the centre of the basal
disc ; and the more contracted portion which lies at the distal side of the circle of rudimental tentacles
is now plainly recognisable as the proboscis or hypostome of the future hydranth. The tentacles
now rapidly multiply by the intercalation of others between those already formed (fig. 15). The
second set may at first be easily distinguished by their shortness ; but the bases of all seem to be
on the same level, and the whole appeal' to constitute a single unintemipted series. The tentacles,

12

90 MORPHOLOGY.

though short and thick, will have thus soon attained the full number which we meet with in the
adult. They consist in this stage of an endodermal and an ectodermal layer, the ectoderm apparently
formed of a single layer of prismatic cells, while the endoderm seems to fill the entire axis with a
mass of minute, spherical, loosely aggregated cells. Just behind the tentacles the body of tiie
young hydranth is seen to be excavated by a large cavity, in which is a multitude of loose spherical
cells, filled with a red granular pigment, and undoubtedly thrown off from the inner surface of
the walls.

The whole of the young hydroid is still completely enveloped by the delicate chitinous
perisarc, which forms a sheath extending over even the distal free extremity, and within which
the various changes just described, including even the formation of the tentacles, have been going
on. We now find, however, that this sheath (which has for some time lain loosely over the
distal parts of the hydroid, and which it seemed to invest as in a sac) becomes ruptured in front
of the tentacles, so that the water gains direct access to the surface of the young hydranth, and the
tentacles have full freedom to extend themselves. It would seem, too, that the distal extremity
of the proboscis had now, for the first time, become perforated by a mouth ; for up to this stage,
no undoubted evidence of an oral aperture could be detected.

The young Eiidendrium has thus acquired the form of a true hydranth borne on the extremity
of a short simple cylindrical stem, which still springs from the centre of the radiating disc (fig. IG).
The ste7u elongates itself, and the body, tentacles, and hypostomc rapidly acquire all the characters
of the adult. It still, however, remains for it to develop from its base a creeping stolon which
will take the place of the primordial disc, and which would seem to originate in the elongation of
some of the lobes of this disc, to complicate itself by the budding of new hydranths and the
development of branches, and, finally, by the formation of sexual zooids, to combine a gonosome
with a trophosome, in order that the little hydroid whose progressive changes we have been
thus following may attain the condition of the adult Eudendrium (figs. I and 2).

BeveJopnent hy Actinidce. — The developmental phenomena above described are, in all their
essential points, so far as we know, universal among the Hydroid a, with the exception of the
genera Tuhdaria, Hydra, and, probalily, also Myriotliela and Adinogonium.

In Tuhdaria a minutely granidar plasma, which, except in its more obviously cellular struc-
ture, is entirely similar to that which in other Ilydroida becomes differentiated into ordinary ova,
may be seen enveloping the spadix of the young gonophore. Instead, however, of becoming
transformed in the usual way into ova, portions become detached from the mass and lie loose in
the cavity of the gonophore, where they undergo a development into free embryos in the manner
to be presently described, while the residual plasma continues to detach from its mass fresh
fragments, which are in their turn transformed into embryos (PI. XX, fig. 3, and PI. XXIII,
figs. 11, 13, 14, 15, 16, 23, 24).

In the portions (PL XXIII, figs. \\g and 23 y) thus successively detached from the central
plasma (/) it is impossible to detect any decided trace of germinal vesicle or germinal spot, and yet
we should certainly not be justified in regarding tliem as mere gemmae, or in attributing to them
any other significance than that of true ova,' though, possibly, of ova after they had already passed

' Agassiz calls the central plasma in Tuhdaria the " gerra-basis," and refuses to regard as ova
the masses which are thrown off from it and become developed into hydriform young. {Op. cit., vol iv,
pp. 255 and 2G9.)

DEVELOPMENT. 91

the earlier stages of their development. The plasma in which the}' originate holds in the gono-
phores which contain it a position precisely similar to that held by the nndonbted spermatogcnons
tissue in the male gonophores (figs. 8/, 9/, and 21/) of the same species; and as nothing else
is presented by the hydroid which can in any way be regarded as ova, we should, by denying to
these the essential attributes of ova, be reduced to the anomalous alternative of admitting the
existence of the male clement without the correlative female one.

The fact, however, that the plasma at a very early period, as well as the masses which have
been detached from it in order to become developed into embryos, consist almost entirely of cell-
like elements, indicates a difference between the matter composing them and the more simple
protoplasmic matter of the unfecundated vitellus in other hydroids.

The phenomena connected with the development of the embryo in Ttibularia indivisa and
T. larynx, in both of which I have carefully examined them,* will afford a good example of the
difference between this form of development and that which is usual among the Hydroida ; they
would seem to be in all essential points similar in the other species of Tuhularia.

In the very young female gonophore of Tubniaria larynx, while yet only ^^th of an inch
in diameter, the spadix may be seen lying in the axis of a cavity bounded externally by a double
wall (PI. XXIII, fig. 19). Surrounding the spadix, and occupying the whole of the space between
it and the wall of the cavity, may be seen the generative plasma (/), consisting of a uniform mass
of small spherical cells, about 2-oooth of an inch in diameter. When liberated from the young
gonophore, and floated in water, these cells seem perfectly transparent, their contents appearing to
consist of a clear colourless fluid, with a somewhat higher refractive power than the surrounding
water. Under the action of acetic acid their contents become granular, and a nucleus-like
particle usually becomes visible in the midst of the granular contents (fig. 20).

At a slightly more advanced stage (fig. 21) the gonophore has reached to about xko^ of an
inch in diameter, and the apical tubercles (a') which characterise the mature gonophore have
begun to make their appearance. The inner layer (i) of the walls of the gonophore may now be
seen to have become separated from the outer («), and thereby i-endered more distinct. This
inner laj'er is plainly composed of minute spherical cells, and is thinner than the outer wall,
which is composed of prismatic cells, among which thread-cells are already developed. At this
period the gonophore begins to become perforated at its summit by an aperture, which opens
externally between the bases of the apical tubercles.

The tubercles continue to increase in size with the enlarging gonophore ; the plasma becomes
more voluminous, and among its component cells may be seen several of somewhat larger size
than the rest ; under the action of acetic acid these larger cells show a very distinct nucleus, with
nucleolus, in the midst of granular contents (fig. 22). It is just possible that these cells may
represent germinal vesicles with the germinal spot and its contained ptmctum, but with no vitellus
as yet differentiated around them. The plasma, retaining the same structure, continues to
increase in volume with the growth of the gonophore ; while the inner layer of the wall — that
which had immediately invested the plasma, and must be regarded as the endotheca — would
seem to undergo absoi-}3tion, and finally to disappear. We now find that a portion (fig. 23 y)
of the plasma has become detached from the mass, and soon undergoes a special development into
an embryo within the cavity of the gonophore. As has been said, no obvious trace of germinal

' "Notes ou the Hydroid Zoophytes," 'Auu. Xat. Hist.,' July, IS.I'J.

92 MORPHOLOGY.

vesicle or spot can lie found either in the entire mass or in any of the detached portions, unless
the nucleated cells just referred to (fig. 22) can be so regarded; so also the phenomenon of yolk-
cleavage, if present at all, is very obscure, but the detached mass may be easUy broken up into
cells filled with secondary cells.

The ovum (for I have no hesitation in so designating the mass detached from the primitive
plasma, notwithstanding its anomalous character) lies in contact with the remainder of the
plasma, and while in this position becomes developed into an actiniform embryo, as has been already
noticed by Van Bcnedcn,^ Mummery,' myself,' and others. In the act of development, as shown
in figs. 11 — 16, which represent the corresponding process in Tubularia indivisa, it becomes first
(fig. 13) extended as a disc over the residual plasma. In this disc we can always recognise a
difierentiation between its peripheral and central portions. Next (fig. 14), from the circumference
of the disc short and thick processes radiate all round, and these soon elongate themselves into
tentacles (fig. 15) ; the disc at the same time gradually becomes more gibbous on the side turned
away from the axis of the gonophore, its interior has already become hollowed out into a digestive
cavity, and a mouth now makes its appearance in the centre of the opposite side, or that in
contact with the plasma. The embryo now retreats from the plasma, the, mouth is seen to be
elevated on a conical prominence (fig. 16 a, fig. 24//), while the side opposite to the mouth
becomes more and more prolonged with the general cavity of the embryo continued into it. The
extremity of this prolongation presents in Tubularia larynx and some other species the appearance
of delicate strise (probably fibres) radiating for a short distance from its central point (PI. XXI,
fig. 6) — a peculiar structure which might easily lead to the belief that an aperture was here present.
The appearance of an aperture, however, I believe to be entirely deceptive. In this state it escapes
from the gonophore, a circle of very short tentacles having first become developed round the mouth
in some species {T. indivisa, fig. 16) ; while in others (7! larynsc, fig. 24) the oral tentacles do not
make their appearance until after the escape of the embryo. After continuing free (PI. XX, fig. 4,
PI. XXI, fig. 6) for a period, the side opposite to the mouth becomes ultimately developed into a
cylindrical stem, which soon clothes itself with a perisarc and fixes the young Tubularia to some
neighbouring oliject (PI. XX, fig. 5, PI. XXI, fig. 7). After the escape of the embryo, or even
during its development within the gonophore, the remaius of the plasma may stiU throw off
portions (PI. XXIII, fig. 24/), which become developed, in a similar way, into free actiniform
embryos. To such embryos the name of actinulm may be given, in order to distinguish them
from \\\& planulm of other hydroids.*

^ " Recherches sur rEmbryogenie des Tubulaires," p. 37, pi. 1, in ' Nouv. Mem. de I'Acad. Uoy.
de Bruxelles,' torn, xvii, 1814.

^ " On the Development of Tubularia indivisa" 'Trans. Micr. See.,' 1853, p. 28.

^ Allman, " On Tubularia indivisa," ' Ann. Nat. Hist.,' July, 1859.

* Prof. H. J. Clark has given a detailed account of the development of the gonophore and
ovarian plasma in Tubularia (" Tubularia not Parthenogenous/' ' American Journal of Science and
Arts,' vol. xxxvii, Jan., 1864). I cannot, however, accept in all points his interpretation of the appear-
ances presented in the microscopic investigation of these parts. He regards as the true ova certain
very minute cells which are visible in the gonophore while yet in a rudimentary state, and which
would seem to be those described above as composing the very young tissue of the plasma. Notwith-
standing, however, a certain resemblance of these cells to ova, I cannot so regard them. They cannot
be followed through any of the changes which characterise the development of a true ovum ; they

DEVEI.OPMENT. 9:3

The generative process in the freshwater Hydra offers some striking resemljlanccs to that
just described in Tubularia. Usually towards the end of autumn, hut occasionally even in spring,
peculiar tubercles may be seen budding from the body of various species of Hydra. They are
produced chiefly towards the anterior end of the body. I have especially examined them in
ffj/dra vulgaris. They are here of a conical form, and when mature have their apex perforated
by a short canal, through which the contents of the tubercle escape. These contents are then
seen to be active spermatozoa of the usual form, and the tubercle must be regarded as a male
gonophore.i Its external wall consists of a single ectodermal layer, and its cavity is traversed by
a process of the endoderm, which, at least in the younger stages, extends from the base to the
summit of the gonophore, where it remains for some time united to the ectodermal wall. Between
this axile process of endoderm, which plainly corresponds to a spadix, and the outer wall of
endoderm, the spermatogenous plasma is developed. The entire plasma has the appearance of
being divided into longitudinal masses, as if by septa, which pass from the outer wall to the axile
spadix. It increases in maturity as we examine it from the base towards the summit of the
gonophore, the reproductive elements being still enclosed towards the base in their generating
cells, while towards the summit they may be seen as free active spermatozoa, ready to escape
through the perforation which is now found in the summit of the gonophore for their exit.

But, besides the spermatogenous tubercles, there also occur, usually on the same specimen,
others which, instead of containing spermatozoa, have their cavity occupied by a peculiar cellular
plasma, destined to give origin to ova. Their position on the body of the Ht/dra, in every speci-
men which has come under my observation, was at the proximal side of that part of the animal
which carries the spermatogenous tubercles. They form rounded elevations, with a broad base of
attachment, and are of less defined form than the others. They seem to be produced by a simple
separation of the ectoderm and endoderm of the Hi/dra, with the plasma interposed between the
two membranes. They certainly correspond to the female gonophore of other hydroids, but they

simply constitute a portion of the general tissue of the plasma, as well as of the masses which are sub-
sequently detached from it in order to become developed into actiniform embryos. As stated above,
however, it is possible that the nucleated cells (PI. XXIII, fig. 22) which make their appearance at a
somewhat later period represent germinal vesicles.

Claparede (' Beobacht. iiber Anat. u. Entwickel. wirbelloser Thiere an der Kiiste von Normandie,'
1863, p. 2) also takes a different view of the development of Tubularia from that given above. His
observations were made on certain minute organisms which he found swimming in the open sea, and
which are undoubtedly the actinula-stage of some species of Tubularia. He compares them to small
medusa;, the body of the actinula representing the umbrella, and the long tentacles the marginal
tentacles of the medusa, while that portion which is subsequently to become developed into the stem of
the Tubularia is viewed by Claparede as corresponding to the manubrium — the mouth of the future
Tubularia, with its circle of short tentacles, being developed on the summit of the umbrella. Claparede
believes that he had found an aperture in the extremity of that portion which is to become the stem,
and he has apparently been thus led to interpret this part as the manubrium of a medusa. I have
little doubt that Claparede has been here deceived by the peculiar structure described above, and which
might easily lead to an error of interpretation.

1 We owe to Ehrenberg the original determination of the nature of these bodies. His account
of them is given in the ' ^Mittheilungen aus den Yerhandl. der Gesellsch. Naturf. Freunde in Berlin,'
1838, p. 14.

94 MORPHOLOGY.

present little or no trace of a spadix ; and if this ever existed, it must have been depressed at an
early stage by the ovarian plasma, which now lies upon an even floor of endoderm.

The contents of these bodies consist in an early period of development of minute spherical
cells distributed through a semifluid granular blastema, and generally exhibiting a distinct
nucleus. As this cellular and granular material increases in volume, we find it becoming broken
up into detached masses. These masses vary much in size and form ; they frequently present a
veiy irregular outline, with projecting lobes and processes of no definite or permanent shape. At
a somewhat later period some of them burst through the confining wall of ectoderm, and then
usually remain for some time in the form of irregularly spherical bodies, attached to the external
surface of the gonophore, as if by the adhesive properties of their constituent blastema.

The escaped masses may usually be seen to be themselves composed of au agglomeration of
smaller masses, reminding us of a segmented vitellus ; but I am, nevertheless, not prepared to
regard this complex condition as a true vitelHne segmentation. Further, no appearance of
a germinal vesicle or spot can at any time be detected in any part of the ovarian plasma, and yet
I believe we should not be justified in denying to the masses which have become detached the
significance of true ova.'

Beyond this point my observations have not extended, but other observers have described
the liberated masses as enveloping themselves with a tough membrane, which in some species
would seem to develop over its surface peculiar forked spines. On the rupture of this membrane
its contents become directly developed into an actiniform embryo, which gradually assumes the
form of the adult Hydra."

The resemblimce between the embryonic development in Hydra and that in Tubularia is thus
very close ; indeed, it is impossible not to regard them both as presenting the same essential
modification of the reproductive process — a modification whose most striking feature shows itself
in the formation of an actinula instead of a planula.

I have had no opportunity of studying the genus Myriothela of Sars ; but from the observa-
tions of Mr. W. P. Cocks, who was the first to meet with this remarkable hydroid genus on the
British shores, as well as from those of Mr. Joshua Alder, it would appear that actiniform embryos
closely resembling those of Tiihidaria are the immediate result of the development of the ovum.

Several years ago M. Van Bcneden described and figured a Cori/ne-Y\ke, hydroid from the
coast of Belgium, and assigned to it the name of Syncoryne pusilla, under the belief that it was
identical with the original Coryne pusilla of Gaertner.* In this determination M. Van Beneden
was wrong ; but his hydroid possesses special interest from the nature of its gonophores, which
are described as giving origin to actinula-like bodies, whose form is compared by the Belgian

' Rouget, who has examined with much care the reproductive system of Hydra (' Mem. de la
Soc. de Biologic,' torn, iv, 1852, p. 387), compares these masses to a Graafian vesicle rather than to a
true ovum.

" Pallas, ' Karakteristik der Thierpfiansen,' p. 53; Elirenberg, ' Abhandl. der Berl. Akad.,' 1836,
p. 115, taf. ii Laurent, ' Froriep's Neue Notizen,' No. 513, p. 10] ; and ' Nouveaux Recberclies sur les
Hydrcs d'eau douce, Voyage de la Bonite,' 1844. See also ' Ou the Generative System of Hydra,^ by
Prof. Allen Thomson, loc. cit., and Hancock, " Notes on a Species of Hydra found in the Northumber-
land Lakes," in the 'Annals of Natural History,' vol. v, 1850; and more especially Ecker, 'Eutwick-
elungsgeschichte dcs Griinen Armpolypen,' Freiburg im Breisgau, 1853.

^ Van Beneden, ' Embryogenie des Tubulaires.'

DEVELOPMENT. 95

zoologist to that of a cuttlefish with four arms. Taking for granted that there is here no error
of observation, the obvious interpretation is that Van Beneden's hydroid affords an example of
development from an actinula instead of a plannla.

This is a very important character, and one which, notwithstanding the general resemblance
of the trophosome to that of a Cori/ne, must remove the hydroid into a new genus, to which the
name <A Aclinogonium may be given.

3. Sipiifcance of the Medusa in the Life-Series of the Hydroid.

In our attempts to determine the significance of the sporosac, and the part it plays in the
life of the hydroid, no difficulty is encountered, for its entire history, from its origin to the fulfil-
ment of the purpose it is destined to serve in the economy of tlie hydroid, passes uninterruptedly
before our eyes, and proves it to be a true generative zooid, giving origin in some cases to
spermatozoa, in others to ova, whose development, as we have just seen, may be followed, and we
are thus enabled to trace back the hydroid in an unbroken series through the egg from which it
is developed, and the sporosac in which this egg originates, to the hydroid trophosome from
which the sporosac buds.

The cases in which a similarly unbroken chain can be traced back through the free generative
bud or planoblast are naturally far less frequent, for in the majority of cases the planoblast
does not produce its generative elements until a considerable time after it has become free, and
has undergone more or less change of form as it continues to develop itself in the open sea ; and
it is very seldom that we can succeed in rearing the free medusa, in the confinement of om- tanks,
up to the period when it shall attain to sexual maturity, either directly, as in the gonocheme, or
indirectly, as in the blastocheme.^ We thus, then, almost always lose absolute evidence of identity
in both gonocheme and blastocheme, when presented at two distant periods of their lives ; and
there is in such cases, necessarily, an interruption in the series of continuous observations.

Some uninterruptedly continuous observations, however, have been made, and we now know
of various instances in which the generative elements have been detected, either in the walls of the
manubrium or in special sexual buds developed from the gastro-vascular canals, in medusaj which
have been themselves traced to hydriform trophosomes ; while in others, though the free medusa
in which the eggs or spermatozoa have been found have not been traced by direct observation to
a trophosome, their resemblance to forms which have been so traced is so close as to justify us in
assigning to both a similar origin.^ There thus remains no longer any doubt that the significance
of the medusa in the life-series of the hydroid is in all essential points identical with that of the
sporosac ; and the assertion here made applies to both gonocheme and blastocheme, with this
difierence alone, that in the latter the generative elements are not produced directly, but only

' lu the Siphonophora the opposite condition is prevalent ; for here the gonopbores, even such as
present tlie more complete medusal or phanerocodonic form, usually become loaded with ova or sperma-
tozoa before they detacb themselves from the trophosome.

" I liave elsewhere brought together all the known instances in which medusie, whether gono-
phores or blastochemes, traceable to trophosomes, have been observed to develop generative elements.
See 'Report on the Reprod. Syst. in the Hydroida,' p. 411, &c.

on MORPHOLOGY.

through the intervention of its sexual buds. We are thus Ijrought up to an important point in
the developmental history of the Ilydroida, and are enabled to enunciate the following fundamental
proposition :

The fixed plant-like Hydroida give origin to sexual buds, not only in the form of

CLOSED sacs {the SpOrOSac), which develop within them the generative elements ; BUT ALSO
in THAT OF A MORE SPECIALISED FORM OF BUD, WHICH BECOMES A FREE (RARELY FIXED) MEDUSA,
AND THIS ULTIMATELY ATTAINS EITHER DIRECTLY {the gomcheme) OR INDIRECTLY [the Uasto-

cheme) to sexual maturity, and produces ova or spermatozoa.

But the point to which we thus arrive does not present us with the entire life-series of the
medusa-producing hydroid, for the important question still remains. What is the result, immediate
and remote, of the development of the ovum produced by the medusa? and how far does this
development correspond with that of the ovum produced in a sporosac without the intervention
of a true medusiform bud ?

A considerable number of facts bearing upon this question have also been accumulated ; and
the development of the ovum formed in the medusa has been traced, with more or less minuteness,
by various observers, so that we are now enabled to present the terms which were still wanting to
complete the life-series of the hydroid.

As the observations which have thus aided in completing our knowledge of hydroid
development are of great importance in the present inquiry, it will be necessary to give here some
account of those cases in which the development of the egg of the hydroid medusa has been
satisfactorily traced.

Dujardin^ observed that a remarkable little medusa, which he described under the name of
Cladonema, was developed as a bud from a hydriform trophosome, to which he gave the
name of Stauridium. He had noticed the production of eggs by his Cladonema, and had also
seen young Statiridia developed from these eggs, though the planula stage seems to have escaped
him.

Krohn," having placed in a jar of sea-water some mature specimens of Dujardin's Cladonema,
observed that after a time they had deposited eggs, which adhered to the sides and bottom of
the vessel. Soon after deposition, the segmentation of the yolk commenced ; and in about forty-
eight hours after the beginning of the cleavage the ovum had become changed into a free-swimming
ciliated infusorium-Uke embryo (planula).

This embryo was successfully watched by Krohn through all the subsequent stages — the
disappearance of its cilia, the fixing itself to the sides of the jar, its conversion into a little circular
disc, the growth of a short column from the centre of the disc, and its final conversion into a
hydroid, identical with the Stauridium from which Dujardin had originally seen the Cladonema
thrown off. To Dujardin and to Krohn are thus due the first grand oliservations by which the
whole circle of hydroid develo])ment, in the case of a free phanerocodonic gonophore, has been
completed.

^ Dujardin, 'Ann. des Sc. Nat.,' ser. iii, vol. iv, 1845, p. 273.
- Miiller's ' ArcLiv,' 1853, p. 420, tab. xiii.

DEVELOPMENT. 97

Gosse^ has seen the medusa dcscrihcd by Forbes under the name of Tiirris nrr/lecln, discharge
from the generative mass formed in the walls of its manul)rium ciliated planuLnc, which, after
some time, fixed themselves to the glass, and became elongated into adherent, branched, stolon-
like bodies, which threw up a perpendicular stem, on whose summit a circle of four tentacles
was developed, and the whole became thus changed into a C7«jj«-like hydroid.

Strethill Wright^ subsequently watched the development of the ovum in this same medusa.
His observations agree with those of Gosse, but he has succeeded in tracing the development a
step further; for he saw the tentacles increase in number by the growth of others behind
those first formed, giving by their scattered disposition a still more C/«u«-like appearance to
the hydroid, while he also noticed the formation of a chitinous periderm which clothed the
creeping stolon.

Gegenbaur' describes the development of the egg in a medusa, which he names Lizzia
KulliJceri. He has seen the segmentation of the vitellus, and the formation of a ciliated planula,
which, after enjoying for a time its locomotive existence, loses its cilia, fixes itself to the side of
the vessel, expands one extremity into a disc of adhesion, elongates the rest of its body into
a cylindrical stem, which after clothing itself with a chitinous polypary, develops a mouth upon
its free extremity, and just below this throws out a verticil of tentacles, while the expanded base
becomes extended into short stolon-like prolongations.

The development of the ova in another medusa, named by Kolliker Oceania armafa, was
also observed by Gegenbaur.* He traced the segmentation of the vitellus, the formation of a
ciliated planula, the fixation of the planula, and its development into a stolon-like body ; but
beyond this point his observations were not carried.

Wright" noticed the occurrence of numerous planula^ which had made their appearance in
a vessel in which he had placed some isolated specimens of Thaumantias inconspicua, Forbes. He
believed that these planulae were produced by the Thaumantias, and he saw some of them fix
themselves to the sides of the vessel and develop a lobed disc of attachment. From this disc
arose a stem, which after developing from its summit a hydranth closely resembled the Cam-
panularia raridentafa of Alder.

In the Zyyodactyla {^'Equorea) vitrina of Gosse, Wright^ also observed free ciliated planulse
to escape from the generative bodies, and, after fixing themselves to the sides of the vessel,
become developed each into a hydroid, with hydranth, hydrothcca, and perisarc, bearing, as he
informs us, a close resemblance to the Laomedea acuminata of Alder.

Alexander Agassiz'' has followed the development of the egg in two forms of hydroid
medusa; — Melicertum campanula, Esch., and Tima formosa, Agass. In both he has seen the
formation of the ciliated planula, the fixation of the planula, and its gradual conversion into a
young campanularian trophosome.

1 'A Naturalist's Katnbles on the Devonshire Coast,' 1853, p. 34<S, pi. 13.
' 'Edinb. New Phil. Journ.,' July 1859, pi. 8, f. 1.
' ' Generationswechsel,' 1854, p. 23, pi. ii, figs. 1-9.
^ Loo. cit., p. 28, pi. 2, figs. 10— IG.
° ' Micr. Journ.,' vol. ii, new ser.
^ 'Micr. Journ.,' vol. ii, pi. iv, figs. 1 — 6.

' 'Illustrated Catalogue of the Museum of Comp. Zool. at Harvard College,' No. II, pp. 115
and 134,

13

98 MORPHOLOGY.

I have myself traced the development of the egg of a Tijaropsis as far as the plaiuila stage ;
but though the planula; continued to live in my jars for some weeks, they ultimately perished
without passing into any further phase of their metamorphosis.

The class of observations here enumerated enable us to complete the circle of hydroid
development, and justify us in the enunciation of a second proposition, which, taken along with
the former one, will express the entire life series of the hydroid :

The ova of the medusiforu bud undergo, like those of the sporosac, a con-
tinuous DEVELOPMENT, BY WHICH THEY BECOME TRANSFORMED INTO HYDRIFORM TROPIIOSOMES,
WHILE THESE TROPHOSOMES ULTIMATELY GIVE ORIGIN, BY BUDS, TO MEDUS.E IDENTICAL WITH
THOSE FROM WHOSE OVA THE TROPIIOSOME WAS DIRECTLY DEVKLOPED.

It will be further seen, from the facts now stated, that the earliest stage of the hydroid tro-
Ijhosome is always free and locomotive, and that it shows itself under one or other of two types.
One of these types is presented by the great majority of the Hydroida, and has been described
above as \hQpIanulaol Sir J. G. Dalyell ; the other occurs in the genus Tubularia, and apparently
also in Myrotltcla, Hydra, and Actinogonium, and has been already described under the name of
adinula. Every hydroid, if we except such forms as may be proved to pass to the medusal con-
dition directly from the egg, thus commences its free existence either as a planula or an actinula.

Direct Development of the Medusa from the Eyg. — In by far the majority of cases in which
the development of the hydroid has been successfully traced, the life series of the individual has
presented a non-sexual hydra-like form, interposed between the ovum and the directly or indi-
rectly sexual medusal form.

Against the absolute universality of this law, however, certain oi)servations have been adduced,
as tending to show that in some cases a direct development from the egg to the medusa takes
place without the intervention of a non-sexual trophosome. There is no reason why this should
not be so, and yet a careful examination of the cases adduced in support of it will render it
evident that most of them afford no evidence which can be relied on as conclusive in favour of the
direct development of the medusa from the egg.^

It is chiefly among the ^Eginidan and Geryonidan medusaj that cases believed to afford evidence
of direct development have been observed. The first observation bearing immediately on this
question is due to Johan. Miiller,' who captured, on several occasions in the sea, at jMarseilles and
Nice, a minute free-swimming hydroid. It was of an oval form, about half a line in its longer
diameter, ciliated over its entire surface, with two tentacle-like processes near one end, and
having at the opposite end an opening which led into a central cavity.

Muller considers this little animal to have been developed directly from the egg, and from its
resemblance to a peculiar two-tentacled medusa which he obtained in considerable abundance at
Nice, he believes himself justified in regarding it as one of the stages in the development of this

' Among the Discophoka a single case also of direct development has been made kuowu ; that,
namely, of Pelaijia, a medusa which has been sliown by Krohu to be developed from the egg without
tlie intervention of a Sciiphosloma or polypoid form.

• -Muller's ' Archiv,' 1851, p. 272.

DEVELOPMENT. 99

medusa, into which lie supposes it to pass by direct metamorphosis. He refers it to the genus
yEf/inoj]sis, Brandt, and names it JEr/inopsh Mediferrcniea, Midi. Miiller docs not seem to liavc
obtained any specimen of his j-E. Med'derranca, so far matured as to present traces of the
generative elements ; but his observations have been in this respect supplemented by Kollikcr,'
who afterwards obtained the same species at Messina in a sexually mature state.

Now, we cannot overlook the fact that Miiller has not, in the above case, traced his ciliated
hydroid through a continuous series of developmental phases into the adult form of jEjinojmH ;
and, without denying the probability that the ciliated bitentacular hydroid is really the larva of
the JEffinopsis, we cannot regard this relation as absolutely proved, wJiile there is no evidence
whatever that the ciliated form is the immediate result of the development of an ovum. Indeed,
its remarkable resemblance to the singular generative zooid of Bicoryne (see above, p. 31)
would seem to show the probability of another origin than that by direct development from
the egg. Midler, led apparently by the analogy of the planula-stage of the Hydroida, considers
the ciliated condition of the surface as affording evidence of such a direct development ; but the
fact that the Dicori/ne-7jOo\A is also richly ciliated over its whole surface shows that this argument
goes for nothing.

Kolliker" found in the stomach-cavity of a ten-tentacled iEginidan medusa, captured in the
sea at Messina, and described by him under the name of Enrydoma rnhiyinosiim, a number of
small organisms resembling medusae in various stages of development, and which he believed he
could follow from stage to stage until he found them assume the form of a sixteen-tentacled
medusa. To this last, which also belongs to the family of the yEyinida, he gives the name of
Stenoyaster complanatus.

The great difference betvveen these two medusae appears to Kollikcr sufficient proof that
the one could not have been produced by the other, and he regards the young stenogasters as
having been swallowed by the Eurystoma. He views, however, the young Stenoyaster, exhibiting
as it does, various steps in a metamorphosis from a very early stage, as affording evidence of the
direct development of Stenoyaster from the egg. It is, nevertheless, plain that there are no more
valid grounds for such a conclusion in this instance than iu Johan. Miiller's case of j^yinopsis,
while Fritz Muller's case of Cunina Kollikeri, as well as the cases described by Gegenbaur and
by Keferstein and Ehlers, and the more recent observations of Haeckel, all of which are cited
above (p. S3), suggest the probability that the stenogasters noticed by Kolliker originated as buds
from the Eurystoma.

Other instances which have been adduced as affording evidence of direct development from
the egg without the intervention of a trophosome have been already referred to as cases where
the medusa passes through a series of metamorphoses before arriving at its adult state. They
are i\rCrady's case of Cunina octonaria, Fritz Miiller's of Liriope cathariensis, and Haeckel's
of Glossocodon euryhia and Carmarina hastata, all of which, from the very imperfect state of
development in which the earliest stages of the medusae present themselves, have been regarded
by their describers as instances of direct development from the egg, though there is no positive
evidence of such an origin ; and, lastly, there is the instance afforded by Trachynema, the ciliated
condition of whose youngest di.^covered stage has led Gegenbaur to consider it also as a case of

^ ' Zeit. fiir wissensch. Zool.,' 1853, vol. iv, p. 327.
' Ibid.

100 MORPHOLOGY.

direct development, though, as we have already seen, the analogy of the Dlcori/ne gonophore is
sufficient to show that no conclusion of the kind can be based on such a character.

Though the cases thus adduced afford no absolute proof of the direct development of the
medusa from the egg, some recent observations leave us no longer in doubt as to the reality of
this phenomenon. The observations alluded to have just been made by M. Mecznikoff of St.
Petersburgh, who has traced the development of a Cunina as well as that of a Geryonia continu-
ously from the egg. In the case of the Cunina, a ciliated planula is immediately developed from
the egg, and the planula becomes gradually changed — in a way very similar to that already
described by M'Crady — into the form of the medusa. In the Geryonia there is no ciliated
planula, and the medusa form is here attained immediately by the development of the ovum.'

Besides these cases of undoubted development from the egg without the intervention of a
hydrifonu trophosome, both of which occur in a group of medusa; of peculiar and exceptional
conformation, an instance has been published in which, if the appearances be correctly interpreted,
we have in the hydroid medusa of the ordinary type a case of direct development from the ovum.
For our knowledge of it we are indebted to Claparede,^ who obtained on the west coast of
Scotland a species of Lizzia whose manubrium is described by him as loaded with eggs, some in
an early stage, with the germinal vesicle and germinal spot still visible, while others appeared to
contain an embryo in various stages of development. Similar ova, with the contained embryo,
are stated to have been found floating free in the sea.

Claparede informs us that the embryo, while still confined within the vitellary membrane,
presented all the features of a young medusa : from the centre of the bell-shaped umbrella there
depended a thick-walled manubrium, whose cavity extended itself into four radiating gastrovascular
canals, which ran in the substance of the umbrella, and opened at the margin into a circular
canal, while round the margin were to be seen the rudiments of eight tentacles. Claparede's
observation on the development of the embryo did not extend beyond this point ; it is clear,
however, that but slight changes were now needed to convert it into the form of the parent
Lizzia.

This observation of Claparede has not been confirmed, and ic is quite possible that the
appearances here interpreted as ova in various stages of development are in reality only buds. A
very young bud might be easily mistaken for an ovum, and in a medusa, by no means remotely
allied to that described by Claparede, buds occur in an exactly similar position, and might easily
give rise to an erroneous interpretation of their nature. (See woodcut, fig. 36.)

' For a knowledge of these facts I am indebted to M. Mecziiikofl', who allowed me not only to
inspect his drawings, but to e.xamine his animals iu the progress of their development. M. Mecziiikotf's
observations necessitate some modification of a statement made in an earlier part of the present work
(see p. 23), and already printed before they came to my knowledge.

" 'Zeit. fiir wisseu. Zool.,' 1861, p. 401.

develop:\ient. loi

4. Relation between zooidal and embryonal multiplication in the Htdroida ; Polymerism
and Ileteromorphism ; Genetic succession of zooids.

Having now examined tlie various modes, whether zooidal or embryonal, by which multipli-
cation is effected in the IIydroida, it remains for us to see how the two forms of reproduction
are related to one another, and how they are associated in the complex phenomenon which con-
stitutes the life of the hydroid.

From all the facts which the study of the IIydroida has made apparent, we may regard it
as certain that however long zooidal multiplication may continue, this is not sufficient for the
perpetuation of the species, but that a period must at last come in the life of the hydroid when,
by an act of true sexual reproduction, new individuals are produced for the indefinite extension
of the species through time.

These facts find their expression in a remarkable law originally propounded by Chamisso,
when he made his memorable discoveiy of the true genetic relation between the solitary Salpa
and the associated chain-like colonies of these animals ; though it was reserved for Steenstrup, by
correlating with Chamisso's discovery, not only the genetic phenomenon of the IIydroida, but also
various analogous phenomena observed in other members of the animal kingdom, to give a wider
comprehensiveness and a more definite enunciation to the law henceforth known as the law of
"alternation of generations," an expression originally employed by Chamisso himself when describing
the genetic phenomena of Salpa. It is true that Steenstrup's mode of stating his law of alternation
of generations was destined to undergo some modification, but it has, nevertheless, received in
all essential points abundant confirmation, and will explain, in a way which it alone can do, a
host of phenomena which would otherwise have appeared isolated and exceptional.'

The law of alternation of generations manifests itself wherever it prevails, in the fact that
every act of embryonal development is followed by one or more acts of zooidal development,
which invariably conduct us to an ovum in which embryonal development, followed by zooidal
development, again occurs, and the entire series becomes thus repeated.

Now, the various series expressing this alternation of sexual with non-sexual development
exhibit among the Hydroida different degrees of complication, which will be more easily under-
stood if we attempt to present them in the somewhat technical shape of formulae.

Let t be the trophosomc, and y the gonosome then —

I. < + /7 X ^H-.y X '+,^ X • • ■ &c.

will be the general expression for the genetic succession in the life of the hydroid, the sign +,
indicating succession by zooidal development, and X by embryonal.

' The true significance of the facts on which the law of Alternation of Generations is founded,
was for tlie first time clearly pointed out by Dr. Carpenter. See ' Brit, and For. Med.-Chir. Rev.,'
vol. i, Jan., 1848, p. 183, &c.

102 MORPHOLOGY.

It is very seldom, liowcver, that the trophosome consists of only a single zooid. Such rare
instances are presented by Corijmorplia, and by certain allied forms whose trophosomes never become
developed into a colony of mutually dependent hydranths, and I believe it better to regard the
hydrorhizal fibres here as elsewhere in the light of mere extensions of the hydrorhizal base rather
than in that of proper zooids — a view supported by their mode of development in the primordial
hydranth. In almost every other case, on the contrary, the hydranths composing the trophosomes
become greatly multiplied by budding ; and in this respect Hydra affords no exception, though
here the troj)hosome, by the subsequent detachment of the buds, may become restored to its
original condition of a simple hydranth.

Still less tendency is there in the gonosome to present an absolutely simple condition.
Indeed, the gonosome is perhaps never limited in its normal state to a single zooid, and we fre-
quently find hundreds and even thousands of zooids entering into the composition of this portion
of the hydroid colony.

But the zooids of which the colony is thus composed, whether in its trophosome or its gono-
some, may not only be numerous, but may also vary inform. Those, indeed, which constitute the
trophosome are always of a different form from those of the gonosome. In the trophosome it is
rare to find any other form of zooid than that of the proper hydranth. In Hydradinia, however,
there is associated with the ordinary hydranths the peculiarly modified ones, whose spiral form
confers upon the trophosome of this genus so striking a featm'e ; while the nematophores of the
Phimularidte can scarcely be regarded otherwise than as special zooids, whose morphological dif-
ferentiation from the other zooids of the colony is carried to an extreme.

While the type of heteromorpldsm or variety of form among the zooids is fixed for every
species, i\\e polymerism or simple multiplication of the component zooids is indefinite, and varies
with the age, perfection of nutrition, &c., of the individual.

If we specialise the general expression already given (I), so as to make it directly applicable
to particular cases of heteromorphic succession in the life of the hydroid, we shall obtain the
following formulas, where // is used for hydranth, his for blastostyle, hlch for blastocheme, ijph
for gonophore :

II. < )■ h-{-gph X h-\- gph X
J

jy_ I I A -f i/s + bich + ffph X h -{- Ms -{- blch -f gph X

CJJ

DEVELOPMENT.

103

These formiiLx present three types of hetcromorphism. In II the heteromorphism is
binary (woodcut, fig. 40), in III ternary (woodcut, fig. 41), in IV quaternary (woodcut,

fig. 4:2).

But the hydranth mny and does in ahnost every instance — either directly or through
the medium of the common basis or hydrophyton — repeat itself indefinitely by budding before the

Fig. 41.

Fig. 42.

Diagr.im of Corymorpha.

A. Au entire colony, consisting of two
forms of zooids. a a a, the trophosorae,
consisting of a single hydrantbal zooid;
6, the gonosome, composed of many medu-
siform zooids.

B. A single zooid of the gonosome after
having become free and attained to sexual
maturity.

Diagram of Dicoryne, Colony
composed of three forms of
zooids.
aaa a. Tlie trophosome, com-

posedof numerous hydranthal zooids;

h c, the gonosome, composed of two

forms of zooids, namely, b, bias-

tostyle, and, c, gonophores.

Diagram of Campanularia. Colony com-
posed of four ibrms of zooids.

A. Portion of a colony, with tropho-
some and gonosome, a a. Trophosome,
showing one of the hydranths of which
it is composed ; 4 c, gonosome, composed
of three forms of zooids, namely, blasto-
styles, blastochemes, and gonophores;
b, blastostyle ; c, c, blastochemes.

B. A blastocheme, which has become
free and attained to maturity, showing
the gonophores springing witliin it from
the radiating canals.

time arrives when an element of the gonosome is to be budded off; and a series of homomorphic
zooids may thus introduce themselves (woodcut, fig. 41) into the heteromorphic succession, as
expressed in the following formula ;

V. A+A+A+

&,c.-{-bls-\-gphXh-\-h-{-h-\-

ii,c.-\-bls+gphX

&c.

where the hydranth becomes indefinitely repeated in the formula of ternary heteromorphism (III)
given above, and the same will apply to each of the other two types of heteromorphism.

Now, in all these cases the succession from the primordial nutritive zooid to the ultimate
generative zooid or gonophore admits of being expressed in a continuous line ; but one or more

104

MORPHOLOGY.

of the zooids of the trophosome may emit buds which will diverge from the direct Hne of succession,
and which may then either form the starting point for another similar line of succession, or may be
destitute of all power of continuing the succession of the zooids. Thus (woodcut, fig. 43) the
primordial hydranth, or any of its derivative hydranths, may repeat itself by a bud which will
diverge from the direct line, produce other zooids by gemmation, and thus start off a new series,
as expressed in the following formula :

VI. h ] +

h -{- h-{- h-\-

&c. 4" 6ls -\- ffph
&c. -)- bis -\- ffph

X h

{}

&c.

And this state of things may also repeat itself indefinitely, giving rise to an indefinite number of
collateral series, diverging from one another and from the primary axis of succession.

Pig. 43.

Diagram of Laomedea.

a, a, a, a. Hydranths belonging to the primary or direct line of succession ; a', a', a', a', hydranths
belonging to a secondary or diverging line of succession ; I, blastostyle of the primary line of succession,
bearing gonophores and surrounded by a gonangium; b', blastostyle, with gonophores and gonaugium
of the diverging line.

As already said, how^ever, the diverging zooid may have no power of continuing the succes-
sion. Thus, the spiral hydranth oi Hydradinia is not intercalated in the direct succession of
zooids. It is a diverging zooid, like that which starts off the collateral series in Formula VI,

DEVELOPMENT. 105

but one which never gives rise to buds, and is therefore incapable of citlicr continuing or
originating a new succession.'

The following formula, where //' is the spiral hydranth, will express the place and jjower of
this zooid in Hi/dractinia :

+ */■« + 'JP^>- X . . . . &c.

va *{ + «• + '"" x^lit^;'

The case expressed in Formula VI is the simple one where only the last hydranth in the
succession of buds composing a period is supposed to give origin to a bud of the gonosome. But
any other hydranth in the succession may just as well bud otf a member of the gonosome, which
may thus form a collateral gonosomal axis. This, which is by far the most usual case, is what is
actually represented in the diagrams (see figs. 41, 42, 43). The axis, however, thus formed will
be necessarily definite, and will contrast in this respect with the indefinitely extended axis of
the trophosome, while it will differ from the diverging bud, // in VII, by the fact of its having
the power of repeating the colony by sexual reproduction, while // has no power of reproduction
either sexual or non-sexual.

This condition may be' expressed by the following formula, in which not only the last
hydranth of the period gives off a bud of the gonosome, but the primordial hydranth itself emits
a collateral gonosomal axis :

) -j- i/5 -i_ ,jph ) I )

Besides the particular cases now given certain other modifications of the plan of gemmation
will at once occur to any one who has made the Hydhoida a subject of study. Those here
adduced, however, will serve to convey a sufficiently adequate idea of the more important features
in hydroid gemmation. It is thus, by the combination of heteromorphic and homomorphic
multiplication, and of direct and diverging series indefinitely repeated, that the animal attains to
the condition of those wonderful complex colonies which impress themselves so strongly on the
mind of the observer.

So also the gonosome may present, not only a heteromorphic, but a homomorphic multipli-
cation of zooids. In no case, however, so far as I am aware, does any zooid of the gonosome
repeat itself by homomorphic gemmation, except in some comparatively rare instances of budding
in the Medusa ; for though the homomorphic repetition of zooids may be in the gonosome, as in
the trophosome, carried to a great extent, it is almost always the resiJt of budding from a zooid
of a different form. Thus, the blastostyle never emits buds destined to repeat its own form, and
this form, however frequently repeated in the gonosome, is always budded oft' from the hydranthal
element in the trophosome, its own buds, however numerous, being always heteromorphic with
itself.

It is a universal law in the succession of zooids that no retogression ever takes place in the
series. In other words, no bud ever becomes developed into a zooid which is of a different

1 The bifurcation occasioually observed in the spiral hydranth of Hydractinia is evidently abnormal,
and cannot be regarded as contradicting the above statement.

14

lOG MORPHOLOGY.

form from the Inicldcr, and lias, at the same time, preceded it in the line of succession. Thus,
true hydranths are never emitted either by blastostyle, blastochemc, or gonophore ; and to this
law the peculiar gemminatc hydriform bodies which are found on the summit of the female blas-
tostyle in certain species of Halecium form no exception ; for though closely resembling true
hydranths, they appear to have a difTerent signification, contributing rather to the generative
functions of the hydroid, while they have no power of continuing the succession either in a direct
or collateral line like the proper hydranths of the trophosome.

Now, a glance at any of the formulae given above renders it evident — 1. That between every
two acts of true generation there are interposed one or more acts of non-sexual multiplication.
2. That the heteromorphic elements in each recurring period of the succession are invarialily con-
nected with one another by a non-sexual and not by a sexual genesis. 3. That the type of
heteromorphisra exactly repeats itself after each true generative act.

A still further fact, however, is apparent in all the cases here adduced, namely, that a certain
number of zooids, incapable of attaining to sexual maturity, and hence becoming multiplied only
by zooidal reproduction, occur in every succession. To the universality, however, of this prin-
ciple, Haeckel regards the case of the Geri/onidcR already referred to (p. 82), as affording an
exception. He has found Geryonidan medusae swimming freely in the open sea, in such an early
stage of their development, that he believed them to have been produced by the direct develop-
ment of an ovum, and yet these medusae have been traced by him into a condition which he
regards as that of sexual maturity, in which state they not only produce generative elements, but
give origin, by heteromorphic budding within the stomach, to ^ginidan medusa {Cunina), these
Cuiiina-hwA?, also attaining to a condition of sexual maturity.

These facts are regarded by Haeckel as presenting an entirely new type of genesis — a
type totally different in its fundamental principles from the phenomena hitherto included under
the head of " Alternation of Generations ;" and, believing a new term to be needed for it, he pro-
poses to distinguish it by the designation of " Alleogenesis." It is worth while to inquire
how far Haeckel is borne out in this mode of viewing the phenomena of Geryonidan
development.

Admitting that Haeckel is right in regarding his Geryonidan as developed directly from the
egg without the intervention of a non-sexual trophosome, I am by no means prepared, as I have
elsewhere^ stated, to take for granted the proper sexuality of this medusa. On the contrary, I
am still disposed to consider the sexual pouches of the radiating canals as truly zooidal deve-
lopments corresponding, notwithstanding their flattened leaf-like form, to the more prominent
pouches developed on the radiating canals of such forms as Obelia ; so that, in accordance with
this view, the Geryonidan medusa would be a true blastocheme. If this be so, then the non-
sexual character of the Geryonidan must be admitted, and a non-sexual element will thus
become intercalated in the series, even though the hypothesis of a non-sexual hydi'iform tropho-
some be given up. Haeckel, it is true, referring to my view of the zooidal nature of the sexual
pouches of Geryonia, argues against it, and states his conviction, from personal examination of
these pouches, that they are simple lateral dilatations of the canal, with the generative products
developed out of the epithelium of their walls.^

' ' Ann. and Mag. of Nat. Hist.,' 1865.

* Haeckel, 'Die Familie der Riisselquallen. Vorwort,' vii.

DEVELOPMENT. 107

Against the direct testimony of so able and conscientious an observer as Haeckel, I sliould
not consider myself justified in insisting on a hypothesis which I have had no opportunity of
verifying by direct examination; but yet I can scarcely avoid seeing that Ilaeckel's description
of the structure of these pouches is in some points favorable, rather than contradictory, to my
view ; thus the currents of nutritive fluid which he has observed flowing in ramified channels
through the mass of the ova appear to me to be explicable only on the admission that these
currents are contained within a ramified spadix, for the supposition that the generative elements are
directly bathed in the fluid of the gastro-vascular canals is so completely at variance with the
analogy of these parts in all the other Hydroida, that we can scarcely bring ourselves, without
very strong evidence, to accept it. If we admit the presence of a true spadix penetrating the
pouch, and surrounded by the ova or spermatozoa, we have all the parts needed to establish a
detailed homology between the leaf-like pouches of Geryonia and the prominent sacs of Obelia,
and these last are, without any doubt, true zooids, strictly homologous with the sporosacs of
Clava.

It is more difficult to recognise a zooidal origin in the generative pouches of the Cunina which
Haeckel has shown to be produced as a bud from the Gerijonia, and there seems no reason
why we should not, with Haeckel, regard the Cimince as truly sexual medussR. What may be
the subsequent history of these Cunina, is as yet entirely unknown ; and until this shall have
been determined, the significance of Haeckel's beautiful discovery of the relation between the
Geryonidans and ^ginidans must remain but partially recognised.

In the genetic phenomena of the Hydroida, so far as these have been accurately determined,
one fact stands out in prominent relief, and its recognition is of great importance in enabling
us to perceive the true import of these phenomena, and the mode in which they are associated
in the life of the hydroid. I again refer to the fact that in every hydroid the groups included
between every two acts of embryonal development (the groups connected by horizontal brackets in
the above forumlae) are exactly similar in the nature and succession of their heteromorphic elements,
— in other words that the life series of the hydroid may be represented by definite groups of
zooids exactly repeated after each generative act.^ It is plain, too, that each of these groups
— which we may conveniently designate as the " periods" of the series — exactly corresponds to
the " individual" which constitutes the proper logical element of the sjjecies in animals which do
not present the phenomenon of alternation, the period here repeating itself by true generation,
and this repetition continuing itself indefinitely like a circulating decimal, so as to represent the
indefinitely extended life of the species, while the life of the individual is expressed by each period
singly. It is further evident that the conception of the individual involved in the above view
is in no respect invalidated by the fact that one or more of its zooidal elements may become free,
and enjoy an independent existence.

For the views of Hydroid individuality, embodied in the above paragraph, we are indebted to
Prof. Huxley, who first assigned to our conception of the biological individual its proper limits
when he defined it as " the total result of the development of a single ovum" — a most
important determination by which alone the genetic phenomena of the Hydroida can be
properly understood and brought into comparison with those of the higher animals. At the same

^ Tlie mere number of zooids in two or more of tliese groups may of course vaiy, depending as
this does on the accident of abundant or deficient nutrition and the like.

108 MORPHOLOGY.

time it must be borne in mind tiiat it is "the individual" in the somewhat technical sense of the
component of a biological species which is to be here understood, and that individuality, in its
more ordinary acceptation, cannot be excluded from our conception of the life-series of a hydroid.
In this sense every zooid has an individuality of its own — an individuality, however, of a very
different kind from that which characterises the successively repeated groups of zooids constituting
the individuals which logically make up the biological species.

The hydranth normally continues the axis in the hydroid colony, just as the leaf-bud in the
plant continues the vegetable axis ; the gonophorc, on the other hand, has no power of con-
tinuing the axis, and constitutes the terminal zooid in each "period" of the series, just as the
flower-bud stops the elongation of the axis in the plant. This analogy, however, must not be
pushed too far, for while the hydrauths and gonophores are simple zooids, the leaf-buds and
flower-buds are complex associations of the corresponding element of individuality in the
plant.

The normal order of succession of the buds in the trophosome is from the proximal to the
distal end of the hydrosoma, so that the older buds are met with towards the base or hydro-
rhizal end of the main stem and branches, the younger ones towards the summit. In the
gonosome, on the other hand, the order of succession is sometimes towards the distal, sometimes
towards the proximal end of the axis. In the calyptoblastic genera the order of succession of the
sporosacs or blastochemes is invariably from the distal towards the proximal extremity of the
blastostyle, on which, in these genera, they are always borne. When a blastostyle is present in
the gymnoblastic genera, the gonophores succeed one another, sometimes {Hydractinia ecldnata)
from the proximal towards the distal end of the blastostyle, sometimes {Bicoryne covfertd) from
the distal towards the proximal. In Tuhulariu their succession is from the distal towards the
proximal end of the common peduncle, which is more or less developed in the various species of
this genus, and the same order of succession occurs in Corymorpliu.

Where no special gonosomal axis is developed the succession is usually from the proximal
to the distal extremity of the branch [Bouyainvillia, Periyonimus), thus corresponding to that
of the zooids of the trophosome. Sometimes, however {Syncoryne, Gemmnria), it is from the
distal to the proximal.

We have thus, then, in the gonosome of the IIyduoida, as in the inflorescence of plants,
both a centripetal and a centrifugal order of succession. It is possible, however, that irregu-
larities may occur, and that a new bud may be abnormally emitted at the distal side of a
centrifu<ral series, or at the proximal side of a centripetal one, so as to disturb in individual cases
the normal sequence of the zooids.

Some further points admitting of comparison with the inflorescence of plants may be noticed
in the gonosome of such hydroids as possess a special gonosomal axis. In Ttihulana indivmi
(PI. XX, fig. 2, 3), and in the male colonies of Tuhidaria larynx (PI. XXI, fig. 1) the gonophores
are— like the flowers of a raceme — carried on short pedicels along the sides of a long common
peduncle which springs from the body of the hydranth (woodcut, fig. 44). Their order of develop-
ment, however, is centrifugal, or from the distal to the proximal extremity of the peduncle, so that
the whole group may be compared to a reversed raceme. In the female colonies of Tiibularia
larynx (PI. XXI, fig. 2, and woodcut, fig. 45), and in Corymorpha nutans (PI. XIX, figs. 1 and
3), the pedicels become branched, with a similar order of development which thus gives us the
compound reversed raceme or cyme.

DEVELOPxMENT.

109

The reversed spike, or spike with a centrifugal development, shows itself in such forms as
Bicoryne conferta (PL VIII, lig. 1, and woodcut, fig. 41) ; while in Laomedca (woodcuts, figs. 18
and 48), ObeJia, woodcut, fig. 19), and other calyptoblastic forms, we have a reversed spike

Diagram ot' Tnhularia indivisa,

a, a. llydranth with its stalk; b, shortly
stalked gonophores, borne on a common
peduncle, and increasing in maturity from
the proximal to the distal extremity of the
pedunrle.

Fig. 45.

Diagram of Tnbularia larynx — female.

a, a. A hydranth with its stalk ; 5,
gouophores, attached by short stalks to
a common branched peduncle, and in-
creasing in maturity from the proximal
to the distal extremities of the branches.

Fig. 46.

Fig. 47.

surronncled by the gonangial sheath, and, were it not for its centrifugal development, strongly
recalling the spadix with its spathe in the inflorescence of an araceous plant.

In certain proliferous medusa; (woodcuts, figs. 30 and 37) the buds are borne on the luanu-
briuin with a centripetal order of development, thus giving us, according as the buds are sessile
or pedunculated, the true spike or the true raceme.

In Eudcmhium (Pis. XIII and XIV) the male gonophores are disposed in an umbel with

the axis, in some cases prolonged beyond

it, while in others there is no extension of

the axis beyond the depressed portion

which carries the gonophores. Though

we cannot here recognise any difference

in the order of development among the

gonophores composing the umbel, we are

probably justified in assuming this order

to be, as in the true umbel, a centripetal

one; for in the female colonies of most

species, such as Eudendrium ramosum

(PI. XIII, fig. 3), the gonophores are

separated from distance to distance upon

the stem immediately below the hydranth,

and here their order of development is

plainly seen to be centripetal.

In Ilijdraclinia echinala (PI. XV,
figs. 1 and 3, and woodcut, fig. 4G) we have the closely approximated gonophores sessile on a
blastostyle, and the development centripetal, as in the true spike, while the axis extends beyond

Rydractinia. A blastostyle
with gonophores clustering
round it, and increasing in ma-
turity fi'om the free or distal to
the attached or pruxinuil end.

data. A hydranth surrounded
by a verticil of globular clusters
of gonophores.

no MORPHOLOGY.

it as a naked prolongation, reminding us of the naked prolongation of the spadix in certain
Aracece.

In Clava sqtianiafa (PI. I, figs. 1 and 3) and in CJavamulticonm (PI. II, fig. 1, and woodcut,
fig. 47) the gonophores form dense clusters surrounding the hydranth in a sort of verticil. Each
cluster consists of sessile gonophores borne on a greatly depressed common peduncle, and thus
recalling the form of inflorescence known as a capitulum. The order of development, however,
appears to be centrifugal, instead of being, as in the true capitulum, centripetal, and would
therefore, perhaps, more truly suggest a comparison with the depressed cyme which constitutes
the axillary inflorescence in many Lahiato'..

In the comparison just instituted between the gonosome of the IItdroida and tlie inflores-
cence of plants it will be noticed that, whenever in the Hydroida the generative buds are borne
upon a special gonosomal axis like the flov/ers in an inflorescence, the order of succession is far
more frequently a centrifugal than a centripetal one. In the calyptoblastic forms, indeed, it is
always centrifugal. This is exactly the opposite of what prevails in plants, for here the centripetal
forms of inflorescence greatly exceed the centrifugal ones.

We must be careful, however, not to assign to the resemblances which may be noticed more
importance than they are justly entitled to. Yet, after setting aside such as are merely superficial
and accidental, many still remain which have their origin in certain deep-seated properties. They
may be referred to the common phenomenon of gemmation, which, by agamic multiplication in the
animal as well as in the plant, gives rise to colonies whose members, in each case mutually
dependent on one another, continue to be organically associated into definitely arranged and
determinate groups.

V. Histology.

Both trophosome, gonosome, and coenosome have now been considered in their broader
morphological features ; their morphology, however, cannot be regarded as complete without some
further anatomical details, embracing a histological examination of the tissues.

The Hydroida possess an exceedingly simple structure. Every hydroid, as we have already
seen, is composed of two layers, an ectoderm and an endoderm. Each of these may present in
itself various degrees of differentiation, and we can, perhaps, best study the minute texture of the
tissues by examining them first in the ectoderm, and secondly in the endoderm.

1. Tlic Ectoderm.

General sfructiire of Ectoderm. — In many cases it is impossible to detect in the ectoderm of
the mature hydroid any well-defined structure. A homogeneous blastema with granules and
scattered nucleus-like corpuscles, and thread-cells more or less thickly immersed in it, are all
that the microscope has as yet in such cases succeeded in demonstrating, though a celliUar struc-
ture of this layer can almost always be observed in the embryo.

HISTOLOGY OF THE ECTODERM. Ill

Sometimes, however, the cctoderiu of the adult hydroid may be seen to be composed of
very distinct cells, which can occasionally be isolated without difficulty under the microscope.
In Hydra viridis, for example, the ectoderm can be broken down under the microscope into
spherical cells, in which a nucleolated nucleus may occasionally be detected. These cells are
provided with distinct membranous walls, within which is contained a homogeneous vacuolated
protoplasm, in which are often immersed secondary nucleated cells, and also in many of them one
or more thread-cells.

In some cases in which it is not easy to detect definite structure in the normal state of the
ectoderm, this membrane, by a natural hystolytic decomposition after death, will become broken
up into very distinct cells (PI. I, fig. 4).

Faljmcils. — Leydig^ has called attention to the occurrence in Ili/dra of a minute bristle-like
projection of the surface over the site of each thread-cell, without having any immediate connection
with the thread-cell itself, and Dujardin" had already noticed and correctly described similar
bodies as existing on the surface of the capitula which terminate the tentacles in his Stauridium. ■
Dujardin further compares them to the processes which are emitted by an Actinophrp or an
Acineta. These bodies, however, had been previously noticed by Corda,' where they lie over the
smaller kind of thread-cell which occurs in the tentacles of Hydra ; but he erroneously describes
them as direct prolongations from the thread-cell, so as to constitute with it a special organ,
which he calls an " organ of touch." They have also been examined by Dr. T. S. Wright,*
who follows Dujardin and Leydig in showing that they are not directly continuous with the
thread-cell, and who also maintains their very simple protoplasmic nature. He proposes for
them the name of " palpocils." I can entirely confirm the views of Dujardin, Leydig, and Wright
with regard to those bodies which I have examined in various hydroids (PI. IV, fig. 4).

In the marine hydroids a very delicate structureless pellicle can usually be shown to
exist for a greater or less extent over those parts which are not covered by the ordinary
perisarc. I believe it to be either a simple excretion from the surface of the ectoderm or the
result of a metamorphosis of the most superficial portions of this structure. Sometimes, however,
the place of this pellicle is taken, at least on the tentacles, by an exceedingly thin layer of a
transparent semifluid substance, which seems to possess the properties of sarcode (woodcut,
fig. 48). The minute filaments just described as occurring over the site of the thread-cell would
seem to be mere continuations of this sarcode layer, which also frequently presents here and there
little conical elevations, which have no relation with the thread-cells, and whose summit is con-
tinued into a filament of extreme tenuity (woodcut, fig. 48 <?, e, <?). These filaments were first
pointed out by Wright, who includes them along with the organs of Corda under the common
name of " palpocils."

Leydig is of opinion that there exists over the whole surface of Hydra, except the surface of
attachment of the foot-disc, a very thin homogeneous cuticle. I am inclined to believe that what
Leydig names cuticle is really the sarcode layer, which here, as in the marine hydroids, is con-
tinued into the palpocils.

^ Miiller's ' Archiv,' 1854.

' 'Ana. des Sci. Nat.,' 1843, p. 370.

* Ibid., 2me serie, 1837, p. 363.

*■ ' Proc. Roy. Phys. Soc. Edinb.,' vol. i, p. 311.

112

MORPHOLOGY.

Fig. 48.

FihriUated tmite. — In a great many cases there is developed upon tlie inner surface of tiie
ectoderm a peculiar tissue, forming an abrupt Ijoundary between the ectoderm and endoderm.
It may be well seen in those elongated hydranths which present an extensive surface uncovered by
the more or less opaque chitinous perisarc, such as Clava, Ili/drac-
tinia, and Clavatella, and in the very contractile body and tenta-
cles of the hydranths of different CorynidcB, and in the stem and
hydranth of Corymorplia. In all these cases it presents the appear-
ance of fine, close, longitudinal and parallel strise between the ecto-
derm and endoderm. From the body of the hydranth these stripe
usually extend into the tentacles, and may be very distinctly seen
in the tentacles of Coryne pusilla, where they can be easily traced
as far as the terminal capitulum.

I have succeeded in isolating the fibrillated tissue of the large
tentacles in specimens of Tiibularia indimsa, which had been kept
some years in spirits. The fibrillated tissue here (PI. XXIII, fig. .5)
consists of two layers, one composed of filu'es which take a
longitudinal course parallel to the axis of the tentacle, and the other
of fibres which take a circular course transverse to the axis. The
circular fibres seem to lie externally to the longitudinal, and both
form a muscular envelope which is intimately connected with the
ectoderm of the tentacle, and comes away with the latter when this
is separated from the endoderm.'

The fibres thus isolated appear to be tubular, having a dia-
meter of about 5Tfg^ of an inch ; they are perfectly smooth, but in
most cases a very distinct oval nucleus, having a greater diameter
of about -25770 of an inch, and with a brilliant nucleolus, may be
demonstrated in them. Tliey may sometimes be seen to taper away at each end to a point, when
they present the appearance of greatly elongated fusiform cells (fig. 6).

That the fibres thus demonstrated in the Hydroida constitute a true contractile tissue would
follow, not only from the analogy suggested by structure, but from the fact of contractility mani-
festing itself with great intensity in those parts where they are best developed.^

Part of a tentacle of Syncoryne
^M^c//e//«, showings the superficial
protoplasmic layer.

a. Large-celled tissne of endo-
derm, forming the axis of the ten-
tacle.the cells containing nncleated,
and often radiating, masses of pro-
toplasm; i, fibrillated layer; c,
ectoderm, with immersed thread-
cells ; d, superficial layer of proto-
plasm, which has become extended
here and there into e, e, e, filaments
of extreme tenuity.

' lu some cases the fibrillated tissue would seem to be more intimately united to the endoderm
than to the ectoderm. I have found, at least, that in specimens of C/ava squamata, which had been
preserved in spirits, the ectoderm could be detached, leaving the fibrillated layer still adlierent to the
endoderm.

^ Reichert (" Ueber die contractile Substanz und den feineren Bau der Campanularien," &c.,
' Monatsbericht der Akad. der Wissens. zu Berliu,' July, 1866, p. 504) denies the existence of the con-
tractile fibrillated layer whose presence is here insisted on, and maintains that its place is taken by a
structureless membrane, to which he gives the name of supporting lamella (" Stiitzlamelle"), and
which he regards as a sort of internal skeleton secreted from the inner side of the ectoderm. He
believes that the whole contractility of the hydroid resides in the ectoderm itself, which he says is
entirely destitute of cells or cell-constituents. My own observations will not allow me to adopt the
views of Reichert.

HISTOLOGY OF THE ECTODERM.

113

Fig. 49.

Tlie fibrillatcd tissue is largely developed in tlie umbrella of the medusa, where the fibres
are arranged circularly, and where, instead of lying between the ectoderm and endoderm, they
are situated on the concave surface of the umbrella, as well as in the membranous velum, of
which they constitute nearly the entire substance. We shall examine this tissue in connection
with the other structures which enter into the umbrella of the medusa.

Umbrella of medusa. — A remarkable and important modification of the ectoderm is seen in
the umbrella of the medusa, where it forms nearly its entire mass, the ectoderm becoming
almost whollv changed into a perfectly transparent, elastic, gelatinous-looking substance, traversed
by the gastro-vascular canals, which are lined by endoderm.

The following observations on the structure of the umbrella have been chiefly made on the
medusa oi Syncoryne pulckella shortly after its liberation from the trophosome. In this medusa
(PI. VI, fig. 3, and woodcut, fig. 49) the convex surface of the umbrella is covered by a distinct
epithelial layer (woodcut, fig. 49 a, b) formed of cells (b), containing a nucleolated nucleus,
and separated from one another by narrow
tracts of homogeneous intercellular substance.
These cells appear to consist of simple masses
of protoplasm, without distinctly differentiated
walls. Under the action of acetic acid they
undergo changes of form, becoming elongated
oval or fusiform or conical, or even extended into
irregular processes, while the protoplasm assumes
a moi'e distinctly granular ajipearance, and the
nucleus, with its nucleolus, frequently becomes
more evident. Immediately beneath the epithelial
layer the proper gelatinous substance (a, a) of the
umbrella begins. It is transparent, colourless,
and elastic, and ap|)arently of the consistence of
soft jelly. I have in vain sought for anything
like definite structure in it, though IlaeckeP has
shown that in the GeryoniclfB the homogeneous
gelatinous substance of the umbrella is traversed
by numerous fine branching fibres in a direction
chiefly perpendicular to the surfaces, while the
researches of jMax Schultze" have proved that in
the BiicojjJwra or stcganophthalmic medusae the
corresponding part consists of widely separated
cells, which send out prolongations from their
walls to meet similar prolongations from the neigh-
boiu'ing cells, and wMth a voluminous intercellular
substance, which is composed of a semifluid matter in the meshes of a loose, elastic, fibrous

Structure of medusii of Syneorytie puhhella Bhortly after
liberation from the troijhoaome.

A. Projected view of medusa from tlie summit of the um-
brella ; a, gelatinous substance of the umbrella; 6, external epi-
thelium-layer, and c, internal epithelium-layer of the um-
brella ; d, subumbrellar muscular sac united to the umbrella
along the lines of tlie four radiating canals and along four in-
termediate lines corresponding to the ibur meridional furrows
e, e, e, e, observable on the outer surface of the umbrella; f,
summit of mauubrium.

n. Cells composing the external epithelium of the umbrella.

C. Fibrillated tissue of the subumbrellar muscular sac.

D. Fibrillie of the muscular sac treated with acetic acid,
ami viewed under a high power.

' Op. cit., p. IGO.

' Schultze, "Ueber den Bau dor Gallertscheibe der IMcdnsen," IMiillcr's 'Arch.," 185G.

15

114 .MORPHOLOGY.

network. Scliultze has recognised the close resemblance between this structure and that of
certain forms of the so-called connective tissue.

Though I have failed in my attempts to detect structure in the gelatinous portion of the
umbrella of those hydroid medusse which I have examined, I do not desire on that account to
insist on its absence. The comparatively small size and soft consistence of most of the hydroid
medusae, and the consequent extreme difficulty of obtaining thin slices fitted for microscopical
observation, and freed from the confusing presence of the epithelium and fibrillated layers,
throw much greater obstacles in the way of a satisfactory examination of the imibrella of the
ordinary hydroid medusae than what we meet with in the larger and more easily manipulated
umbrella of the Geryonidce and Stef/anophtlialniafa.

Lying immediately on the concave surface of the gelatinous substance of the umbrella, an
inner epithelial layer (a, c) may under circumstances favorable for observation be demonstrated.
It consists of a single layer of cells, and corresponds to the epithelium of the convex siuface, but
is much more difficult to detect.

The concavity of the umbrella is lined by a sac (a, (/) which lies immediately upon the inner
epithelium layer, and consists of a distinctly fibrillated membrane. The fibres composing it take,
when at rest, a circular course parallel to the margin of the umbrella, and ai-e usually in close
contact with one another, though occasionally they become separated at intervals, so as to leave
numerous fusiform spaces between them (c). They are very fine, measuring about the ^o^oq of
an inch in diameter, and under a high power of the microscope each fibril appears resolved into
a single series of corpuscles, a structure which under the action of acetic acid becomes distinct (d).
At the margin of the umbrella the fibrillated layer leaves the gelatinous bell and is inflected
inwards over the codonostome, so as to constitute the perforated diaphragm or velum.

The fibrillae of the umbrella and velum, which are thus much more minute than those which
have been described in the trophosome, present a marked resemblance to the ultimate fibrillae
of striated muscle ; but, instead of being united into fibres, they are spread out into a broad
membrane.^ That the fibrillated layer forms a true contractile tissue, conferring on the medusa
those active natatory powers which constitute one of its most striking characters, there cannot
be any doubt.

' Busk, in a paper full of excellent observations on the structure of some hydroid medusfe (' Trans.
Mic. Soc. Lend.,' vol. iii, p. 14), describes the muscular fibres in the umbrella of Turris neylecia and
in that of a Thaumantias-V\ke medusa as " distinctly marked with transverse strise."

In the swimming-bells of the Siphonophora the fibrillated tissue is very well developed. In a
small species of Diphya, captured abundantly on the Irish coast, it was easy in very fresh specimens to
get a good view of the contractile fibres which are largely developed in the swimming-bell of this
Siphonophore. Besides the circular fibres, a longitudinal set seems also to be present. The circular fibres
are flattened, and marked with close transverse striaj, which are rendered particularly evident by the
application of acetic acid, which also brings out in the walls of the fibre distinct but distant nuclei
with contained nucleoli.

According to Haeckel, the fibrillated layers of the umbrella and velum of the Gerijonidcc, as well
as that which invests the stiff tentacles which always e.xist in the young state of these medusfe, consist
of very distinctly striated fibres, while smooth muscle-fibres occur in the walls of the manubrium.
He also refers certain fibres and nucleated fusiform cells, which he has detected in the extensile marginal
tentacles of these medusee, to the group of smooth muscle-fibre.

HISTOLOGY OF THE ECTODERM. 115

The muscular sac which thus lines the umbrella in Syncoryne pidehella is not uniformly
adherent to it, and receives its chief attachment along eight meridional lines. Four of these cor-
respond to the direction of the four radiating canals, and the remaining four arc so distributed
that one lies exactly midway between every two radiating canals.

In medusae, such especially as have been kept alive for some time in our jars, the muscular
sac may be occasionally seen to be separated by a considerable interval from the rest of the
umbrella in all the spaces which intervene between the eight meridional lines of attachment,
tiiough it still continues closely adherent to these lines (woodcut, fig. 49 a). A few delicate
bands may here and there be seen, near the summit of the umbrella, stretching transversely across
the spaces betw'cen the umbrella and the detached portions of the sac.

It is in this condition that the inner epithelium layer becomes apparent. I have failed to see
it as long as the muscle-sac is uniformly in contact with the rest of the umbrella.

An inner epithelium layer has been shown by Haeckel to exist in the Geryonidm. There,
however, he describes it as lying on the concave surface of the nuiscular layer.

In the medusa of Obelia geniculata shortly after liberation, the inner epithelium is very
distinct, but I have failed in satisfying myself of the existence of a distinct nmscular layer in the
almost disc-shaped umbrella of this medusa, which, moreover, presents the very exceptional
condition of being entirely destitute of a velum. On the other hand, a distinctly fibrillated
layer may be seen in the marginal tentacles (woodcut, fig. 59 a a), which indeed would seem by
their fin-like action to be far more efficient than the umbrella in the locomotion of the medusa.

The inner epithelium consists here of distinctly nucleated cells with narrow intercellular spaces.
In many of the cells the nuclei were plainly seen to be in process of division (woodcut, fig. 59 c).

The gelatinous portion of the umbrella thus lies between the two layers of epithelium, and is
probably a product of one or both of these layers. In many hydroids the medusa immediately
after liberation has this gelatinous portion still thin, but it often acquires great thickness as the
medusa advances towards maturity — a phenomenon of which Bouguinvillia affords a striking
example (see PI. IX).^

NematojjJiores. — The general form and relations of the nematophores have been already
described (p. 28). The matter which fills the chitinous sheath of the nematophore is a clear semi-
fluid substance with scattered granules, and without the slightest trace of structure, but having
frequently imbedded in it a cluster of thread-cells. It differs in no respect from the sarcode
matter composing the bodies of the Rldzopoda, and like it is capable of emitting true pseudopodia.
When a specimen of Antenmilaria antennina (woodcut, fig. 50) or Aglaophenia pluma (woodcut,
fig. 51 J is examined in the zoophyte trough of the microscope shortly after removal from the sea,
and before it has lost any of its original vigour, the contents of the nematophore-sheaths will be

^ Haeckel (op. cit.) has described in the umbrella of the Geryonidce certain structures which he has
shown to be nearly allied to cartilage, not onh' in consistence, but in histological composition. This
medusa-cartilage forms — 1. A ring which runs in the gelatinous substance of the umbrella parallel to
its margin, and below the circular canal. 2. Certain rib-like structures (mantel-spangen) which are
imbedded in the outer surface of the umbrella, and extend from the margin in a meridional direction
for a greater or less distance towards the summit. 3. The rod-like axis of the stiff, solid tentacles.

In all these cases the medusa-cartilage consists of rouuded, nucleated, occasionally vacuolated,
masses of granular protoplasm, contained in cavities of a clear homogeneous iutercellular substance
having a cartilaginous consistence.

IIG

MORPHOLOGY.

seen extending themselves in the form of long processes and threads of sarcode, sometimes simple
and undivided, sometimes breaking themselves up into branches, sometimes stretching themselves

out as free processes into the surrounding water, and
sometimes seeming to flow over the surface of the
hydrosoma in simple or branching streams ; and then
again the whole will slowly withdraw itself into its
chitinous receptacles, leaving not a trace visible of
those wonderfully extensile processes and filaments of
sarcode into which it had just before transformed itself.
In all this the clusters of thread-cells, when they exist,
remain quite stationary, being never carried out with the
sarcode in its pseudopodial prolongations.'

Thread-cells. — The most characteristic elements of
the ectoderm are the thread-cells. They occur under
various forms throughout the whole of the Ccelenferata,
and though analogous bodies are occasionally found in
some other invertebrate groups, they are nowhere
so abundant and cliaracteristic as in the Ccelenteratar

The form of the thread-ceil varies in different
species of hydroids, and even in different parts of the
same animal ; but it consists essentially of a containing
capsule, and a contained filament, which admits under
certain conditions of being projected from the capsule
(woodcut, fig. 52). The investigation of the thread-
cell, with the view of obtaining a knowledge of its struc-
ture and mode of action, is one of the most difficult
tasks in the anatomy of the Hydroida. The minute
size, great transparency, and almost entire xmiformity of
action on the light, of all the parts of these really com-
plex bodies, and the rapidity with which their charac-
teristic function is performed, renders their study one
which requires no ordinary patience and practice in
microscopical observation, and, notwithstanding the labour
which has been bestowed upon them, our knowledge of
the thread-cell is by no means in all points satis-
factory.

The large thread-cells which occur in the tentacles
and body of Hydra, or in the capitula which terminate
the tentacles in Coryne, may be taken as presenting the
most usual type of these bodies among the Hydroida.

' See my " Report on the Reproductive System of the Hydroida," in ' Brit. As. Rep.' for 1863,
and a paper " Ou the Occm-rence of Amocbiform Protoplasm and the Emission of I'seudopodia in the
Hydroida," 'Ann. Nat. Hist.' for March, 18Gi.

"' Among the more recent authors who have studied the thread-cell in the Coslenterala, refe-

Portion of a ramulus of Antennularia antennina,
with hydrauths and nematophores.

a, Hydranth extended ; b, bydraath retracted ;
e, hydrotheca; (/, d, d,d, consecutive segments of the
ramulus ; e, e, azygous or mesial nematophores, with
tiieir sarcode contents quiescent ; e', e', e', azygous
nematophores, with the sarcode contents emitting
pseudopodial prolongations; f, gemminate or lateral
nematophore-s with pseudopodial prolongations of the
sarcode.

HISTOLOGY OF THE ECTODERM.

117

If we examine one of the large thread-cells from the tentacle of Cori/nr pmilla prcvionsly
to its being exposed to the conditions which result in the emission of its filament, we shall find it to
consist of an external perfectly transparent oval capsule with rigid walls, and of certain remarkable
contents which play au important part in the special function of the thread-cell (PI. IV, fig. 4). The

Fig. 51.

Hydrothecae of Aglaoplienia pluma, with hydrantbs and neraatopliores.

A. Hydrotlieca, with extended hydranth and with the sareode contents of the nematophores quiescent.
a, extended hydranth; c, serrated mar<^in of hydrotheca; d, segment of tlie ramulus carrying tlie liydrotheca;
e, mesial or azygous nematophore; f^ lateral uematophore; g, lateral aperture through which tlie mesial
nematophore communicates with the cavity of the hydrotheca.

B. Hydrotheca, with retracted hydranth and the sareode contents of the nematophores emitting pseudo-
podial prolongations, o, retracted hydranth; c, margin of hydrotheca; d, segment of the ramulus carryings
the hydrotheca; e, mesial nematophore, with its protoplasm projected in an irregular pseudopodial mass;
y, through its lateral aperture into the cavity of the hydrotheca; f^ lateral nematophore, with the commence-
ment of a pseudopodium.

C. Same parts with pseudopodial processes more advanced, a, retracted hydranth ; c, margin of hydro-
theca ; rf, segment of ramulus ; e, mesial nematophore from which a long clavute process, g, of protoplasm is
projected into the cavity of the hydrotheca; fy lateral nematophore from which a long pseudopodium is
projected into the surrounding water.

D. Same parts showing different states of extension of the pseudopodia. a, retracted hydranth ; c, margin
of hydrotheca; d, segment of ramulus; c, lateral nematophore with a branching process,^, of its sareode
projected into the cavity of the hydrotheca ; /, lateral nematophore with the pseudopodium entirely withdrawn.

In all the figures a cluster of thread-cells is seen imbedded in the distal end of the protoplasm within the sheath
of the nematophore.

capsule is completely closed, and its longer axis is occupied by a membranous tube somewhat wider
near the centre than at either end. At one end of this tube, its walls are continuous with those
of the capsule, and it is this part of the capsule which usually lies most superficially when the
thread-cell is imbedded in the ectoderm ; it may, for the convenience of description, be
distinguished as the anterior end. The opposite end of the axile tube loses itself in a perfectly
transparent mass which occupies nearly the whole of the posterior half of the cavity of the
capsule, and in which I have in vain sought for anything like definite structure.

The peculiar phenomena, however, which characterise the evolution of the thread-cell,
together with observations upon the structure of the thread-cell in other species, render it
almost certain that the apparently homogeneous mass in which the posterior extremity of the

rence may be made to Leidy (" Marine Invertebrate Fauna of the Coasts of Rhode Island and New
Jersey," in ' Acad. Nat. Sci. Philadelphia,' vol. iii, 1855), Gosse (' British Sea Anemones,' 1860),
Clark (Agass., ' Cont. Nat. Hist. United States,' vol. iv, p. 209), and more especially to Mcibius
(" Ueber den Bau, den Mechanismus, und die Entwickelung der Nesselkapseln," from the ' Trans-
actions of the Natural History Society of Hamburg,' 18C6), who has given us a very full account of
these bodies, which he Las studied chiefly in Caryophyllia Smiihii.

118

MORPHOLOGY.

Fig. 52.

axilc tube loses itself, is in reality hero, as in other cases, a filament rolled upon itself, but of such
extreme tenuity and transparency as to render it impossible to distinguish the individual coils.

By carefully adjusted illumination, the whole of the capsule appears to l)e lined by a
membrane of extreme delicacy. The existence of an external capsule distinct from its lining
membrane is particularly apparent in the thread-cell after the emission of its contents, for then
the outer capsule is seen to terminate by a distinct slightly everted margin round the orifice
through which the contents have been expelled (PI. IV, fig. 5).

The characteristic action of the thread-cell may be brought into play under the microscope
by some force artificially applied, such as the pressure of the compressorium, or the contact of

alcohol or acetic acid. This action consists
in a sudden change of form, the capsule
opening at its anterior end, from which a
very remarkable body is at the same time
projected with a rapidity which renders it
impossible for the eye to follow it in its
progress (PI. IV, fig. 5, and woodcut, fig.
52 B, b). When, however, this act is com-
pleted, the projected body may be seen to
be an elongated sac whose walls are at one
end continuous through the opening in the
capsule with the membrane by which the
latter is lined, and whose opposite end
tapers away to a point. Just where this sac
begins to thin away towards its point, three
rigid spines are fixed in a verticil to the outer
surface of its walls, their points being directed
backwards like the barbs of an arrow.

The projection of the barbed sac does
not, however, constitute the whole of the
phenomenon presented by the thread-cell
under the influence of the force which
gave 01 igin to its evolution ; for this first
act is followed — in most cases instantane-
ously— by a second, which consists in the
projection, from the pointed summit of the
sac, of a filament (woodcut, fig. 52 c, c)
of great tenuity and transparency, which
rapidly shoots across the field of the micro-
scope until it attains a length of between
thirty and forty times that of the longer
diameter of the original capsule ; at the termination of the emission, one end of the filament
always remains in connection with the free end of the barbed sac. The whole act of evolution is
thus completed, and the capsule is now seen to contain nothing but an absolutely colourless homo-
geneous fluid.

the

Diagrammatic Views of the Thread-cell.

A. The thread-cell previous to the emission of its contents. _, ...
double wall of the capsule; b, reflected portion of the inner membran.,
of the capsule-wall, forming the axile tube of the unevolved thread-
cell ; c, the filament lyiug in complicated coils in the bottom of the
capsule.

B. The thread-cell immediately after the first stage of evolution.
a, the double wall of the capsule ; h, the barbed sac formed by the
eversion of the axile tube; c, the filament still lying within the cap-
sule and barbed sac.

C. The thread-cell after the complete emission of the contents.
'(, the double wall of the capsule; I, the barbed sac; c, the ejected
filament.

inSTOLOGY OF THE ECTODERM. 119

Such are the phenomena presented hy the thread-cell in the act of evolution, but the
amazing rapidity with which the whole takes place renders it extremely difficult to determine
in what the remarkable series of changes now described essentially consists. There can, however,
be no doubt tliat the sudden appearance of the barbed sac is the result of a process of eversion
of tiic membranous tube which in the original state of the thread-cell occupies its axis, and
which is formed by the lining membrane of the capsule inverted into itself at its anterior end.
The eversion of this tube, by which it becomes freed from the restraint of the capsule, is
necessarily accompanied by its sudden expansion and development ; and that this is the real
nature of the phenomenon in question, receives ample confirmation from what may be (vitnessed
in another kind of thread-cell to be presently described.

It is still more difficult to determine the nature of the act which shows itself in the emission
of the filament, than of that which constitutes the emission of the barbed sac. Previously to
emission, no trace of a filament can be detected in the thread-cell now under consideration ;
but, as already said, there can be little doubt that the transparent homogeneous-looking mass in
which the posterior part of the axile tube loses itself before its eversion has commenced is
nothing more than the filament closely coiled on itself (as shown diagrammatically in woodcut,
fig. 52 A).

It is probable that the expulsion of the filament, like that of the barbed sac, consists also
in an act of eversion ; a view which would of course render it necessary to regard it, notwith-
standing its extreme tenuity, as a flexible membranous tube.

It must be borne in mind, that the barbed sac and the filament are in no way continuous
with the outer rigid wall of the capsule, but only with the delicate sac by which this is lined.
The outer wall opens at its summit by a definite orifice which appears to have been previously
closed by a minute lid which is thrown off" in the act of evolution, and which may be occasionally
seen after the completion of this act adhering in the form of a httle disc to the edge of the now
expanded orifice, through which the contents of the capsule had been emitted.

Another form of thread-cell which throws light on the structure and action of that just
described, occurs also among the Hydroida. In Hydra, it is in the form of a minute oval
capsule, much smaller than the former, and having its cavity occupied by a spirally coiled
filament which may be easily seen through its transparent walls. The act of evolution consists
in the emission of the filament, but the barbed sac which constitutes so important a feature in
the evolved thread-cell of the first kind cannot here be distinguished.

In some other Hydroida, the second kind of thread-cell acquires a larger size than in
Hydra. I have carefully examined it in Gemmaria implexa, where it is met with scattered in the
ectoderm of the hydranth, and where, from its comparatively large size, it is well fitted for
observation. It here consists of an oval capsule (PI. VII, fig. 9), within which may be seen, pre-
viously to emission a long thread, consisting of a straight and a coiled portion. The straight
portion crosses the capsule diagonally, and is continuous at one end with the walls of the capsule,
while at the opposite end it begins to be rolled up in distinct coils which almost completely till
the capsule. The coiled portion of the thread, however, does not surround the straight portion,
but is placed entirely on one side of it, as had been already noticed by Clark in the thread-cells
of Syncoryne mirahilis}

' Clark, loc. cit.

120 MORPHOLOGY.

The plienonienon of emission consists here, as in tlic form of thread-cell first described, of
two distinct acts (see PL VII, tig. 10). The immediate result of the exciting cause is the projec-
tion of a long tulmlar thread from the anterior end of the capsule. This thread attains the length
of about eight times that of the longer diameter of the capsule ; and on the completion of its
emission one end remains attached to the capsule, while the free end presents a small oval dila-
tation, behind which the thread is furnished for a short distance with very minute spines, which
seem to be arranged spirally on its walls.

The projection of this thread thus constitutes the first step in the evolution of the thread-
cell ; but it is no sooner completed, than a still iiner thread is shot out with great rapidity
from its free extremity to a length about equal to that of the first. The whole act of emission is
thus completed, and the capsule seems entirely emptied of its contents.

It is probable that the smaller kind of thread-cell in Hydra lias a structure similar to that
now described, though the minuteness of the parts unfits it for a satisfactory demonstration. It
is evident, too, that the long thread with its armature of minute spines, whose emission con-
stitutes the first stage in the action of the last-described thread-cell, is the exact homologue of
the barbed sac of the first kind, the fine terminal filament in the two being also homologous
organs.

The first stage in the emission of the contents of the capsule in the second kind of thread-
cell takes place occasionally so slowly, froui accidental resistance, that the course of the thread
may be easily followed by the eye ; and then it will be plainly seen to roll outwards by an
act of eversion, reminding us of the mode of extension of the tentacles of a snail. The
emission of the finer thread from the extremity of this is, probably, also by an act of eversion ;
but I have never succeeded in following it so as to obtain direct evidence that it is so.^

A fact which has an important bearing on the structure of the thread-cell is the difference
of capacity presented by this body according as it is examined in its unevolved or evolved state.
It would seem that, in some cases, the capacity of the capsule after evolution is less than it was
previously, and then we can understand that no necessity may exist for its emitted contents being
replaced by others. In certain cases, however, the total cavity of the thread-cell after the evolution
may occupy a space more than double that which characterised it before the commencement of this
act. Now, as this occurs in a sac, one portion, at least, of which must consist of a flexible
membrane which would yield to any pressure from without not counterbalanced by an equal
pressure from within, and as no tendency to collapsing can be detected in the evolved thread-
cell, it is plain that, simultaneously with evolution, a fluid must be admitted into it from without,

' The observations of Mobius (loc. cit.), made chiefly on the thread-cells of CaryophyUia Smitldi,
have led him to describe the axile tube as consisting of three tubes, included by invagination one within
the other. I cannot find direct evidence of tliis triple invagination in the thread-cells of any of the
Ilydroida I have examined. It must, however, he borne in mind that the much larger size of the
thread-cells of Car?/o/i/(?///«a render them more favorable for the determination of structural details ;
while this view of the axile tube receives support from the fact that the proximal portion of the exserted
filament in the thread-cell of Gemmaria greatly exceeds in length the straight portion of the unexserted
filament; for this portion is the representative, in the non-evolved state, of the proximal portion of
the evolved filament, and the difference between the two can scarcely be otherwise explained than by
supposing the straight unevolved portion to consist of a tubular filament several times invaginated into
itself.

HISTOLOGY OF THE ECTODERM. 121

unless we suppose tliat the cavity of the thread-cell is occupied by a compressible and expansible
aeriform fluid. This latter supposition cannot be maintained, and if the former view be
accepted, we must either admit the power of rapid imbibition in the walls of the cell, or
suppose the presence of some definite opening, not yet detected, through which the surrounding
water may gain access to its cavity.

The two forms of thread-cells now described may be regarded as the principal types
under which these bodies occur in the Hydroida. Of the nature of the force by which the
emission of their contents is effected we cannot yet speak with certainty. Everything that has
been observed, however, is opposed to the supposition that the act of evolution is a vital one,
dependent on the irritability of certain tissues which enter into the structure of the thread-cell.
It is far more probable that it is simply physical, depending on the mere elasticity of the parts,
and brought into play when the internal structures are mechanically released from the tension in
which they had been held during the previous state of repose.

One fact may here be mentioned which is inconsistent with the idea of the evolution of the
thread-cell being a vital act ; namely, that prolonged immersion in certain re-agents, such as
alcohol, will sometimes have no effect in destroying its characteristic properties ; for I have seen
the thread-cells of hydroids which had remained for months immersed in alcohol retain the
power of emitting their contents on the application of some other re-agent, such as acetic
acid.

There is reason to believe that, in some cases at least, the thread-cell when brought into
use is at the same time forcibly ejected from the ectoderm in which it had been previously
imbedded, and not merely drawn out of its berth by its attachment to the prey ; and there is no
reason why this act should not be referred to an irritability residing in the ectoderm, and
receiving its special stimulus from the conditions which rendered necessary the employment of
the thread-cell, such as the contact of living prey.

The employment of the thread -cell by the hydroid would, under these circumstances, involve
both a vital and a physical act — the vital manifested in the ejection of the thread-cell from the
body of the hydroid, the physical in the emission of its contents.

The special purpose fulfilled by the thread-cells in the economy of the animal, and their
probable employment as urticating organs, will be considered under the section which treats of
the physiology of the Hydroida.

While the thread-cells in the Hydroida are entirely confined to the ectoderm and its
appendages, they are by no means uniformly distributed in it. It is in the tentacles of both
hydranth and medusa that they are usually most abundant ; and here we find them either
generally scattered through the ectoderm, or distributed through it in knot-like or wart-like or
verticellate groups, or collected together in the spherical capitula in which the tentacles of some
genera terminate (see Pis. IV, V, &c.). In many medusae the ectoderm forms, at the base of the
marginal tentacles, bulb-like thickenings which are loaded with thread-cells (see woodcut, fig. 56e).
In many cases, minute thread-cells singly, or in clusters, are scattered superficially in the walls of
the umbrella ; in the genus Gennnaria (PI. VII) we also find four piriform chambers extending
from the circular canal of the medusa into the umbrella-walls, and filled with small oval thread-
cells which have the appearance of lying loose within their cavity (figs. 3 and G) ; while numerous
similar but smaller sacs, filled with thread-cells, may be seen in irUlia opening into the circular
canal, and thence extending in a meridional direction along the walls of the umbrella.

16

122

MORPHOLOGY.

Fig. 53.

In Gemmaria are also found certain remarkable pedunculated sacs which are developed from
the marginal tentacles of the medusa, and filled with thread-cells (PI. VII, figs. 3 and 4). Thread-
cells are also, in many cases, specially developed in the nematopliores- — those peculiar sarcode
appendages already described as characterising the family of the Plmnularido!.

It is almost certain that the thread-cells are always developed in the interior of proper
cells, and they may be frequently separated from the ectoderm with the generating cell still
surrounding them.

From observations which I have made on the larger kind of thread-cell in Hydra, it would
seem that a portion of the protoplasm of certain ectodermal cells (generating cells) becomes

differentiated as a spherical or oval mass which may be seen
to. occupy a vacuole in the midst of the remaining proto-
plasmic contents of the cell (woodcut, fig. 53), and in which
one or more nuclei are usually apparent. This little mass is to
become developed into the thread-cell. It continues to in-
crease in size, and becomes invested by a distinct cell-wall,
while its contained protoplasm becomes, in a way which 1
have not been able to follow, metamorphosed into the proper
contents of the thread-cell. After this the now mature
thread-cell becomes free by the rupture of the cell in which it
had been generated.^ The development of the thread-cell
may thus to a certain extent admit of comparison with that
of spermatozoa — a comparison already made by Leuckart —
but we are still far from having a satisfactory notion of the
origin and mode of development of these remarkable bodies.
To the ectodermal structures also belong the sense-
These will be considered in the section on the physiology of the

Development of thread-cell in Sydra.

a. Generating cell of the thread-cell detached
from the ectoderm, containing a very pale
yellowish protoplasm, from which a small mass
has become separated, and occupies the interior
of a vacuole, where it looks like a nucleus in a
cell.

b. The separated mass of protoplasm has
hecome larger and more oval ; two nucleolus-
like corpuscles were in this instance apparent
within it.

c. A developed thread-cell occupies the place
of the protoplasm-mass in the vacuole of
a and h.

organs — ocellus and lythocyst.
Hydroida.

2. The Endoderm.

Cellular structure of endoderm — contents of the cells. — The demonstration of a definite struc-
ture in the endoderm is much easier than in the ectoderm. Here, indeed, we seldom meet with
any difficulty in making out a very distinct cellular composition.

Most usually, two sets of cells may be detected in the endoderm : an external set, with mostly
clear, colourless contents, having but few granules ; and an internal set, which forms the immediate
boundary of the somatic cavity, and contains abundance of coloured granules.

The separation between these two sets of cells is often quite abrupt. In many hydroids
those composing the external set are large, elongated, and with their longer axis in the direction
of the radius of a transverse section of the hydioid, while those composing the internal set are

' Miibius has noticed that iu the development of the thread-cells of Lucernaria and of certain
Aclinozoa amoeboid changes of form are occasionally exhibited by the generating cell. Op. cit., p. 10.

HISTOLOGY OF THE EiNDODERM. 123

raucli smaller and more spherical (see PI. VII, fig. 5). Sometimes, liowcver, the two sets pass
imperceptihly into one another without any distinct boundary-line.

In Hydra viridis, the cellular structure of the endodcrm is veiy distinct. The contents of
most of the cells composing it are here peculiar, and consist of a colourless protoplasm with very
definite green corpuscles imbedded in it, and in almost every instance with one and occasionally
with two clear vacuolae excavated in it.' It was the occurrence of these vacuolaj which caused
Ecker,- by overlooking the proper cell-wall, to adopt the erroneous view that the whole tissue of
Hydra was merely a mass of vacuolated protoplasm.

Among those cells which lie most internally and form the immediate boundary of the
somatic cavity in Hydra viridis, are many which are destitute of green corpuscles, but contain
brown, irregular granules, mostly included in a secondary cell which is itself imbedded in the
vacuolated protoplasm of the mother-cell.

The green corpuscles possess a very definite form— a circumstance in which they contrast
strongly with the brown granules. They are spherical, and present in then- interior a lighter-
coloured space which gives them a close resemblance to thick-walled cells. They will be again
referred to. (See below, p. 136.)

Along with those cells which contain the green corpuscles, there also occur, especially in
badly fed Hydra, others in which the green contents are replaced by smaller spherical but
colom-less bodies, which are probably the green corpuscles in an undeveloped or transformed con-
dition. Occasionally, also, an irregular mass of brown granules may be seen in the same cell with
the green granules.

The cells which thus compose the endoderm of Hydra possess but a weak union with one
another ; they are easily separated by a slight force, and on becoming free immediately assume
the spherical figure, without any trace of their having been previously united into a tissue.

No green matter is developed in the cells of any other species of Hydra. In Hydra fusca,
there occur among the cells which form the boundary of the stomach cavity, many which are of an
elongated piriform shape, with the broad thick end projecting into the cavity, and with their thin
ends imbedded among the others. Within these piriform cells, secondary cells may usually be
detected ; these are spherical in shape, sometimes having clear colourless contents with a nucleus,
while in other cases they are filled with irregular brown granules, and present no evident nucleus.
Sometimes the piriform cells contain only free brown granules, while we may also often meet
with instances in which, besides the free granules, the same piriform cell will contain the clear
nucleated secondary cells, and the secondary cells filled with brown granules. This may be
regarded as the typical condition of the same parts in the other IIydroida, though it is seldom
so distinctly demonstrable as in Hydra.

The walls of the somatic cavity are probably in all Hydroida, if we except Hydra, clothed
to a greater or less extent with vibratile cilia. These ciha are remarkably distinct among some of
the TuhularidiE ; while iu other cases in which their existence has not been proved by direct
observation, the peculiar currents visible in the fluid contents of the somatic cavity leave no doubt
that it is to the agency of such ciha that these currents are mainly due. Hydra would appear to
constitute a solitary exception in the non-ciliated condition of its somatic cavity.

1 Allman, "On the Structure oi Hydra viridis," in 'Brit. Assoc. Reports' for 1853.

- Ecker, " Zur Lelire vom Ban mid Leben der cuutractilen Sub.stanz der niedersteu Thiere,"
' Zeitsch, f. wiss. Zool.,' Bd. 1, 1849.

124 MORPHOLOGY.

The endoderm of the proper digestive cavity is probably in all Hydroida thrown into
more or less complicated lobes or ridges, which disappear where this cavity passes into the common
cavity of the coenosarc.

This condition may be well seen in Ili/drafusca, especially when, as sometimes happens, we
have an opportunity of looking down into the stomach of the uninjured animal through the widely
open mouth. Numerous ridges may then be witnessed, extending from the walls of the cavity
almost to its centre. These ridges are rendered papillose by the projection from their surface of
the peculiar piriform cells already described, while the furrows between them are comparatively
smooth. I have witnessed a similar condition of the endoderm in the digestive cavity of
Cordyhphora lacustris, Coryne pusilla, St/ncoryne eximia, and many others, where thick, irregularly
sinuous and lobulated ridges of endoderm project into the stomach from the inner surface of its
walls. A furrowed and lobulated state of the endoderm may also be witnessed in the manubrium
of the medusa, and in the spadix of the sporosac of many hydroids.'

The hydranth of Tubularia indivisa presents a very remarkable condition of its endodermal
layer (PI. XXIII, fig. 1). Immediately within the mouth, the cells containing the coloured granules
form a narrow smooth zone of vermilion dots, immediately behind which the surface of the endoderm
is disposed in irregidarly oval prominent vermilion patches, separated by paler narrow sulci. In
tracing these patches backwards, we find that they become smaller and more numerous, gradually
resolving themselves into small spots, and ultimately into minute scattered puncta which at the
posterior extremity of the hydranth, where its cavity passes into that of the stem, become
more densely grouped, and are here arranged in radiating bands of a bright vermilion colour.
Along the line where the anterior contracted portion of the hydranth passes into the wider
basal portion, the endoderm is thrown into numerous remarkable pendulous lobes of a piriform
shape and bright vermilion colour (PL XXIII, fig. 3). They consLst "each of a cluster of very
distinct spherical cells, containing vermilion-coloured granules, among which are numerous small
clear spherical elements, apparently oil-drops. They present a close resemblance, both in their
form and in the nature of their contents, to the zone of gland-like lobes which occupies a very
similar position in the interior of the hydranthal zooids of certain Siphonophora.

Canaliculation of endoderm. — In some Hydroida we have a very peculiar and exceptional
condition of the somatic cavity and of its endodermal lining ; for while this cavity in most
Hydroida consists of a simple tube, it is in the cases here alluded to composed of numerous
intercommunicating canals.

This condition is well seen in different species of Tubularia. The stem of Tubularia indivisa
for example, presents immediately within the perisarcal tube a continuous layer of ectoderm
enclosing the endoderm which extends to the very centre of the stem, and thus obliterates all trace
of a central somatic cavity (PI. XXIII, fig. 7). The place of this cavity, however, is suppHed
by numerous canals, which are excavated in the endoderm and take a longitudinal course through
the stem, occasionally communicating by lateral offsets with one another, and finally all merging
in a common central cavity at the base of the hydranth.

In a transverse section of the stem, the canals present a wedge-shaped form, the narrow end

* In certain Geryonidce (Glnssocodon eurylnu, Carmarina hasiata), Haeckel (op. cit.) describes, as
gastric glands, peculiar leaf-shaped organs which occur in the inner walls of the manubrium (four in Glosso-
codon. six in Carmarina). They are composed of clusters of large cells with dark-coloured contents.

HISTOLOGY OF THE ENDODERM. 125

being directed towards the axis, and arc seen to be arranged in a single zone at some distance
from the centre. They are manifestly simple tubular lacuuic interposed among the cells of the
endoderm, and destitute of special walls. They are irregular in size, one of them especially being
usually considerably larger than the others ; and Agassiz, who was the first to call attention to this
difference of size, lays considerable stress upon it, for he regards it as constant, and considers
the large channel to represent that which alone constituted the cavity of the Tubularia in its
young state. I have made many sections of Tubularia stems, and have always found the tubes
irregular in size, and in most cases one of them considerably surpassing the others, as affirmed by
Agassiz. Wright, who first distinctly drew attention to the tubular lacuna, represents them as of
equal size ; but his figur* must be regarded as merely diagramatic. I am not, however, prepared
to assent to Agassiz's view of the origin of the large channel. I believe, on the other hand, that the
simple somatic cavity of the young Tubularia is represented in the adult by the common chamber
at the summit of the stem, into which the longitudinal channels all open. It is only in the free
stage of Tubularia that its cavity is simple ; and immediately after it has become fixed the hydranth
is carried upwards by the development of a stem whose cavity exhibits, at the moment it can be
detected, the compound character of the adult.

That portion of the endoderm which occupies the axis of the stem consists of large cells of a
spherical form, or by mutual pressure more or less polygonal ; they are filled with a clear colour-
less or slightly granular fluid, while the more peripheral portion of the endoderm is composed of
small spherical cells containing abundance of minute vermilion-coloured granules ; it is in this
peripheral portion of the endoderm that the lacunaj are excavated. The walls of the lacunae
are clothed with very long and distinct vibratile cilia.

A structure which in all essential points resembles that just described in Tubularia indivisa
occurs also in Corymorplia nutans, the endoderm of the stem being here, as in Tubularia,
excavated by longitudinal lacunar canals (PI. XIX, figs. 1, 6, 7). These canals inosculate here
and there with one another ; they are much more numerous than in Tubularia, but towards the
base of the stem they liecome by mutual coalescence less numerous than in the distal part ; they
liejust within the ectoderm, and open above into a common cavity in the basal portion of the
hydranth. Here the endoderm which forms the floor of the hydranth-cavity rises as a broad
conical projection so as to nearly fill the posterior half of the cavity. The axis of this projection is
perforated by a narrow canal by which the cavity of the hydranth is continued to the summit of
the stem. At this point the canal becomes wider, and receives the longitudinal channels of the
stem. The peripheral portion of the endoderm, or that in which the canals are excavated, is com-
posed of small spherical cells with reddish-brown granular contents, while the rest of the endoderm
,is composed of loose cells with clear colourless contents, and forms a thick pith-like column
which occupies the whole axis of the stem. I have not succeeded in detecting cilia in the canals,
but, from the very distinct currents which may be witnessed in these canals in the living hydroid,
I have no doubt of the presence of cilia here as in Tubularia.

The superficial position of the endodermal canals both in Tubularia and Corymorpha renders
them conspicuous without dissection, even to the naked eye, for they give to the stem the appear-
ance of being marked by longitudinal bands from one end to the other.

In Anfennularia antennina the coenosarc in the main stems presents also a canaliculated
condition, one, however, which differs in some important points from that of Tubularia and Conj-
morj]ha ; for, while in the two latter the canals of the stem are mere lacunae excavated in a

126

MORPHOLOGY.

Fig. 54.

fomnion endodemi, the whole cfcnosarc is in Anfemmlaria hri)ken up into sepiu'ate tubes, cacli
with its own endoderm and ectoderm, and all surrounded by a common chitinous perisarc
(woodcut, fig. 54).

Twenty or more tubes will thus be found in
the main stem of a specimen of Antenmdaria.
They lie close upon the inner surface of the peri-
sarc, and are connected to one another by an
extension of their ectoderm (woodcut, fig. 54 b).
In some parts of their course they are straight and
parallel ; in others, more or less sinuous and re-
ticulated by inosculation. The whole centre of
the stem is empty, and Las no communication with
the somatic cavity, instead of being occupied, as
in Tiihidaria and Corymorpha, by an endodermal
j)ith-like core, or instead of being, as in most of
the Hydkoida, occupied by the somatic cavity
itself. Very distinct currents may be witnessed
in these tubes, especially in the yoiuig hydroid,
thus affording evidence of the existence of vibra-
tile cilia on their walls.

Endoderm of tentacles. — In the tentacles of
the Hydroida, both those of the hydranth and of
the medusa, the endoderm usually undergoes a
])eculiar and important modification. In Hydra
the tentacles are quite pervious throughout their
entire length, while they are lined by a simple
continuation of the endoderm of the digestive
cavity ; but in every other hydroid with which
I am acquainted, the endoderm of the tentacles
of the hydranth presents a peculiar septate ap-
pearance, looking in some cases like a solid coi'e
divided by numerous transverse septa into cir-
cular discs, and entirely filling up the tube of the
tentacles, while in others, though the same septate condition exists, there is still a narrow pervious
canal in the axis.

I believe that in all these cases the apparent septa are really the opposed walls of large
adjacent endodermal cells which encroach more or less on the cavity of the tentacle, often even to
its entire obliteration.

This structure may be well seen in the tentacles of various species of Corynidce [Coryne
pusilla, PI. IV, fig. 3, &c.). For some distance from the distal extremity of the tentacle,
the cavity is usually entirely obliterated by thin membranous septa which stretch transversely
across it. Further towards the base of the tentacle, the septa become less regular, and the
obliteration of the tube is now plainly seen to be caused by the large irregular cells of which the
endoderm is here composed, and which encroach upon its cavity to such an extent as entirely to

Canalliculatioii of Endoderm iu Antennularia antennina.

A. Very young hydroid, just after the fixation of the planula
and the elevation of the primordial stem, a, a, the canals of
the endoderm, inosculating here and tliere with one another;
b, common cavity, formed by the coalescence of the endo-
dermal canals at tlie summit of the stem ; c, commencement of
a ramulus; d, lobed hydrorhizal disc, forming the surface of
attachment of the young hydroid ; e, circular groove, commence-
ment of segmentation in the stem.

B. Part of a transverse section of the adult stem, a, ecto-
derm, and, i, endoderm of coenosarcal canals ; o', extension of
the ectoderm in the intervals of the canals ; c, perisarc.

HISTOLOGY OF THE ENDODERM. 127

fill it; while at a little distance from the base the encroachment of the cndodcrmal cells is usually
not so complete, and the cavity of the tentacle, though greatly contracted, still continues
j)ervious, and admits into it fluid with suspended particles from the digestive cavity of the
hydranth.

Many of the cells which thus constitute the peculiarly modiiicd endodcrin of the tentacle
contain a very distinct uucleolated nucleus, roimd which is generally collected a colourless finely
granular protoplasm, which is often kept in connection with the cell-wall by radiating pro-
longations. A few large coloured granules are also usually contained in the cells. Towards the
distal extremity of the tentacle, where the septa are most regular and complete, we generally find a
little granular mass collected in the centre of the distal surface of each septum, or else forming a
little axile column stretching from septum to septum. These septa, with their intervening
chambers, have much the appearance of large cells which stretch entirely across the tube of the
tentacle ; it is possible, however, that they may be formed by the coalescence of two or more cells
from opposite sides, as in the more proximal portions, but here accompanied by the disappearance
of their walls in the axis of the tentacle, and the consequent confluence of their cavities.

In some other cases the endodermal cells are not developed to such an extent as to obliterate
in any portion the cavity of the tentacle, which thus remains pervious in its entire length ; but the
large size and regular form of these cells continue to confer on the tentacle a distinctly septate
appearance. I have found this condition in the tentacles of Garveia nutans (PI. XII, fig. (5).

In the planoblast also the tentacles usually present the same septate appearance as in the
hydranth, the endoderm being frequently so developed as entirely to obliterate the tube. Examples
of this may be seen in the marginal tentacles of most medusae, as, for instance, those which
constitute the planoblasts of Obdla dichofoma, and of Podocoryne carnea, as well as in the ten-
tacles of the free sporosac of Dicoryne. Those cases, however, in which the tentacle pre-
serves a complete continuity of its tube are much more common in the medusa than in the
hydranth. In the medusa, indeed, it is far from uncommon ; we find it, for example, in the
marginal tentacles of the medusa of Syncoryne eximia (PI. V, fig. 4), Gemmaria implcxa (PI. VII,
fig. 4), and Sarsia stranyula/a (woodcut, fig. 17), and, according to Haeckel, in the long extensile
tentacles of the Gcryonidce}

In the bulb-like expansion which usually exists at the base of the marginal tentacles of the
medusae, at the point where the radiating canal meets the circular canal, the endoderm acquires
increased thickness, and often forms lobulated projections which encroach upon the cavity, but never
obliterate it (woodcut, fig. 56). These projections are usually rich in cells filled with coloured
granules, which, like the coloured granules of the digestive cavity, appear to be secreted in
secondary cells developed in the interior of mother-cells.

The endoderm lining the remainder of the gastrovascular canals is, on the other hand, entu-ely
destitute of lobes or ridges. It is very thin, and would seem to have a purely distributive
function, the fluid contents of the canals being propelled by vibratile cilia which clothe their
walls. These cilia may be well seen in Syncoryne eximia (woodcut, fig. 50).

' Eeichert (loc. cit.) maintains the entire absence of the endoderm in the tentacles of the sertu-
larian and campanularian hydranths, ivhile he regards the septate appearance as produced by simple
extensions of his " supporting lamella," which, according to him, forms a layer secreted on the inner
surface of the tentacular ectoderm. This view, however, is opposed by the structure, which easily
admits of demonstration in the tentacles of many tubulariau hydroids.

128 PHYSIOLOGY.

PHYSIOLOGY OF THE HYDROIDA.

In attempting a classification of the various phenomena of hydroid life, we are at once met
by the difficulty of arranging them under definite physiological heads. The low grade of speciali-
sation on which the Hydroida stand, renders it often impossible to assign to these phenomena a
definite physiological significance, and many of the acts which make up the life of such simple
organisms can scarcely claim to be referred to one more than to another of the great classes under
which the functions of animals are usually distributed.

Notwithstanding this, however, we shall find it convenient to speak of the vital acts of the
Hydroida under the following heads : — 1. Digestion. 2. Circulation, Nutrition and Growth of
the Tissues, and Respiration. 3. Secretion. 4. Contractility. 5. Sensation. 6. Phosphores-
cence. 7. Reproduction. We must, however, keep in mind that these classes do not all neces-
sarily possess the definiteness which characterises them in the higher and more specialised
members of the animal kingdom.

1. Bigestion.

So far as observation has taught us anything regarding the food of the Hydroida, we may
conclude that it consists mainly of living animals, though Diatomaceae and other minute free
vegetables contribute also to their subsistence. The patient observations of Trembley and the
older naturalists on the fresh-water Hydra had long ago shown that animals of considerable size,
such as entouiostraca, and even aquatic larvae and worms, are seized by the tentacles of this
voracious little hydroid, and carried to the mouth, through which they are borne into the cavity
of the body ; and subsequent observations have proved that in other Hydroida the nature of the
food and the capture of the prey are in all essential points similar to what has been noticed in
the Hydra .

It is in this act that the functions of the thread-cells seem to be manily called into play ;
and repeated observations have shown that no sooner do the tentacles come in contact with the
living prey, than all power of resistance in the latter is at an end ; its efforts to escape seem
suddenly paralysed, and it becomes an easy victim to the rapacity of its captor.

If the prey be at this stage released from the grasp of the hydroid and placed under the
microscope, all the soft tissues of its surface will be found pierced with discharged thread-cells.

Observers are by no means agreed as to the true functions and mode of action of the thread-
cells. By most they are supposed to penetrate the tissues of the prey, and those which, like the
larger thread-cells of Hydra, are provided with an apparatus of barbed spines, have been described
as plunged beyond the barbs into the soft tissues of the victim. Other observers, and more

DIGESTION. 129

especially Mobius, reject tbe idea of penetration, and maintain that the thread-cells arrest the
motions of the prey by the mere adhesion of the ejected filament.

I believe, however, that it will be found that there is an actual jicnetration, but not to the
extent that is usually insisted on. If some soft body, such as a worm, be brought into contact
with the tentacles of a Hydra, the surface of contact may be seen immediately afterwards to be
covered with large discharged thread-cells, many of which will be found with the freed end of the
barbed sac inserted into it as far as the roots of the barbs. I have never, however, witnessed a
deeper penetration than this, and the barbs themselves were never immersed.

I believe that in such cases the action of the thread-cell consists in the sudden ejection of
the barbed sac against the tissues of the prey, which, if these be soft enough, allow the point of
the sac to penetrate as far as the roots of the barbs. This act is instantly followed by the ejection
of the filament, for which the barbed sac has opened the commencement of a passage, and which
now worms its way among the tissues, recalling the mode in which the delicate filaments which
form the mi/cclium of certain parasitic fungi penetrate the organic structures infested by them.

It is impossible, however, to believe that such effects as follow the action of the thread-cell
can be produced by mere mechanical penetration, and the conclusion is irresistible that the pene-
tration is accompanied by the injection of some potent virus which acts by rapidly destroying the
irritability of the living tissues, the tubular filament of the thread-cell as it continues to insinuate
itself, affording at the same time a channel by which the special secretion of the cell is conducted
into the deeper parts of the tissues of the prey.

In judging of the functions of the thread-cells, however, it must not be forgotten that they
also occur in parts which preclude the possibility of their being employed as offensive or defensive
organs, as, for instance, when they exist, as they frequently do in great numbers, in that part of
the ectoderm which is under cover of the hard chitinous perisarc, or when they are included in
the interior of cavities which can have nothing to do with the capture of the prey, as in the
sac-like receptacles which occur in the umbrella of Gemmaria.

Though we have scarcely yet sufficient data to enable us to determine the exact limits of
that part of the somatic cavity on which the function of digestion specially devolves, we shall
perhaps be justified in considering as such that portion of it which is included within the hydranth,
and which is almost always distinguished by its form and by some special structure of its walls
from that which belongs to the cmnosarc; while in the medusa the cavity of the manubrium, or
at least its basal portion, must also be regarded in a special sense as devoted to the digestive
functions of the free planoblast.

When the food has once passed the mouth and entered the cavity of the hydranth, or that
of the manubrium of the medusa, it is there subjected to a process of solution. Of the nature
of the solvent we as yet know almost nothing. There can be .little doubt, however, that it is
secreted from the walls of the cavity; and though it must be more or less diluted with water which
has obtained admission through the mouth, its action on the food is powerful and rapid. This
process is doubtless aided by the motion to which the contents of the digestive cavity are subjected
by the contraction of the walls and the vibration of the cilia which clothe them. The soluble
and nutritious parts are speedily separated, and the insoluble and non-nutritious dchris are ejected
through the mouth.^

' Corda (" Anat. Hydra; fuscJE," in 'Ann. des Sci. Nat.,' 1H37) has described the body-cavity of
Hydra as communicating with the external medium by an orifice situated at the end opposite to tbe

17

130 PHYSIOLOGY.

The product of tliis process, consisting of the dissolved and disintegrated nutritious portion
of the food, mingled with certain materials which it has received from the walls of the digestive
cavity, and diluted with water which has been introduced from without, is propelled from the
cavity of the hydranth into that of the coenosarc, or from the manubrium into the radiating canals
of the medusa, in order to be distributed to the several tissues. It thus becomes the somatic fluid,
to be presently considered in its relations to the circidatory functions.

Examined under the microscope, this fluid is seen to be of a very heterogeneous nature. Its
basis is a transparent colourless liquid, and in this solid bodies of various kinds are suspended.
These consist partly of disintegrated elements of the food, partly of solid coloured matter which
has been secreted by the walls of the somatic cavity, partly of cells, some of which have undoubt-
edly been detached from these walls, though it is possible that others may have been primarily
developed in the fluid, and partly of minute irregular corpuscles, which are possibly some of the
effete elements of the tissues.

Beyond this point it is plain that digestion cannot be traced as a specialised function, and its
phenomena here become coincident with those of circulation.

2. Circulation, Nutrition, and Respiration.

Between the phenomena associated above under the genei'al head of digestion, and those
which might with equal justice be claimed by circulation, it is impossible to draw any well-
marked line of demarcation. The somatic fluid, whose relations to digestion we have been just
considering, has relations quite as intimate with circulation ; for though in one aspect we may
compare it to the chyme or chyle of the more highly specialised animals, in another it admits of
just as close a comparison with the blood.

No trace of a differentiated blood vascular system has been detected in any hydi'oid. The
nearest approach to it is that which is presented by the radiating and circular canals of the
medusa, and yet these are simple oS'sets from the digestive cavity.

The place of the blood is taken by the fluid which pervades the somatic cavity, and which
consists of the digested food largely mingled with water which has gained admission by
the mouth, as well as with certain materials which have been secreted by the walls of the
somatic cavity.

The fluid which thus permeates the somatic cavity and extends through all its ramifications
must not be supposed to be in a state of rest. On the contrary, it is subjected to constant
motion, which manifests itself in currents more or less regular in their velocity and definite in
their direction.' In some cases the currents present a remarkable definiteness and regularity.

mouth. 1 have not been able to confirm this observation ; and though Leydig (" Ueber den Bau der
Hydren," in ' Miiller's Archiv,' 1854) and Hancock ("Notes on a Species of Hydra," in 'Ann. Nat.
Hist.,' 1850) support Corda's statement, I cannot avoid believing that the orifice in question is
accidental.

' Cavolini was the first who distinctly noticed the currents of the somatic fluid in the Hydroida.
He has described them in a eampanulariau liydroid (Sprengel's 'Cavolini,' p. 56). They were subse-

CIRCULATION, NUTRITION, AND RESPIRATION. 1:31

This may be well seen in the stems of Tu/jularia indivisa, where the contained fluid is distriljuted
in numerous parallel streams ascending through some of the channels which are excavated in the
endoderm of the stem of this hydroid, and descending through others. The ascending streams
become confluent at the summit of the stem, and from this point the descending streams issue.
Every now and then the direction of the streams is reversed, the fluid flowing down the channels
which just before carried ascending streams, and flowing up in those through which the streams
had previously descended. Here and there the longitudinal channels conununicatc with one
another by transverse branches, and thnmgh these branches the fluid may be seen flowing from
one longitudinal cliannel to another. Occasionally the fluid will remain for a short period at rest
in one of the channels, while it continues to flow in its ascending and descending streams in the
others. The somatic currents of Tiibularia, however, are by no means always easily seen, and it
is only now and then that a specimen occurs, which, from the transparency of its tissues, and
the abundance of floating corpuscles in the somatic fluid, presents conditions favorable for
observation.

In Cori/morpha mduns, the structm'e of whose stem resembles that of Tuhidaria indimm,
similar definite currents may be witnessed ; and in Anfennuluria aidennina, whose cceuosarc, as
already mentioned, is composed of numerous tubes within a common perisarc, currents of an
entirely similar kind may be seen ascending in some of these tubes, descending in others, passing
over from one to another by intercommunicating branches, and again resting in others, but ready
to start ofi' in a fresh stream upwards or downwards while we continue to watch.

It is rare, however, to find the currents of the somatic fluid presenting the regularity and
definiteness which characterise them in the instances just mentioned. In those far more nume-
rous cases in which the somatic cavity is simple instead of being composed of separate and dis-
tinct channels, the contained fluid may occasionally be seen in a single broad stream flowing
through the axis of the stem or branch, then coming to a state of rest and after a short peiiod
of repose, starting off' anew in a reversed direction. Most usually, however, the movement in
such cases, instead of being in the form of a definite current, consists in an irregular commotion
in which the floating corpuscles are whirled about in all directions. Within more limited cavities,
such as that of the spadix in the sporosac and of young buds in process of development, the
motion of the fluid is generally in circular streams, which may be seen coursing round the walls
of the cavity and returning into themselves. These streams in the developing bud usually pre-
sent great activity. A similar disposition of the streams in circular currents may be witnessed
in the cavity of the hydranth ; but the motion here ought probably to be referred to digestion
rather than to circulation, as it does not seem to have any proper distributional office, being most
likely specially connected with the preparation of the aliment.

In the gastrovascular canals of the medusa, the currents are very distinct, the transparency
of the surrounding tissues rendering it easy to observe them. In all these canals, whether radiat-
ing or circular, they constitute simple streams, which in some of the radiating canals may be seen
flowing from the proximal towards the distal end, and in others from the distal towards the proxi-
mal, and, as it would seem, reversing themselves every now and then in all these tubes. In the

quently described by Lister, who also detected them in the stem of Tubuluria indivisa (' Phil. Trans.,'
183-4). Since then the currents in the somatic cavity of the Hyduuida have become familiar to every
observer of these animals.

132 PHYSIOLOGY.

circular canal the fluid may also be seen flowing at one time in one direction, and then again in
the opi^osite.

That the motion of the somatic fluid in all these cases is mainly caused by the impulse of
vibratile cilia, there can be but little doubt. Dujardin appears to have witnessed these cilia in
the stem of his Syncoryne decipiens ; ' and in the stem of Tuhularia indivim, where the cilia
clothing the lacunar channels and their relation to the currents were first pointed out by Dr. T.
S. Wright, ° they are very distinct, and can be demonstrated with the greatest ease in a transverse
section of the stem (PI. XXIII, fig. 7). By carefully applied pressure, I have succeeded in ejecting
the spadix in an inverted state through the proximal end of a detached sporosac of Tubidciria
iiidivi-m, when its internal surface being thus exposed, a rich clothing of very distinct actively
vibrating cilia has been brought into view (PL XXIII, fig. 10).

It must, however, be admitted that in the greater number of hydroids the existence of endo-
dermal cilia has not been proved by direct observation. In most hydroids the tissues are too
opaque to atford a view of the delicate cilia which may clothe the cavities, while sections of the
stem will aflbrd none of that aid which we obtain from them in Ttibtdaria ; for the section is
accompanied in most other species by a collapsing of the cavity which would necessarily interfere
with the action of the cilia and remove them from observation. In every case where a satisfactory
view of any portion of the free surface of the endodenn in the Hydroida has been obtained, as,
for instance, where the transparency of the tissues is such as to afford no obstacle to a view of the
deeper seated parts, this smface is invariably — except in Hydra, where cilia appear to be really
absent — seen to be clothed with vibratile cilia. Thus, in certain meduste the cilia may be easily
seen vibrating along the walls of the radiating and circular canals.

It is probably, then, a nearly universal fact that the free surface of the endoderm in the
Hydroida is ciliated, and that the fluid in contact with this surface is kept in motion by the
impulse it receives from the cilia.

It uHist not, however, be overlooked that the contractility of the walls has its share in this
motion. The fluid contained within the gastric cavity of the hydranth may, indeed, be frequently
seen to be forcibly expelled from this cavity by its contraction into that of the coenosarc — the
mouth of the hydranth being at the same time kept closed — and then again drawn back into the
gastric cavity, when the contraction which had expelled it is succeeded by an expansion.

As every portion of the somatic fluid is thus successively brought in contact with the walls
of the somatic cavity, it yields to the tissues by a process of direct absorption the nutriment
necessary for their growth and maintenance, and receives from them such portions of their sub-
stance as in the performance of their vital functions had become effete.

That the formation and growth of the tissues is intimately connected with cell-formation
is rendered obvious by attending to the phenomena which accompany rapid growth among the
Hydroida.

A healthy colony of Obclia dicliotoma was placed in a jar of sea-water. In a few days most
of the hydranths had disappeared, but adventitious branches had begun to be sent out in great
profusion from various parts of the surface. These branches elongated themselves with

' ' Ann. des Sc. Xat.,' tome iv, 1845, p. 275.
- ' Proc. Roy. Ph. Soc. Edinb.,' 1855-50.

CIRCULATION, NUTRITION, AND RESPIRATION.

13:3

astonishing rapidity, in some cases at tlic rate of tlircc quarters of an incli in thirty-six

Tiie distal extremity of the branch was its growing point, and here it terminated in a slight
enlargement which formed a cul de sac, in whose walls both cndoderm and ectoderm were dis-
tinctly difTerentiatcd (woodcut, fig. 5")). The somatic
fluid penetrated to the extremity of the branch. The
ectoderm formed a considerably thicker layer than the
endoilerm, and was composed of very distinct, nearly
spherical cells, with clear granular contents, and with
but little intercellular plasma. Numerous slightly
curved, cylindrical bodies, resembling undeveloped
thread-cells, appeared to lie free in the intercellular
])Iasma, where they were scattered among the true
cells.' Under the action of acetic acid a distinct nu-
cleus was rendered evident in the ectodermal cells.

The endoderm of the growing point was also dis-
tinctly composed of spherical cells in a very scanty
intercellular plasma. The endodermal cells, however,
were much smaller than those of the ectoderm, and,
instead of the clear contents of the latter, the endo-
dermal cells contained coloured granules.

On following the cells of the ectoderm backwards
from the growing point, they were found to become
less and less distinct ; it was, indeed, plain that they
had undergone a metamorphosis, and at a short dis-
tance from the extremity no true cells could any longer
be detected in the ectoderm, which now presented in
a sectional view only an obliquely striated and granular appearance. The endoderm, on the other
hand, retained its distinctly cellular structure throughout. The perisarc, which on the proximal
part of the branch possessed considerable thickness, showed itself on the growing part as a scarcely
perceptible pellicle.

I did not succeed in detecting anything like a process of cell-division in the ectoderm of the
growing point, nor could I satisfy myself that the cells became multiplied by endogenous forma-
tion of cell-broods within the cavity of the older ones. The appearances were, on the other hand,
rather in favour of the direct formation of the cells out of the plasma.

The cells of the endoderm probably originate in the same way, but they appear soon to
become ruptured and to discharge their contents into the somatic fluid, which filled the cavity of
the growing stem.

In certain other tissues, growth would seem to take place by cell-division. This, at all
events, appears to be the case in the epithelium of the umbrella of ObeUa (see below woodcut,
fig. 59, c).

Growing extremity of an nilvcntitions branch of
Obelia dichotoma.

a. Ectoderm overlaid by a delicate pellicle of
cliitine, und composed of distinct cells, where it occu-
pies the growing point ; b, endoderm ; c, ectoderm,
somewhat removed from the growing point, its cells
no longer evident ; rf, some of the cells from the ecto-
derm of the growing point, treated with acetic acid,
more highly magnified, and with thread-cell-like
bodies lying scattered among them.

^ This structure recals that of certain young sponges, with the spiculae which are to form the
skeleton scattered among more or less individualised masses of protoplasm.

134 PHYSIOLOGY.

The currents descriljcd aliove, as existing in tlie somatic cavity of the Hydroida, serve not
only for the distribution of the nutritive fluids among the tissues — an office in which they admit
of comparison with the somatic circulation of the blood in the more specialised members of the
animal kingdom — but they also contribute to liring this fluid into more immediate relation with
the surrounding medium, and thus subserve the function of respiration.

No differentiated respiratory organs occur anywhere among the Hydroida, and the inter-
change between the nutritious fluid and the surrounding medium must take place through the
general surface of the body, more especially through those portions of it which, not being under
cover of the chitinous perisarc, are more directly exposed to the conditions under which such
interchange may be effected.

The cilia which cover the external surface of the hydroid when in the stage of planula, as
well as those which clothe the free sporosac of Dicoryiie, and those which occur on certain hydroid
medusBe not yet traced to a trophosome, such as Tracltynema, Gegeng., are probably subservient
to respiration as well as to locomotion.

3. Secretion.

That distinct secretions are found among the Hydroida, and that even special structures
are set aside for their elaboration, there cannot now be any doubt.

One of the most marked of these secretions consists of a coloured granular matter which is
contained at first in the interior of certain spherical cells, and may afterwards become discharged
into the somatic fluid. These cells, as already mentioned, are developed in the endoderra, in
which they are frequently so abundant as to form a continuous layer upon the free surface of this
membrane. It is in the proper gastric cavity of the hydranth and medusa, in the spadix of the
sporosac, and in the bulbous dilatations which generally occur at the bases of the marginal
tentacles of the medusa, that they are developed in greatest abundance and perfection ; but they
are also to be found more or less abundantly in the walls of probably the whole somatic cavity, if
we except that portion of the gastrovascular canals of the medusa which is not included within
the bulbous dilatations.

In the parts just mentioned as affording the most abundant supply of these cells, they are
chiefly borne on the prominent ridges into which the endoderm is thrown in these situations ;
when they occur in the intervals between the ridges, they are smaller and less numerous.

The granular matter contained in the interior of these cells varies in its colour in different
hydroids. In many it presents various shades of brown ; in others it is a reddish-brown or light
pink, or deeper carmhie or vermilion or orange ; or occasionally a fine lemon-yellow, as in the
hydranth of Coppiiiia arcta, or even a bright emerald green, as in the bulbous bases of the mar-
ginal tentacles of certain medusa;. No definite structure can be detected in it ; it is entirely
composed of minute granules, irregular in form, and usually aggregated into irregularly shaped
masses in the interior of the cells. It is to this matter that the colours of the Hydroida, varying
as they do in different species, are almost entirely due.

The coloured granular matter is undoubtedly a product of true secretion ; and the cells in
which it is found must be regarded as true secreting cells. These cells arc themselves frequently

SECRETION. 135

to he seen as secondary cells in the interior of parent cells, from which they escape by rupture,
and then falling into the somatic fluid, arc carried along by its currents, until ultimately, by their
own rupture, they discharge into it their contents.

We have no facts which may enable us to form a decided opinion as to the purpose served
by this secretion. Its being always more or less deeply coloured, and the fact of its being
abundantly produced in the digestive cavity, might suggest that it represented the biliary secretion
of higher animals. This may be its true nature, but as yet we can assert nothing approaching to
certainty on the subject: indeed, considering how widely the cells destined for the secretion of
coloured granules are distributed over the walls of the somatic cavity, it would seem not impro-
bable that the import of the coloured matter may be different in different situations ; that while
some of it may be a product destined for further use in the economy of the hydroid, more of it
may be simply excretive, taking no further part in the vital phenomena, and intended solely for
elimination from the system.

The fluid which acts so powerfully as a solvent on the food which has passed into the
digestive cavity, must certainly be also regarded as a secretion. We are ignorant of its exact
source, but it is in all probability derived, like the coloured matter, from cells developed in the
walls of the cavity.^

Under the head of specific secretion must also probably be classed the fluid contents of the
thread-cells.

That the remarkable green matter which is contained in the endodermal cells of Hydra
viridis is a special product of these cells there can be no doubt. Its very definite structure, how-
ever, would seem to take it out of the class of ordinary secretions, and place it rather in the same
group of products as spermatozoa and thread-cells.

The researches of Cohn' have led him to believe that the green colouring matter of Hydra is
identical with that of Euc/lena, and with that of Loxodes, Stenfor, and certain other green Infusoria;
and, further, that in all these cases it is indistinguishable from the chlorophylle granules of certain
Alija, especially of Faticheria. He has shown that the green granules in these animals, as well as
in a green Flanaria which he has also examined, present precisely the same appearances under
the action of sulphuric acid as those which we witness in the chlorophylle of plants when sub-
jected to the same treatment ; for instead of remaining unchanged, or merely becoming charred,
the granules, when brought into contact with the acid, pass in a very characteristic way fi-om
verdigris-green to a more intense bluish-green, and at last, in solution, become almost blue.

From these facts Cohn concludes that the green matter in the organisms mentioned, at least
in Eughna and the green ciliate Infusoria, performs a function similar to that of the chloro-
phylle of plants, and he regards it as destined for the excretion of oxygen.

The chitinous perisarc which, to a greater or less extent, invests the surface of almost
every hydroid, is perhaps rather a product of metamorphosis of tissue than of true secretion, the
most external portions of the ectoderm becoming converted into the perisarc, which increases
in thickness by successive additions to the inner surface, these additions giving a distinctly

' The cells described by Haeckel as arranged in peculiar leaf-like groups in the stomach-walls for
certain medusaj belonging to the family of the Geryonidai, -srould seem to be destined for the elaboration
of some special secretion.

- Cohn, "Beitr. zur Entwckel. der Infusorien," ' Zeit. f. Wiss. Zool.,' vol. iii, 1<Sj1.

136 PHYSIOLOGY.

laminated condition to the otherwise absolutely structureless perisarc. To the same class of
products belongs the gelatinous-looking investment which envelopes the acrocyst in Serfuhria
pumila, &c. (see p. 50), and in which a distinctly laminated character may frequently be
detected.

4. Coniractility .

The existence of a fibrillated tissue in the IIydroida has been already mentioned ; and there
can be no doubt that this tissue is endowed with contractility, and is tiie proper seat of the more
energetic motions performed by these animals. The act by which the hydranth becomes suddenly
retracted when touched is evidently due to the contraction of the longitudinal fibres which are
developed on the inner surface of the ectoderm; and the rhythmical and exquisitely graceful
movements of the mednsre have their seat in the contractile fibres which are developed on the
concave surface of the innbrella and in the velum, and are antagonised by the elasticity of the
gelatinous substance which constitutes the chief mass of the umbrella.

It is, indeed, in the motions of the medusae that we find contractility manifesting itself in
its highest degree of intensity among theHi'DRoiDA. In these beautiful zooids the contraction of
the sub-umbrellar fibres necessarily diminishes the cavity of the umbrella, and by thus expelling
in a jet a portion of the water which had filled it in its expanded state, causes the propulsion of the
medusa in an opposite direction, or that in which the convexity of the umbrella is turned forward,
wdiile the relaxation of the fibres immediately following their contraction permits the elasticity
of the umbrella to come into play, so that its cavity instantly resumes its original capacity, and
receives within it a fresh supply of water, to be again expelled as a propelling force by the ener-
getic contraction of the fibres. The part played here by the velum would seem to consist chiefly
in controlling the diameter of the aperture through which the jet of water is propelled from the
umbrella cavity.

But contractility is by no means confined to the fibrillated tissue. There arc many parts of
the Hydroida in which no trace of fibres can be detected, and which are yet eminently contrac-
tile. Thus the broad four-lobed lip of the medusa of Obelia geniculata is remarkable for its
mobility and its power of constantly changing its form while under observation, and yet it
consists exclusively of a single layer of differentiated masses of protoplasm (membraneless cells)
without the slightest trace of fibres.

Ecker has shown that the body of Hydra viridis is to a great extent composed of a contrac-
tile, semifluid homogeneous substance, agreeing in all essential points with the sarcode matter of
the lowest animals. Ecker, however, is wrong in denying a cellular structure to Hydra, for the
sarcode is not only differentiated into distinct cell-masses, but these masses are included each —
in most cases along with certain granular products — within a proper cell-wall. By the rupture
of the cell-wall the protoplasm can be liberated under the microscope ; and Ecker has seen it
then undergoing evident contraction, forming isolated masses, which continually change their
shape like an Ammha.

There can be little doubt that many of the motions of the Hydroida are due to the con-
tractility of this homogeneous sarcode ; and while we may refer the sudden retraction of the
hydranth when touched to the action of the fibrillated tissue, its subsequent slow extension would

SENSATION. 137

seem to be the result of tlic contractility of the sarcodc, probably combined with the general
elasticity of the tissues.

I have already shown that the substance which fills the nematophores of the Plumularkla; is
mainly composed of a similar sarcode, which here, however, is not confined in cells; and it can,
therefore, extend itself beyond the surface of the hydroid in the form of long ])sendopodia, which
will occasionally even branch exactly as in certain JRhizopoda (sec woodcuts, figs. 50, 51). The
remarkable capsules tilled with thread-cclls, which are borne along the marginal tentacles of the
medusa of Gemmaria implexa (PI. VII, fig. 3, 4), are supported on peduncles of extraordinary
extensibility. In these peduncles we have also an example of true sarcode identical with that of
the BJdzopoda. Strethill Wright has also shown that those filiform processes of the ectoderm of
the Htduoida, which he calls " palpocils," are composed of a true rhizopodal sarcode ; and these
processes are probably offsets of a very thin sarcode layer, which, as already mentioned, can be
seen, in certain cases, to extend over the surface of the hydranth.

Under the head of contractility the phenomena of ciliary motion must also be included.
Vibratile cilia, as we have already seen, exist in almost every case on the walls of the somatic
cavity, and mainly contribute to the production of the currents so well known in the nutritive
fluids of the Hydroida ; while the planula or early locomotive stage of most hydroids, as well as
certain hydroid medusae {Trachynemd) and the free sporosac of Dicori/ne cojiferta, are provided
with an external covering of vibratile cilia.

5. Sensation.

Of late years several observers have believed themselves successful in demonstrating a
nervous system in the Hydroida, while others have refused to admit the existence of a differen-
tiated nervous system in these animals, and consider the arguments which have been adduced in
favour of its presence as resting upon imperfect or incorrectly interpreted observations.

The advocates for the existence of a specialised nervous system in the Hydroida all agree in
regarding as its principal part an apparent filament with ganglion-like enlargements, which may be
seen running in the form of a ring round the margin of the medusa just below the circular canal.
Agassiz maintains the existence of such a ring-like cord, and assigns to it the function of a nerve-
ring, in Sarsia, Tinropsis, Sfaurophora, and BoiicjainvilUa ; M'Crady in Euclmlota ; Fritz Miiller in
Tamoya, a genus belonging to the family of the Cliaryhdrnda, as well as in Liriope and Cunina,
true hydroids ;* and Leuckhart in a medusa which he refers to Gegenbaur's genus Eucope ; while
Hensen also admits the presence of a nerve-ring in the Eucope of Gegenbaur. But by far the most
complete description we possess is that by Haeckel," who has studied the structures in question

' The hydroid structure of the (Eginida, in which Cunina is included, must follow, from the
remarkable observations of Haeckel on the structure of Cunina and on its genetic relations with the
Geryonidan Medusse.

" Haeckel, op. cit.

18

1 3S PHYSIOLOGY.

with great care in certain medusa3 belonging to tlie family of the Geryonida, more especially
Glossocodon {Liriope) curyhia and Carmarwn liastata, where it would appear that they are very
conspicuous, and well fitted for examination.

According to Haeckel the nervous system in the Geryonidce consists of a very slender ring-
like cord, which runs round the margin of the umbrella immediately below the circular canal, and
under each marginal vesicle swells into a ganglion. From these ganglia filaments are sent off,
one along the course of eath radial canal as far as the stomach, and one to each of the tentacles,
while another penetrates the marginal vesicle in order to undergo within it a peculiar distribution.
In the system thus constituted he believes that he has succeeded in demonstrating nerve-elements.
The nerve-cells of the nervous ring are contained in the ganglia only ; in the intervening portions
the ring presents merely a longitudinally striated appearance.

While with Haeckel's very detailed description we should hardly be justified in denying
the existence of a marginal nerve-cord with ganglia in the Geryonidce, I am by no means
prepared to attribute a similar significance to the cord-like structure which may be seen
running round the margin of the mnbrella in other medusae (PI. XVI, fig. 8, and woodcuts,
figs. 58 and 59). I have in several cases carefully studied this part of the medusa, and have
arrived at the conviction that in all these the apparent cord is only the ectodermal layer, which
lies immediately upon the distal side of the circular canal, and constitutes the extreme margin
of the umbrella, presenting, when viewed along the plane of the codonostome, the appearance of
n chord. In the medusa of Campanularia, which Gegenbaur refers to his genus Eacope, and in
that of Obelia, which he includes in the same genus,^ I have no doubt as to this being the true
interpretation of the supposed nerve-cord, while the structures assumed to be ganglia are only
thickenings of the ectoderm at the points which give attachment to the litltocyds or sense-
capsules, to be presently described.

Besides the structures to which the significance of a nervous system has been thus
assigned, certain bodies which have been long known and regarded by common consent as organs
of sense, hold a prominent place in many hydroids. They occur only in the meduste, and are of
two kinds — the ocellus and the lUhocyst.

The Ocellus. — The ocellus consists of a little mass of pigment, forming a well-defined
coloured spot, in some species black, in others vermilion or deep carmine. In most cases (PI. V,
fig. 3, PI. VI, fig. 3, and PI. IX, fig. 8) no other structure can be detected in the ocellus ;
l)ut sometimes (PI. XVII, fig. 5, and PI. XVIII, fig. 6) a transparent, refracting body may
be seen immersed in the pigment mass. The ocellus is always situated in the walls of the
bulbous dilatation which exists at the root of the marginal tentacle of the medusa (woodcut,
fig. 50), where it lies very superficially, being imbedded exclusively in the ectoderm, while a

' Tiie genus Eucope was founded by Gegenbaur for certain small medusse ■which are known to be
the planoblasts of campauularian hydroids. He subdivides Lis genus into a deep-belled and a shallow-
belled section — sections, however, which differ from one another by characters which are at least of
generic value. The deep-belled forms are the planoblasts of the true Campamdaria, while those with
shallow bells are the planoblasts of another campanularian genus {Laomedea of authors, in part), and
liad already been described by Peron and Lesueur, under the generic name of Obelia. Eucope, as a
generic appellation, must therefore be suppressed in favour of the older names of Campamdaria and
OheUa, by which the deep-bcUed and shallow-belled forms must be respectively designated.

SENSATION.

139

thin layer of the latter passes over it so as to separate it from direct contact with the surrounding
water. In Ti/aropsi.s alone does a similar definite pigment mass occupy a different position, hi
this genus it is situated not at the root of a

tentacle, but at the base of a lythocyst, in ■*^"^- '■• •

the interval between two neighbouring
tentacles (woodcut, fig. 57). I cannot,
however, regard the pigment-shot in T^a-
ropsis as tlie equivalent of the ocellus in
other medusa; ; for besides its connection
with the lithocyst rather than the tentacle,
it is imbedded in a thickening of the endo-
derm of the circular canal instead of being,
as in the true ocellus, an exclusively ecto-
dermal structure.

When the pigment of which the ocellus
is composed is examined under a high power
of the microscope, it is seen to possess a
more definite structure than the granular
coloured matter which is secreted in the
walls of the somatic cavity. The ocellus is,
in fact, composed of an aggregation of very
minute cells, each filled with a homoge-
neous coloured matter. Such, at least, is
the stnicture of the ocellus in Sijncoryne cximia, and it is probable that in other cases a similar
structure may be demonstrated.

As just said, a transparent body, capable of refracting the light, may be sometimes seen
immersed in the pigment. A very distinct refracting body, of a lenticular shape, has been
described by Quatrefage as imbedded in the outer side of the pigment in Eleidheria. In the
nearly allied Clavcdella I have found a minute, spherical, transparent, and refracting body imbeddeii
in the same way in the pigment of the ocellus (PI. XVIII, fig. 5) ; and Krohn and Claparede
had already made a similar observation. It is, however, here of soft consistence ; it seems to be
easily broken down, and I have occasionally failed in detecting any trace of it. A similar refracting
lens-like body may be seen in the ocellus of Cladonema (PI. XVII, fig. 6). In every other case
with which I am acquainted the ocellus consists merely of a mass of pigment-cells, without any
structure which can serve as a refracting medium, unless the transparent layer of ectoderm whicli
is continued in front of it may be regarded as contributing to the functions of the ocellus by its
refractive action on the rays of light. In Ehuthcria and Clavafella, indeed, the ectoderm presents
at this spot an abruptly prominent convex surface, which Quatrefage has compared in Eleutheria
to a cornea.

We have no means of forming anything like a certain conclusion as to the proper function of
the ocellus. The universal presence of a definite pigment, and the occasional occurrence of a
refractile, lens-like body, have suggested a comparison with an organ of vision, and against the
justice of this comparison no sufficient argument has been yet adduced.

The L'dliocyd. — Like the ocellus, the lithocyst is invariably developed on the margin of the

Piirt of the umbrella-margin, with basal bulb of margiual tentacle,
iu Syncorijne exlmla.
rt. Umbrella wall ; i, distal extremity of a radiatin^^ cannl ; —
vibrutile cilia maybe seen clothing its walls; c, part of circular
canal ; rf, endoderm of bulbous dilatation of base of marginal ten-
tacle ; — it is thrown into prominent lobes, which project into the
cavity of the bulb ; e, cushion-like thickening of ectoderm loaded
with thread-cells;/, marginal tentacle ; ^, ocellus, imbedded in a
thickening of the ectoderm.

140

PHYSIOLOGY.

Fig. 57.

Part of the umbrella-margin of Tyaropsis scotica.

c, lithocyst ;
endoderm ;

a. Circular canal ; b, b, marjrin.il tentacles ;
d, ectoderm of distal side of circular canal ; (
f, projection of the ectoderm, lying at the inner side of the
lithocyst ; ff, ocelliform spot, imbedded in the endoderm of
the circular canal.

umbrella. It consists of a transparent, mostly spherical capsule, williin which are contained one
or more transparent refractile concretions generally of a spherical or oval form. In Cunina alone

among the Hydroida the concretions are, ac-
cording to Haeckel, in the form of ciystals, the
usual condition of the analogous bodies in the
Biscophora or steganophthalmic medusse. In
most cases there is but one of these bodies in each
capsule. Sometimes, however, they are more
numerous. In Tuna Bairdii the number varies
from four to twenty in different lithocysts of the
same specimen. In T^arojjsis they are arranged
in a regular crescent parallel to the distal wall of
the capsule, as Agassiz has pointed out in a
North American species of this genus, and as I
have myself found in a Tyaropsis from the
Scottish coast (woodcut, fig. 57), in which I have
counted about thirteen concretions in each capsule.
The concretions offer resistance to pressure ; and, according to Gegeubaur, would seem to
consist of carbonate of lime in an organic basis, which retains the form of the original
body after the mineral matter has been removed by acid. The mineral constituent is, according
to Haeckel, a phosphate of lime in the Geryonida. At least, it here dissolves in acids without
effervescence. The concretions, in every case, are quite motionless, and never lie free in the
capsule.

A careful study of the lithocysts in two very common hydroids — Campanularia Jo.h/sfoni nr\d
Ohelia [Laoniedea) geniculala — has rendered apparent the following facts, which may probably be
regarded as representing, in all essential points, the usual structure of these bodies among the
Hydroida.

In Campcamlaria Johnsloni there are eight lithocysts which alternate with the eight marginal
tentacles of the medusa ; each lithocyst (woodcut, fig. 58/<) is immersed for a little way in the
cord-like structure, which forms the extreme margin of the umbrella, and which sends a very
delicate extension of its substance over the whole of the free surface of the lithocyst. It
consists of a spiierical transparent and structureless vesicle or capsule, the greater part of whose
cavity is occupied by a soft spherical pulp, in whose distal pole, or that opposite to the point
of attachment of the vesicle, there exists a deep well-defined excavation ; and within this, but
not entirely filling it, is the spherical highly refractile concretion. In the pulp itself I could
detect no trace of structure, but when seen in profile it has a slightly wavy outline, possibly occa-
sioned by a special layer which intervenes between it and the walls of the capsule. Its surface
is marked by twelve or fifteen delicate striae, which take a meridional course at exactly equal
distances from one another. Towards the distal pole they all terminate distinctly on the margin
of the excavation, and may be thence traced to within a short distance of the opposite pole,
though I have never been able to follow them exactly to it. The striae generally appear
light-coloured when contrasted with the darker intervening spaces. It is often difficult
to detect any trace of them, but with a high power and carefully adjusted illumination

SENSATION.

Ml

they frequently appear with great distinctness, more especially under the action of dilute acetic
acid.'

Fig. 58.

Part of the Umbrella-margin in the Medusa of Campanularia Johnatoni.

a, cavity of circular canal; b, cavity of a radiating canal; c, c, e, c, endodermal lining of
radiating and circular canals ; rf, rf, d, d, ectoderm forming tlie extreme margin of the umbrella ;
e, ectoderm of marginal tentacle ; /, thickened ectoderm at the base of the tentacle loaded with
thread-cells and directly continuous with d, the ectoderm of the umbrella-margin ; g, marginal
ectoderm thickened where it lies over the spot from which a new tentacle is to spring;
ht lithocyst.

In Obclia genicidcda, the litbocysts are also eight in number, two being situated in each of the
four interradial spaces (woodcut, fig. 59 Kg, and By). They are sessile on the cord-like margin
of the umbrella.and are placed each at the subumbrellar side of the base of a tentacle, but not
exactly in its meridian plane. The structure of the lithocyst resembles that just described in
CampamiJaria Johnsfoni, consisting, like it, of a spherical capsule with a contained pulp, which
nearly fills it, and which is excavated at the distal pole of the capsule into a cavity in which
is the spherical refracting concretion. I could, however, obtain no distinct evidence of the meri-
dional striae visible in the lithocvst of C. Johidoni?

' Busk appears to have noticed these striffi, for in a paper which he read before tlie Microscopical
Society of London (' Trans. Mic. Soc. Lond./ vol. iii, p. 22), and which gives the most complete
account of the lithocyst we had possessed up to the time of its publication, the author describes " au
indistinctly fibrous, radiating appearance," which might sometimes be detected in the more external
parts of the contents of the lithocyst iu a Thaumantias-\^& medusa.

- Keferstein and Ehlers ('Zoolog. Beitrage,' p. 88, pi. xiii, fig. 4) describe the Eucope polystijla
of Gegenbaur, which is, however, really an Obelia. They give a good figure of the lithocyst, but they
mistake the cavity in which the concretion is contained for a peduncle projecting into the capsule from
its distal wall, and carrying the concretion on its summit. A similar mistake seems also to have been
made by Fr. Miiller (" Ueber die Randblaschen der Hydroidenquallen," in Max Schultze's ' Arch. f.
Mic. Anat.,' 1865, vol. i, p. 143), in his account of the lithocyst of an allied Medusa.

142

PHYSIOLOGY.

In most cases no del'uiitc structure can be detected in the walls of the external capsule. In

an nndescrihed species of Tynropsis, how-

FiG. 59.

Medusa of Obelia geniculata.

A. Part of the umbrella-margin, a, a, marginal tentacles; b, h
thickened ectoderm at the base of the marginal tentacles; c, c, large
cell forming the proximal extremity of tlie marginal tentacle where it
is plunged into the gelatinous substance of the umbrella; d, d, d, ecto-
derm forming the extreme margin of tlie umbrella; e, circular canal;
/,/, gelatinous substance of the umbrella ; g, lithocyst.

B. Meridional section of umbrella-margin, a, marginal tentacle;
A, thickened ectoderm at base of marginal tentacle; c, large cell form-
ing the proximal extremity of marginal tentacle; d, ectoderm of ex-
treme margin of umbrella ; e, circular canal in transverse section ; /,
gelatinous substance of umbrella ; (j, lithocyst.

C. Structure of epithelium clothing the inner surface of the umbrella.
In feveral of the cells the nuclei have undergone division.

ever (Tj/aropsis scotica, woodcut, fig. 57),
these walls are composed of very distinct
prismatic cells, lined with a finely granular
layer, which immediately invests the pulp
and tlie zone of concretions imbedded in its
surface, and I have noticed a similar struc-
ture in another medusa allied to Tliaumantias.
Gegenbaiir has pointed out the existence of
a layer of polygonal cells, which lines the
capsule in one of the Geri/on'uhe, an observa-
tion which has been confirmed by Haeckel in
other members of this family ; and in the
medusae of Campanularia and Obelia de-
scribed above there is indication of a special
lining of the capsule.

Hensen, in his remarkable memoir on
the auditory organ of the Decapod Crusta-
cea, describes the lithocyst of a medusa,
which he refers to Gegenbaur's genus
Eitcope, and maintains that the concretion
is connected to the walls of the capsule by
a bundle of very fine hairs (" auditory
hairs").' This observation of Hensen, how-
ever, has not been confirmed by other in-
vestigators, and I consider it almost certain
that the meridional strife described above
have given rise to the appearance which
Hensen has mistaken for hairs.^

' " Studien iiber das Gcliiirorgan der Decapoden," ' Zeit. fur Wissensch. Zool.,' vol. ^\\\, p. 355,
note.

- The lithocyst has been studied with great care by Haeckel (op. cit.) in Glossocodon eurybia and
Carmarina hasiata, Geryonidan Medusse which, from their large size, offer special facilities to the
observer, though they belong to a peculiar type, and their litliocysts, as described by Haeckel, present
features which differ in many points from the structure which, as we have already seen, these bodies
present in Campanularia and Obelia.

In the Medusaj which form the subject of Haeckel's observations the lithocyst is completely included
within the gelatinous substance of the umbrella-margin. It is described as a transparent sphere seated
on the upper side of a ganglion of the nerve-ring. Its walls are formed of a homogeneous membrane lined
with a pavement epithelium, and it is filled with what appears to be a watery fluid, into which projects
from the upper free wall of the capsule a clear spherical body, attached to the wall by means of a broad
short stalli, and containing within it one or more small, concentrically laminated concretions.

Withiu the capsule, on its base, there is seated a cushion-like body, apparently in immediate con-

SENSATION. 143

Tlie lithocysts are almost always freely exposed to the surrounding water upon the margin
of tlie umbrella. In Glotssocodon and Car marina, however, they are entirely ineluded within the
gelatinous substanee of the umbrella. They are usually sessile, but are occasionally {Ciinina)
borne upon a distinct peduncle ; they generally occur in the interval between two marginal
tentacles. In Obclia, however, they arc situated on the inner side of the base of a
tentacle.'

The distribution of the litliocysts on the umbrella margin is at tirst always synuuetrical, and
their number is equal to or a multiple of that of the radiating canals. It is only in certain cases
in which the lithocysts become very numerous, as, for example, in Tima, that this lunnerical law
ceases to be apparent.

We know as little of the function of the lithocyst as we do of that of tiie ocellus. Consisting
as it does of a capsule and contained concretions, it presents a structure which has been
compared with that of certain organs which are met with in the moUusca and in the Crustacea,
and which have in both these groups been generally regarded as organs of hearing, and a similar
function has from analogy been attributed to the lithocyst. The analogy, however, is by no
means so close as to justify us in attributing an auditory function to the organs in question,
while their structure would seem to indicate a relation to light at least as intimate as to
sound.-

nection with the ganglion of the nerve-ring. It is composed of roundish and fusiform cells, and
Haeckel regards it as a nervous ganglion. On each side it is prolonged into a flat, longitudinally
striated, band-like cord, which is evidently composed of close parallel fibres, and which he considers
to be a nerve of sense. These two sense-nerves, after passing up like meridian lines on the walls of the
capsule, unite at the opposite pole. Here the fine fibres which compose them appear to become inter-
woven into a cord, which, after forming the stalk of the spherical l)ody which immediately encloses the
concretions, passes into its interior. Haeckel regards this body as a second internal ganglion. It
appears to be a round vesicle, containing within it a mass of small, closely aggregated cells. In the
midst of these cells are the concretions, consisting of a calcified organic basis, the organic matter being
apparently united with phosphate of lime. He is unable to satisfy himself as to the mode in which
the fibres of the sense-nerves terminate within tlie spherical body, though lie believes it probable that
their extremities are in connexion with the cells existing in the interior of this body.

Haeckel's interpretation of the various structures which he thus finds in the interior of the litho-
cyst, and regards as ganglia and nerve-cords, may be open to criticism, though it must be admitted
that it is not easy to offer any very satisfactory view of them. The two supposed sense-nerves which
he has observed running in two opposite meridians on the inner side of the wall of the capsule will
suggest a comparison with the meridional striae described above, as apparent on the central pulp of the
lithocyst in Campanularia.

^ In Oceania octona and 0. iurrita, Forbes (' INIonograph') describes a cavity at the base of each
tentacle, just below the ocellus, and having within it a vibrating mass. He regards this cavity as a
lithocyst, with its contained concretion. A recent opportunity of examining the O. iurrita of Forbes
has convinced me that what Forbes took for a lithocyst is only a dilatation of the gastrovascular system,
with its contents set in motion by the vibratile cilia of the walls, a suspicion which had already been
entertained by Busk and by Gegenbaur.

" The visual functions of the lithocyst, though maintained by Agassiz, Fr. !Miiller, and partly also
by Haeckel, who is inclined to attribute to them a double function of sight and hearing, has been by no
one so well supported as by Busk (" On the Anatomy of a Species of Thaumantias," ' Trans. Jlic. Soc.

144 PHYSIOLOGY.

The two forms of sense-l)0(lies now described are never associated in the same species, and
the case of Tyaropais, as already shown, affords no exception. It has been already stated, as a
nearly universal rule, that the ocelli are confined to the true sexual medusa or gonocheme while
the lithocysts are found only in the blastochenie. Exceptions to this rule, however, are afforded by
Tliauinanlim, as limited by Gegenbaur, which, though a blastoclieme, has ocelli instead of litho-
cysts, and by Staurophora and LaoiUcea, which are also blastochemes, though, according to
Agassiz, the ocellus takes the place in them of the lithocyst. On the other hand, Goodsirea,
Wright, has, according to Wright's observation,^ lithocysts instead of ocelli, though it is a true
gonocheme. MeUcertiim, another blastochenie, is, according to Agassiz, also very exceptional,
for neither lithocysts nor ocelli are found in it.

The ocellus, however, is by no means so constantly present in the gonocheme as the lithocyst
is in the blastocheme, many planoblasts, with the other characters of the gonocheme, having no
definite pigment spot, though the bulbous expansion at the base of the tentacles, with its
contained coloured matter, has in such cases been frequently confounded with a true ocellus.

Touch. — Analogy would lead us to regard it as probable that sensitiveness to touch resides
in all parts of the surface of the Hydroida — such parts, at least, as are not under cover of a
thick perisarc. It would seem, however, that this sensitiveness is in a special degree conferred
upon the tentacles of both hydranth and medusa. The slightest contact of the tentacles of
the hydranth with a foreign body not destined to be appropriated as food, will frequently be
followed by the instant contraction of the entire hydranth, accompanied, in the calyptoblastic
hydroids, by retraction within the hydrotheca, an action which is plainly subservient to the
safety of the animal by causing its withdrawal from sources of injury. The ordinary contact,
however, of pabulum with the tentacle, excites only the prehensile functions of one or more
tentacles, and is not accompanied by any general contraction of the hydranth. In like manner
the marginal tentacles of the medusae instantly contract on being touched. We are not, however,
obliged to conclude that these acts are necessarily accompanied by consciousness on the part of
the hydroid. Like many other apparently conscious acts in the animal economy, they may be
purely reflex and automatic.

Corda and Wright have described as organs of touch the minute bristle-like bodies to which
the latter has given the name of palpocils (see above, p. 107). Haeckel" has also described as organs of
touch certain long fine and stiff hairs (" Tastborsten"), which he has observed in Cunina, and which
Gegenbaur^ and Keferstein and Ehlers* have observed in OSginela ,- they are developed from the
epithelium lying over the ganglion-like swelling on which the lithocyst is seated in these medusae ;

Loud.,' vol. ill, p. 22), who adduces in favour of this view the action of the lithocyst on polarized light,
when the concretion exhibits a well-defined black cross, indicating, as lie believes, a gradually decreas-
ing density from the centre to the circumference, and pointing to a close resemblance between these
bodies and the crystalline lens of higher animals. Evidence of the sensibility of the hydroid plano-
blasts to light is found in their habit of congregating on the light side of the jar in which they may be
confined. I have noticed, on the other hand, that the planulse of Plunmtaria pinnata, in which, how-
ever, nothing like sense organs exist, avoid the light, always crowding towards the dark side of the jar.

^ 'Edinh. New Phil. Journ./ July, 1859.

" ' Die Familie der Riisselquallen,' p. 137.

' ' Versuch eines Systenies der Medusen,' p. 266.

* ' Zoologische Beitrage,' p. 94.

PliOSPlIORKSCENCE. 145

Avhilc riackel lias foiiiul similar hairs arranged in three verticils upon tlie knob-like extremity of
tile tentacle in another medusa, Tlliopalonema imbUicafiim. It is probal)le that in attributing to
these various structures a specially developed sense of louch, tlieir true function has been assigned
to them ; but it must, at tlu^ same time, be admitted that we have no direct evidence of their real
purport in the economy of the animal.

(). PhospltoreHCcnce.

Among the most reniarkal)le faculties possessed by the Hyduoida is the power with which
many of them are endowed of emitting light — a power which, like contractility and sensation,
ought, perhaps, to find its place among the functions of irritability, as it appears to be always
manifested in obedience to the action of a stimulus.

If a healthy colony of Oldia dicliotoma, for example, be irritated in the dark by being
roughly touched, a beautiful pale-white light will be seen for an instant to flash along the branches.
Here the power of emitting light would seem to reside exclusively in the trophosome, and I have
sought for it in vain in the free planoblasts of this hydroid, where one might, a jjriori, expect to
find it.

On the other hand, in species of Thanmuntias and allied forms which I have met with
swimming in the open sea, and which are almost certainly planoblasts liberated from some tropho-
some as yet unknowai, the luminosity was very striking and beautiful. In these medusae, when
an individual confined in one of my jars was touched in the dark, the whole umbrella-margin
became instantly lighted up by a multitude of luminous points. The luminosity was entirely con-
fined to the margin of the umbrella, and, indeed, might be seen to have its exclusive seat in the
bulbous bases of the marginal tentacles.^ These phosphorescent hydroid planoblasts must be
regarded as one of the chief sources of the luminosity of the sea.

In some of the most brilliantly phosphorescent hydroid trophosomes, the luminosity presents
a singular intermittence, the light appearing, when the hydroid is touched, to palpitate in a very
beautiful way over the surface. Its duration, however, is very transitory; within a few seconds
it will have entirely vanished, and will then need a repetition of the stimulus to call it forth.
After a few such repetitions the power seems to exhaust itself, and rest for some time is needed
before it can be again excited.

The vapour of alcohol exerts a very marked influence on the emission of the light. On
exposing a campannlarian trophosome to alcoholic vapour, given off at a temperature of about
70° F., I was surprised to find that not only was a brilliant luminosity called forth, but that the
light had accpiired a persistence very different from its usual transitory manifestation. It con-
tinued, indeed, for about five minutes while held over the vapour, after which, though still exposed
to the vapour, it gradually faded away.

It must be here borne in mind, however, that the action of the stimulus in the experiment

' Busk ('Trans. ]\Iic. Soc. Lond.,' vol. iii, p. 22) has described the luminosity of a Thaumantias-
like medusa, in which lie also fouud the seat of the light to be confined to the raaigiual tentacular
bulbs.

19

146 PHYSIOLOGY.

witli the alcohol is continuous, while the mechanical stimulus exerts itself only for a moment. A
stimulus presenting a closer parallelism with that of the alcohol is afforded by contact with
atmospheric air. I have always found that the hydroid, when sudilenly removed from the water
into the air, becomes brilliantly luminous at the moment of the change of medium ; the luminosity
here, however, as in the case of simple mechanical stimulus, lasts only for a few seconds, notwith-
standing the continuous action of the atmosphere on the animal. After its disappearance the
vapour of alcohol will again call it forth, accompanied by the persistence which characterises the
application of this stimulus.

Many hydroids, however, are destitute of phosphorescence. It does not exist in the fresh-
water Hydra, and I have never witnessed it in any gynmoblastic fomi ; while among the
CalijptobJastea it would seem that only some are endowed with it. Observations, however, are
still wanting on this point, and phosphorescence will doubtless yet l)c found in species in which it
is not at present known to exist.

Of the immediate source of the phosphorescence we know scarcely anything, and though in
some other phosphorescent animals it would seem to reside in a special luminous secretion we
have no evidence of such a secretion in the Hydroida. Its dependence here on the operation
of a stimulus would remove it from simple physical luminosity, such as may result from a process
of slow combustion or from insolation. It may, it is true, be asserted that the immediate result
of the stimulus is to excite the formation of a luminous secretion, but in the absence of all evidence
of any such secretion this explanation cannot be accepted. The luminosity would here seem
rather to be, like electricity in other cases, the direct accompaniment of certain vital actions.'

7. Reproduction.

The reproductive faculty, whose exercise gives rise to the calling into existence of a new
being, belongs properly to the department of Physiology, while development, or the successive
changes of form which this being undergoes, is a subject of Morphology, and has been already
treated of in its proper place under that head.

Reproduction in the livDROiDA may be either sexual or non-sexual; the sexual showing
itself in the production and fertilization of the ovum, the non-sexual in the production of buds
and in fission.

' Some observatious which I have made on the lumiuosity of Beroe have brought out the some-
what unexpected result that the faculty of emitting light is not possessed by these animals at all
times during the twenty-four hours. It does not exist in the presence of daylight, and a previous
seclusion of the animal for some time in darlcness is always necessary for its manifestation (see ' Proc.
Roy. Soc. Edinb.,' 1863, p. 519). I have no reason to believe that this is the case with the Hydroida,
though I have made no observations which can here be regarded as decisive. It certainly is not the
case with NoctUuca, one of the most vividly luminous of animals, and one well fitted for observation.

REPRODUCTION. 147

a. Sexual Reprodudion — Generation.

Comparison of the Sexes in the Hi/droida. — The existence of differentiated sex in the
Hydroida was first announced by Ehrenberg,' who cnaintained that the so-called "tgg-capsules"
in Cori/ne, Sertularia, &c., had the significance of special fertile animals, to which he gave the
name of females, while he regarded the ordinary hydranths as the sterile individuals of the
colony.

With this announcement we may date a well-marked era in the history of progressive
discovery among the Hydroida ; for it is to the happy conception of Ehrenberg that we must
refer the more jihilosophic views which within the last few years have so greatly advanced our
knowledge of the structure, functions, and relations of these animals.

The celebrated German micrologist, however, did not grasp the full meaning of the facts of
which he had thus so nearly given us the exact interpretation ; for he regarded the central
column (blastostyle) of the gonanginm in Sertularia as the equivalent of the central diverticulum
(spadix) in the gonophore of Coryne, while he viewed the gonophores borne on the sides of the
blastostyle in Sertularia as merely eggs equivalent to the true eggs contained in the gonophore of
Coryne.

The doctrine of the sexual differentiation of the Hydroida was confirmed by Loven in a
remarkable memoir, originally published in the ' Transactions of the Royal Swedish Academy' for
1835, and thence translated into Wiegmann's 'Archiv.'- In this memoir Loven gives an account
of those singular extracapsular medusiform gonophores which are described above (p. 57)
under the name of "meconidia;" he found them in a Campanularian hydroid (Gonotl/yrea
Loveni, Allm.), and recognised in them their true sexual function. He also describes the occur-
rence of medusi-form gonophores in two species of Syncoryne ; and having observed that in
the gonophores of one of these species the cavity of the umbrella was filled with ova, he dis-
tinguishes them from mere organs, and regards the gonophores in both instances as special female
animals.'

Naturalists had now not only become familiar with the presence of true ova in the
Hydroida, but they saw in the portions of the colony set aside for their production something
more than mere organs. No one, however, had as yet discovered any trace of spermatozoa ;
Ehrenberg at this time makes no mention of a male element, while Loven calls the nutritive poly-
pites male, and in this view of their nature falls behind Ehrenberg, who more truly names them
sterile or sexless individuals.

' ' Corallentliiere, Abhandl. der Kiinigl. Akad. der Wiss. zu Berlin,' 1832.

^ ' Beitrage zur Kentnniss der Gattungen Campanularia und Syncoryne, Wiegm. Arch.,' 1837.
Erster Band, S. 239.

" It may here be noticed that AV'agner had already (Isis, 1833, § 256, tab. .xi) found medusa-like
gonophores, filled with ova iu a hydioid which be names Coryne aculeata, apparently a species of
Podocoryne ; but, not being aware of tlie doctrine of Ehrenberg only just announced, tiie exact
significance of these bodies escaped him.

149 PHYSIOLOGY

The doctrine of the sexuality of tlie Hydroida now waited only for the discovery of tlic male
element in order to receive its complete development. This discovery was made by Ehrcnberg,
who, in 183S, pointed out the real nature of certain conical tubercles which at particular seasons
are developed on the body of the freshwater hydra, and had been by previous observers regarded
as a peculiar disease to which this animal was supposed to be subject, but which were now shown
by Ehrenberg^ to be true spermatophorous capsules, while a further and imi)ortant step in this
direction was made by Krohn, who a few years afterwards announced that he had, in the
Peniiaria Cavolinii, Ehren., found certain receptacles similar in form to the ovigerous ones long,
ago described by Cavolini in the same remarkable hydroid, but containing spermatozoa instead of
ova. Similar observations were made on Tahidaria indivisa and on Eudeiidriuni racemosum, as
well as on Aglaophenia pluma and the Serttdaria [Eudendriiim ?) missenensis of Cavolini, in all
of which Krohn succeeded in detecting spermatozoa."

It is now certain that every species of hydroid gives origin to male and female zooids (or, in
case of such medusa:' as may be directly developed from the egg, to male and female se.xually
generated individuals), one destined for the production of ova, the other for that of spermatozoa.
The separation of the sexes in distinct generative zooids, or in distinct individuals of a sexually
generated offspring, is thus absolute and universal among the Hydroida. In by far the greater
number of cases the separation is carried even further than this ; for we scarcely ever meet with
male and female gonophores in the same colony. As an almost universal rule, then, the
Hydroida are dioecious ; in other words, every colony is unisexual.^

Some few cases of a monoecious condition, however, occur. This has been noticed by many
observers in the freshwater Hi/dree,"^ where, indeed, it is the most usual condition. I have found
it also in Plumularia pinnata, which sometimes carries on the same stem both male and female
gonophores, and Hincks has observed it in some other sertularian hydroids.** In Bicoryne
conferfa too there may generally be found, among the dense forest of stems with which this
hydroid invests the surface of univalve shells, some stems carrying male and others female
gonophores. Each stem, howevei-, carries gonophores of one sex only, though it would seem that
both male and female stems are united by the creeping stolon into a common colony. In
Hydradinia, on the other hand, whose habit is entirely similar to that of Bicoryne, we
never meet with the two sexes in a common colony ; perhaps even never investing the same
shell.

Orlyiii of the Generative Elements. — Throughout the whole of the Hydroida the generative
elements originate between the endoderm and ectoderm, and, with one exceptional condition to
be presently described, are always formed in the walls of an organ strictly homologous with the
manubrium of a gynuioplithalmic medusa.

' ' ^Nlittlieil. aus (!eii Verliandl. der Gcsellsch. iiaturf. Frcuudc in Berlii),' 1838.

" Krohn, " Eiiiiiie Bemerkungen und Beobaclitungen ubci- die Gesclileclitverlialtuisse bei den
Sertulariiien," 'Miillei's Archiv,' Jahrg. 1843, S. 174.

■' Ki'ohn had already noticed lliat, in all the species examined by liim, the tnale and female gono-
phores were borne on separate colonies (loc. cit., p. 181).

* See especially Prof. Allen Thomson " On the Coexistence of Ovigerous Capsules and Sperma-
tozoa in the same Individuals oi Hydra viridis," in ' Proc. Roy. Soc. Edin.,' No. 30, 1845-47.

" ' Quarterly .Juunial of Science,' July, 18G5, p. 409, note.

REPRODUCTION.

149

Fig. 00.

This organ forms the axile diverticulum in the young adclocodonic gonophorc, and the
manubrium of the sexual medusa, while it is represented by the entire sexual zooid which buds
from the radiating canals in the blastocheme or non-sexixal medusa.

It is not at first easy to say whether the generative elements have their proper origin in the
ectoderm or endoderm of this body, as in most cases they can be merely seen filling the
space between these two membranes, which become more and more separated from one another
iis the included mass of ova or spermatozoa increases in volume.

From some favorable observations, however, which I have succeeded in making on certain
species of hydroids, I have convinced myself that the true origin of the ova and spermatozoa is
to be found in the endoderm, while the ectoderm serves merely as a confining and protecting sac
until such time as the generative elements acquire sufficient maturity to allow of their liberation,
which always takes place by simple rupture or absorption of the ectodermal sac.

Thus, in the gonophores of the male colonies of Sertularia polyzoiiias the spermatogenous
tissue may be seen filling the entire space between the long cylindrical axile spadix and the
surrounding walls of tiie gonophore. In most specimens it may be easily seen that the sperma-
togenous mass is far from being of uniform maturity throughout ; for while towards the axis of
the gonophore it is still very inuuature, the mother-cells being here distinctly visible with the
ultimate spermatic cells within them, we find that towards the periphery it consists of free active
spermatozoa. The youngest portion of the mass is thus that which is still in contact with the
spadix or endodermal portion of the gonophore, while the oldest portion is situated externally,
being in contact with the confining ectoderm — a condition
which would be scarcely possible if the ectoderm, rather
than the endoderm, gave origin to the spermatic cells.

A state of things exactly parallel to this may be seen
in the female gonophores of Cori/ne pmiUa, in which,
moreover, the actual formation of the ova may be
satisfactorily traced. At an early period in the deve-
lopment of these gonophores, the large thick spadix
may be seen to be surrounded by a granular plasma,
throughout which numerous minute nucleated cells are
scattered (woodcut, fig. GO). These cells I regard as
the germinal vesicles and spots of the future ova, round
which no distinctly differentiated vitellus can as yet be de-
tected. In a more mature stage of the gonophore, while
the same peculiar tissue continues to invest the spadix,
the peripheral portion of this tissue may be seen to be
thrown off in the form of undoubted ova, consisting each of
a germinal vesicle and spot precisely similar to those ob-
served in the more central portion of the mass, but now
with a portion of the common plasma differentiated round
each germinal spot in the form of a a very definite vitellus.
AVhen the gonophore has attained complete maturity, the

whole of the jjlasmatic mass, with its immersed nucleated cells, has become metamorphosed
into fullv formed ova.

Young Spoi-osac of Coryn

, show

certidn early stages in the foruiatiou of the

a, Outer wall f perigonium) of the sporosac ;
b, cavity of spadix ; c, plasma, investing the
spadix, and having imbedded iu it the germinal
vesicles of the future ova; within the germinal
vesicles are seen the germinal spot and the
puttctum gprmanitivnm ; d, fully formed ova,
in each of which a portion of the common
plasma has become tlidercntiated as a vitellus
round the germinal vesicle.

' '>0 PHYSIOLOGY.

I have spoken above of an exception to the all hut universal fact that the generative
elements originate between the ectoderm and endoderm of a body homologous with the manubrium
of a naked-eyed medusa. The exception referred to consists in the origination of ova in the blas-
tostyle, as may be seen in Sertidaria pimila and one or two other species of Sertularia.

In Sertularia pumila a solitary gonophore of the ordinary form, and containing in the usual
way ova or spermatozoa, originates, as in other cases, by a bud from a blastostyle. In the female
colonies, however, nucleated spherical bodies, in no way distinguishable from young ova, are
found in the walls of the blastostyle itself, between whose ectoderm and endoderm they seem to
lie (woodcut, fig. 21, k). I have not succeeded in satisfactorily tracing the destination of these
bodies ; but I have reason to believe that the true gonophores bud forth from that part of the
lilastostyle in which the nucleated bodies occur, and that these, as young ova, pass from the
blastostyle into the budding gonophore, where they would then naturally occupy their normal
position between the endoderm and ectoderm of an organ representing the manubrium of a
medusa, destined to undergo there a further development before being discharged into the acro-
cyst, which, as we have already seen, exists in this species. Each gonophore, after having per-
formed its duty as a receptacle, in which certain intermediate stages of development take place,
would seem to disappear, and be succeeded by another, which in a similar way receives its young
ova from the blastostyle on which it buds.'

b. Non-sewiial Reproduction.

Gemmation. — As already saiil, non-sexual or agamic reproduction may manifest itself in the
Hydroida either by budding or by fission. There is scarcely any part of the external surface of a
hydroid which may not give origin to a bud, though the actual conditions which determine the
formation of this bud are entirely uidcnown. Buds capable of becoming developed into one or
other of the various forms of zooids, which make up the hydroid colony, may be emitted by the
hydranth, by the hydrocaulus, or by the hydrorhiza. In the gonosome we find that not only does
the blastostyle give origin to buds destined to take part in the generative functions, but that the
medusae themselves have the power of emitting buds from various parts of their surface. The
form and development of these buds have already been considered in the morphological section of
the present Monograph.

As an almost universal fact the bud, from whatever part of the hydroid it is emitted, has its
somatic cavity in open communication with that of the budder, so that the common somatic fluid
passes freely from the one into the other. Cases, however, have been recorded (see above, p. 82)
in which certain .^ginidan medusae would seem to give origin to buds from the internal surface
of the manubrium. It is possible that there may be here some error of observation, and, though

' Bodies, undoubtedly of the same nature as those here described, but without any indication
of a nucleus, are figured by Agassiz in an American species, wliich he regards as identical with the
Sertularia pumila of Europe (op. cit., pi. xxxii, fig. 9). Tliey had also been already described by
Lindstrora (op. cit.).

REPRODUCTION. 151

we owe the statement to able and trustworthy inquirers, it is yet to be desired tliat we had further
verification of a fact so much at variance with the phenomena of gemmation as presented elsewhere
among the Hydroida.

It is rarely that the medusa has been noticed to emit buds simultaneously with the produc-
tion of ova or spermatozoa. Instances, however, are on record in which the sexually mature
medusa has also multiplied itself by budding. This has been observed by Rusch' in a medusa
which he refers to the Sarsia prolifera of Forbes, and in which the basal I)ulbs of the tentacles
gave origin to medusa-buds, which were coexistent with the presence of generative elements in
the walls of the manubrium ; by Krohn" in the medusa of Clavalella, which he has seen to be
loaded with ova at the same time that medusa-buds were emitted from the margin of its umbrella;
and by Sars,' who in a blastocheme {Thaumantias multicirratus, Sars) saw medusae budding from
the radiating canals simultaneously with the existence of the convoluted generative pouches.

No multiplication by budding has ever been noticed in the sporosac, a zooid which, it is to
be borne in mind, is almost from its first appearance engaged in the production or protection of
the generative elements.

Fission. — Though budding thus constitutes a highly characteristic and all but universal phe-
nomenon among the Hydroida, multiplication by spontaneous fission is, on the other hand, rare
and exceptional. Kolliker* observed a process of true fissiparous multiplication in a medusa {Stomo-
hracMum mirahile, Koll.) obtained in abundance at Messina. The fission always connuenced by a
vertical division of the manubrium, which thus became doubled ; and this stage of the process was
followed by a similar division of the umbrella, separating the animal into two independent halves.
The process, however, did not stop here, but was followed by a further division of each of the
two first-formed segments into two others, by a fission at right angles to the direction of the first ;
while Kolliker's observations led him still further to conclude that the process does not terminate
with even the second cleavage, but, on the contrary, that it still goes on, the animal continuing
to multiply itself by frequent acts of fission.

Developed generative bodies were not observed in Stomohrachium mirahile, and Kolliker is
of opinion that this medusa is only the young of another {Mesonema carulcscens, Koll.) found in
the same seas, and in which no division takes place, but in which well-developed generative sacs
occur along the course of the radiating canals.

But besides this case of fissiparous multiplication in the medusa 1 am enabled to give a
very well-marked and interesting one which I met with in the trophosome of an undescribed
campanularian hydroid (woodcut, fig. Gl), to which I have assigned the name Schizocladiinu
ramosum."

I have not as yet met with this hydroid more than once. It is a profusely branched form,
with its trophosome having much resemblance to that of Obelia dickotoma ; but as no gonosome

' Busch, ' Beobaclit. iiber Anat. u. Entwick. einiger ■wirbellosen Seethicrc,' p. 1, pi. i, fig. 1.

' Krohn, ' Wiegmann's Arcliiv,' 1861. In the gemmiferous specimens of the Clavalella medusa
examined by myself, there were no visible generative elements.

' Sars, ' Bescrivelser.'

* ' Zoologische Beitrage, 1861.'

° "On a Mode of Reproduction by Spontaneous Fission in the Hydroida." ' Reports of Brit.

Assoc, for the Advancement of Science,' 1870.

152

PHYSIOLOGY.

was present in any of the sj)ecinicns collected, its exact systematic position cannot now be more
than provisionally assigned to it.'

Besides the ramuli which here, as in the hydroids generally, support the hydranths, others
are developed in abundance from all parts of the hydrocaulus. These (a, a) commence just like

Fig. 01.

Schizocladimn ramosum, showiug reproduction by spontaneous fission.

A. Part of an adult colony magnified about six diameters, rz, One of the fissiparous ramuli, still entirely
invested by tbe chitinous perisare ; b, a fissiparous ramulus, in \\ hicb the contained L-ceuosarc has extended
itself beyond the distal extremity through the ruptured perisare ; a constriction (the commencement of
fission) has begun to show itself in the ccenosarc of the ramulus where still covered by the peri:*arc ; c, the
fission is completed, and the separated portion is escaping from the distal extremity of therauniUis ; rf, the
separated portion has entirely disengaged itself, and has become a free frustule in the surrounding water.

B. Gemmation of the bydroid from tbe free fission frustule. _/", The free frustule, after having excreted
a mucous tube, from which it has partly withdrawn itself; g, a bud has begun to be emitted from the side
of the frustule ; e, the bud has become developed into a hydJrantb with its hydrotbeca and hydrocaulus, and
the young trophosome has bcguu to complicate itself by the emission of a branch.

the ordinary ramuli as o£Fslioots from the hydrocaulus, and consist, as usual, of a continuation of
the ccenosarc invested by a chitinous perisare. Unlike the ordinary branchlets, however, they
never carry a hydranth, but are destined for the multiplication of the colony by a process of
spontaneous fission.

After the entire ramulus has attained some length, the contained ccenosarc continues to

It is quite possible that ia Schizocladhim ramosum spontaneous fission never occurs simulta-
neously with true sexual geueratiou.

REPRODUCTION. 153

elongate itself. In doing so it ruptures the delicate pellicle of chitine which closes the extremity
of the ramuhis, and extends itself quite naked into the surrounding water.

It is now that the process of fission commences. A constriction takes place in the coenosarc
at some distance below its distal extremity, and in the part still covered by the chitinous peri-
sarc [b). The constriction rapidly deepens, and ultimately cuts off a piece (c), which slips
entirely out of the perisarcal tube and becomes a free zooid {d), while the surface of disseveration
soon heals over, and the axial cavity of the free frustule becomes here as completely closed as at
the opposite end.

The detached segment is now about the yj]^ of an inch in length, and strikingly resembles
a planula in all points except in the total absence of vibratile cilia. It attaches itself by a mucous
excretion from its surface to the walls of the vessel, and exhibits slight and very sluggish changes
of form. It now slightly advances along the surface of support, withdrawing itself from the first-
formed portion of the excretion, which remains behind as a tube of great tenuity, adhering to the
sides of the vessel (/).

In tracing the further history of the frustule it was foiuid that this never directly developes
a mouth or becomes transformed into a hydranth. After a time a bud springs from its side (y),
and it is from this bud alone that the first hydranth of the new colony is developed.

The bud which thus becomes developed into the primordial hydranth remains attached to
the fission-frustule, which forms for it a sort of hydrorhiza, but which would seem ultimately to
perish and give place to true hydrorhizal filaments. In the mean time the primary bud emits
others (e), and a complex branching colony is the result.

The fission-frustule thus admits of a comparison with the free medusiform element of other
hydroids, with which it agrees in never becoming directly developed into a hydriform trophosome,
but from which it differs in the very important fact of taking no part in the true generation of
the hydroid, and in giving origin to a new colony only by a non-sexual multiplication.

The fissiparous multiplication of Sckizodadium Avould seem to throw light on the nature
of certain bodies which made their appearance in a jar containing living specimens of Cori/morpJia
nutans (see Plate XIX, figs. 12 — 14). These bodies presented a close resemblance to the fission-
frustule of ScMzocladlum , and were seen to become developed into hydranths, which it is almost
certain ultimately repeat the form of the adult Corymorpha. Their origin was, at the time I
noticed them, very enigmatical, but I now regard it as highly probable that they are produced
by a process of spontaneous fission from the filaments which are emitted towards the base of the
stem in the Corymorpha. They would seem, however, to differ from the fission-frustules of
Scltizocladium in becoming directly developed into a trophosome.

The decapitation and successive renewal of the hydranths, referred to above (p. 69) as
occurring in various species of Tuhalaria, may be compared with the phenomenon of fissiparous
multiplication just described. In the decapitation of Tuhularla, however, the separated hydranth
is not destined to undergo any further development ; it has matured its sexual buds, and has
accomplished all the objects of its existence before being cast off, and it then perishes, to be
replaced by another.

The decapitation of Tiihiduria admits of a still closer comparison with the formation and
detachment of discs {cphyrai) from the hydriform stage [scypliostoma) of Aurelia and other
BlscopJiora. Here, however, the discs into which the scypliostoma breaks up by a process of
transverse division which has its equivalent in the budding, by which the sexual zooids are formed

20

154 REPRODUCTION.

from the hydriform trophosome of the Hydroida, are destined to undergo further development
and enjoy an independent existence like that of the hydroid planoblasts. But notwithstanding
this difference the analogy is still close and interesting ; for the more simultaneous occurrence of
the transverse constrictions which result in the formation of a pile of discs before their ultimate
detachment in the scyphodoma has but little significance ; while, as we have already seen, the
hydranths successively produced and detached from the stem \\\ Tubularia, are formed not by a
process of budding, but by one of metamorphosis which shows itself in growth with change of
form in the distal extremity of the stem in this hydroid, — a mode very similar to that by which
the successive terminal discs or rpJiyra are developed from the scypliosioma.

DISTRIBUTION IN SPACE. 155

DISTRIBUTION.

1. DlSTRIBUTlOX IN SPACE.

The data from which an adequate knowledge of tlie Geographical Distribution of the
IIydroida may be derived arc as yet very imperfect, and our assertions regarding the horizontal
and vertical extension of the group in the present seas of the globe cannot be regarded as
possessing more than a provisional value, liable to modifications as additional facts come to our
knowledge ; while — if we except the graptolites, whose hydroid relations will be presently dis-
cussed— we know almost nothing of the geographical distribution of hydroids in former periods
of the earth's history. Indeed, the localities which have yielded the very few fossil hydroids —
not being graptolites — hitherto discovered cannot be viewed as representing, even approximately,
the former distribution in space of the species which have thus come down to us, and the following
remarks on the distribution of the IIydroida in space are therefore confined to the existing forms
of the order.

o. Horizontal Distribution.

In a complete exposition of the horizontal distribution of the IIydroida our statements
ought to embrace not only the fixed elements of the hydrosoma, but also tlie free planoblasts. It
is true that a large number of planoblasts still remain untraced to tlieir hydriforni trophosomes,
but our want of knowledge in this respect can scarcely afford grounds for rejecting the medusa
from our general survey of hydroid distribution, more especially when we bear in mind the
probability that the planoblasts, notwithstanding the pelagic habits of these free-swimming buds,
never wander far from the rooted trophosomes in which they originate ; and it is only because
our positive knowledge of the distribution of the planoblasts is so very deficient that our assertions
regarding hydroid distribution must be understood as applying chiefly to the rooted trophosome.

But even though we take all the facts which have come to our knowledge regarding the dis-
tribution of both trophosome and planoblast, we are still far from possessing the data necessary for
a satisfactory determination of the geographical distribution of the Hydroida. In a great many
important geographical areas we know absolutely nothing of the state of hydroid life, while from
many others the facts which have come to us are so few that they are quite inadequate as the
basis of a generalisation.

It is upon the coasts of the British Isles that the most numerous and complete observations
have been made, and next to these in importance must be mentioned the north-west shores of
the continent of Europe, the Atlantic shores of North America, and the ]\Iediterranean Sea ; while
others of considerable value have come to us from the Pacific shores of North America, and from
Australia, New Zealand, and South Africa.

Several hydroid trophosomes have been recently obtained from deep dredgings in the
western parts of the Gulf Stream, while dredgings at great depths in the more eastern areas

21

156 DISTRIBUTION.

of the North Atlnntic have yielded others of much interest ; and others still have Ijeen found
tenanting the great Nortli Atlantic seaweed meadow, where, borne upon the floating weeds of the
Sorgasso Sea, they lead alnmst the pelagic life of the planoblast.^

Some scanty observations have come to us from the Pacific and Atlantic shores of South
America, from the Falkland Isles, from Madeira, from the Islands of the East Indian Archipelago,
and from Kerguellan's Land, and the Auckland Islands ; while we have somewhat fuller ones
from the West Indies. From all these localities we have evidence of the occurrence of hydroid
life, though from none have we data sufficient for the determination of definite hydroid provinces.
We have hardly yet sufficient information to enable us to assign any special physiognomical
facies to the hydroid faunas of special zones, though the beautiful plumnlarian group represented
by the genus JgJaophenia may be considered as especially characteristic of intertropical and
warmer temperate seas, where it acquires a far greater development than in the colder w^aters of
the higher latitudes ; and it may perhaps also be asserted that the largest hydroid forms are as a
rule confined to the w'armer seas, while those of temperate and colder latitudes consist for the
most part of hvmibler and less conspicuous species.

Under the name of Phmidaria angiilosa Dana" mentions a hydroid from the East Indies,
which attains a height of three feet, while Semper' has described a gigantic plumularian hydroid
from the Pelew Islands, where it forms submarine forests nearly equalling in height that of a man,
and with the base of the stems more than an inch in thickness. It is armed with formidable
stinging cells, and the incautious bather will have reason to repent his too rash incursions
within the precincts of this marvellous grove when he finds himself suffering for hours afterwards
a sense of intolerable burning excited by the envenomed darts to which he has unwittingly
exposed himself.

Nothing which can be compared with this has been found in any of our northern or
temperate seas, where, however, Ac/laophenia mi/riophi/lhim may attain a height of between two and
three feet, and where Tubiilaria and Corijmorpha present forms which strike us by the large
size and conspicuous beauty of their hydranths.

The genera of the Hydroida would appear to be far less rigidly confined within limited
areas of distribution than we know to be the case with those of other nearly allied groups, such
as the coral-forming Advnozoa.

In a collection from New Zealand, which was placed in my hands for determination by Mr.
Busk, and which consists of twenty-five different species of calyptoblastic hydroids referable to
seven genera, I cannot find more than two generic forms (both new) which are not also British.
Among thirty-one species from Australia, collected during the voyage of the " Rattlesnake,"
referable to seven genera, and determined by Busk,* five of these genera are, as pointed out by
Busk, represented by British species, while of the remaining two {Paxythea, Lamx., and Idia,
Lamx.), Pmijthca, though not British, is according to Lamouroux, a form belonging to the

^ Suice the above was writteu the Hydroida obtiiined during tlie e,\ploration of the Gulf Stream
by tlic United States Coast Survey have been entrusted to me for determination. So far as I have as
vet been able to e.xamine tliem I find tliem full of interest. Tlie collection is a very large one, and
gives promise of a most important addition to our knowledge of hydroid zoology.

' Dana, " Structure and Classification of Zoophytes.''

' Semper, 'Zcitschrift f. w. Zool.,' Bd. xiii, p. 560.

* Busk, in the Appendix to the ' Voyage of the Battlcsuake,' p. 385.

DISTRIBUTION IN SPACE. 157

Northern Hemisphere, liaving been met with in the Nortli Atlantic, tliongh its exact locahty is
doubtful, as Lamouroux's specimens were found on floating seaweed. To the genera peculiar to
Australia must also be added the genus Lineolaria of Hincks.'

Again, in a collection containing sixteen species of calyptoblastic hydroids from the Pacific
coast of North America, and examined by Busk and myself, besides one undescribed genus, there
is not another in the collection which is not also British ; while twelve species determined bv
Alexander Agassiz,^ Andrew Murray,' and Trask,' from the same shores, belong all to British
genera.

A collection of twenty-five species of Cali/pioblastea from Soutli Africa, given to me by Mr.
Busk for examination, are also all referable to British genera.

While almost all the generic forms known to us as occurring in other latitudes than our own
are thus widely distributed over the globe, this is by no means the case with the species. The
species of hydroids are as a rule confined within determinate and limited areas. Thus of the
twenty-five species composing the New Zealand collection just mentioned, three only, namely,
the cosmopolitan SertulareUa -pol ponias, with its nearly allied Sertidarella Gay'ii, and the equally
cosmopolitan Sertidaria jjimila, have been recorded from European seas ; while among the thirty-
one species collected by the " Rattlesnake" in Australian seas. Busk has detected only three
European species, namely, Sertularia operculata, Lafoea damosa, and Laomedea voliihiUs ; the last
of which he regards as doubtful. Of the remaining species of this collection three only, as pointed
out by Busk, occur beyond the limits even of the Australian seas, namely, Sertularia elongata,
Lamx., which has been also collected in New Zealand, Sertidarella divaricafa, Busk, in the Straits
of Magellan, and A(/laoplienia MacgUUvraiji, Busk, in the Philippine Islands. To these, however,
must be added the Sertidarella hupinosa. Gray, which is also a New Zealand species, while it
has been collected by Dr. Hooker in the Auckland Islands, and an Aglaophenia {Aglaophenia
formosa, Busk), obtained from Australia by Dr. Landsborough, and which I have found in the
collections from New Zealand and from South Africa ; while Mr. Hincks records as Australian
the Plumularia ohllqua of our own seas.

Among the hydroids referred to above from the Pacific coast of North America, there are
only two which can with any probability be regarded as also European. One of these is Sertularia
put/iila, found in California by Dr. Scouler, and contained in ilr. Busk's collection. The specimen
is destitute of gonangia, but its trophosome is indistinguishable from that of our British form.
The other is Lafoea dumom, forming part of the same collection, and also obtained in California.
This is also destitute of gonangia, but so are all the European examples of this species hitherto
examined.

Indeed, the limitation of groups of hydroid species to definite areas is strikingly illustrated
by the marked difference between the hydroid fauna of Australia and that of New Zealand.
Among all the hydroids hitherto examined from these two regions, only three species, Sertularia
elongata, Lan.x., Sertulurella hispiiiosa. Gray, and Aglaophenia formosa, Busk, are common
to both. Perhaps when a greater number of Australian and New Zealand species shall have

' Hincks, in 'Ann. Nat. Hist.,' 1861.

^ Al. Agassiz, ' Illustrated Catal.'

' A. Jlnrray, ' Ann. Nat. Hist.,' vol. t, .3rcl ser., p. 250.

* Trask, ' Proc. Cal. Acad.,' 1857.

158 DISTRIBUTION.

become known, other points of agreement will he fonnd ; bnt it is not likely that the percentage
of species common to the two will be much altered ; and this will appear all the more remarkable
when we bear in mind that a very large proportion of Australian PoJij'zoa are, as ascertained by
Busk, common to that country and New Zealand,

Even within the region of the British seas a few s])ecies occur which are not only
exclusively British, but which are either absolutely confined to the Northern or the Southern shores
of the British Isles, or have at least the metropolis of their distribution on one or other of these
shores. Among these we may cite, as examples of Northern forms, Syncoryne eximia, Syncoryne
pulchella, Etidendriiim unnulatum, Tubularia bellis, Halecium lahrosum, and Sertularia fuscci;
while Coryne vnyinata, Periyonhnns serpens, Hydrnntliea wargarica, OjjJnoides mirabilis, and
Aylaoplienia permatala may be cited as especially belonging to the Southern shores of our islands.
It is not improbable, however, that further observations will show that many species have a much
wider range than what our present knowledge would justify us in attributing to them.

There is, however, one region which affords a remarkable exception to that limitation of
species within definite areas which is here insisted on. I have examined a collection of twenty-
five species of calyptoblastic hydroids from South Africa in Mr. Busk's possession, and find them
not only all referable to British genera, but with no less than ten of the species indistinguishable
from hydroids occurring on the British shores. These ten species are Sertularia pumila, Sertu-
laria abietina, Sertularella polyzonias, Bipliasia afferiuata, Diphasia pinnala, Antenmdaria
antennina, var.,^ JylaopjJienia tubulifera, Filelluni serpens, LaJ'oea pygmea, and Lafom parvula -^
while another British species, Sertularia operculata, has also been recorded by Busk from the
same seas.

Among the ten hydroids, however, which I have thus identified with British species, the
specimens referred to Sertularella polyzonias and Bipliasia jjinn at a are the only ones in which the
gonosome is present ; but as the trophosomes of the others are indistinguishable from those of the
species to which I have referred them, we are justified for the present in assuming the identifica-
tions as correct.

Admitting the correctness of this determination, the proportion of South African species
identical with British ones is quite exceptional, and unexpectedly large ; so much so that I have
little hesitation in explaining the correspondence between the two faunas by referring it to the
transporting agency of the large number of European ships which frequent our South African

* This Antennularia, though branched, appears to come nearer to Antennularia antennina than to
Antenmdaria ramosa. The short intervening iuternode which is characteristic of Antennularia anten-
nina is here occasionally present in the ramuli, though more frequently absent ; while the rarauli them-
selves agree with those of Antennularia anlennina in being more distant tlian in Antennularia ramosa.
The form is plainly intermediate between Antennularia antennina and Antennularia ramosa, and would
seem to shake the validity of the latter as a true species.

^ Though I believe I am right in the determination of these two species of Lafoca, it must not
he forgotten that the smaller species of Lafoea are very obscure, and by no means easily distinguished
from one another. No gonosome having as yet been found in any of them, we are, in the absence of
the hydranths which may possibly afford diagnostic features, forced to characterise them from what are
in many cases very slight differences in the form of the liydrothecre — diffcicuces which have scarcely
the constancy necessary for a specific cliuracter.

DISTRIBUTION IN SPACE. 159

colonies, and to whose bottoms the liydroids may adhere. It is interesting, and^ confirmatory of
this view, to find the specimens of Scrfulnria abiefina infested in precisely the same way as British
ones with SjArorhis cowuiiiiiis and Filellum serpens, both quite chai'actcristic dwellers on the
specimens of Sertidaria ahietina which occurs in such abundance round the British shores.

It is an interesting and significant fact that the distribution of tlie most widely disseminated
species is not to be accounted for by any locomotive faculty or pelagic habits of their sexual
buds, for the most cosmopolitan hydioids never produce planoblasts at all, their se.\ual l)U(Is
being in the form of fixed sporosacs.

Indeed the hydroid medusit- are nnicli more local in their distribution than what one might
at first be led to suspect, and it is highly probable that the planoblasts, notwithstanding their
powers of locomotion, never, unless drifted by winds or currents, wander far from their fixed
trophosomes. I can find no evidence that any one of the hydroid medusa? enumerated by Forbes
in his Monograph of British Naked-eyed Medusae has been met with beyond the province dis-
tinguished below as the Boreo-Celtic. Several species of hydroid medusae have been observed
by Mertens in the North Pacific and other seas during his voyage round the world, while none of
them have been as yet recorded from localities at a distance from those in which this excellent
observer had noticed them. A similar localization may be affirmed of the species which Peron
and Lesueur as well as other circumnavigatory voyagers have noticed in intertropical and other
sens traversed by them ; while those which Gegenbaur has so fully described from the Mediter-
ranean appear to be similarly limited in their distribution.

There can be little doubt that some facts in hydroid distribution must be attributed to the
agency of man, and that hydroid trophosomes are not unfrequently carried to a distance while
attached to the bottoms of ships and to floating timber. The exceptional correspondence between
the hydroid faunas of South Africa and Britain has been just accounted for in this way. It is
possible that Cord^Iophora hiciisfris owes its introduction into our docks, canals, and rivers to a
migration of this kind, which would have its exact parallel in the case of Dreisse.na polymorpha,
a mollusc which had been undoubtedly introduced into this country in a similar way, and
is now by no means sparingly distributed through the lirackish and fresh waters of the British
Isles.

What has been just asserted regarding the wide distribution of generic forms must be
regarded as referring only to the calyptoblastic hydroids, for if we except the collections made by
the North American zoologists on their own shores, it is only the calyptoblastic species which
have been collected from distant regions of the globe in sufficient numbers to justify generalisa-
tions as to their distribution, or— with very few exceptions — in a sufficiently good state of preser-
vation to render possible a satisfactory determination of them.

Even among the calyptoblastic hydroids we are aluiost confined to the Serfidarince for
facts on which to base any reliable conclusions as to distribution. Almost all the specimens of
hydroids which have been brought to us from other parts of the world than our own latitudes
have either been picked up dead from the seashore, or if obtained in a living state by means of
the dredge, have been brought to this country in a dried condition. Now no gymnoblastic hydroid
can under such circumstances retain any characters of value ; and even among the calyptoblastic
species the Canpanularina, with their delicate and easily detached hydrothecae, are in their dried
state often little better fitted than the dried Gymnohlastea for satisfactory determination. Until,
therefore, collectors make it a point to secure fresh specimens and put them at once into spirits.

160 DISTRIBUTION.

or (lelermliie tlicm on the spot, our data for tlie Geographical Distribution of all but the Serlu-
larince, with their firm, chitinous, ])ersistent hydrotliecae, must continue very defective.

It is, indeed, by no means improbable that the distributional relations of the Gymnohlastea
are in many respects very diff'erent from those of the CalyptoUastea, and that not only the species,
but a large number of the genera of the gymnoblastic hydroids are confined within limited areas
of distribution. Numerous liydroid medusae have been noticed by navigators in tropical and
other distant seas, and though these have not received the attention they deserve, there is reason
to believe that among them are many gcnerically different from any which frequent our own
latitudes. It is probable, as we have already said, that these medusae do not s])ontaneously
migrate far from the rooted trophosomes in which they originate ; and if so their presence would
indicate not only the existence of these trophosomes in the seas frequented by the medusae, but
their limitation within definite areas corresponding to those of the medusaB.'

Now, with the exception of such few as may have originated directly from the egg, it is
almost certain that all these medusae are the planoblasts of either gymnoblastic or campanularian
trophosomes,— the two groups of which, as has been said, collections from otiier countries, have
hitherto afibrded us such few and imperfect examples.

That a more complete knowledge of the gymnoblastic hydroids of other parts of the world
would result in the discovery of numerous generic forms, limited in their distribution to definite
areas, is rendered further probable by the fact that in almost the only region beyond tiie European
seas where they have been subjected to a scientific study, many generic forms, entirely diff'erent
from those of Europe, have been established. I refer to the researches of the American naturalists,
more especially to those of the two Agassizs, and of Clark, McCrady, and Stimpson, which have
resulted in the discovery of seven genera of gymnoblastic hydroid trophosomes inhabiting the
Atlantic shores of North America, and unknown upon the eastern side of the Atlantic.

But few facts regarding the distribution of the Eleutherohlastea are known to us. The
freshwater genus Ili/dra, the only generic representative of this group, appears to be very widely
distributed, for it is not only spread througiiout nearly the whole extent of Europe, but it has
been recorded from North America and from Northern Afiica. The extra-European examples of
Hydra, however, require further examination before their diff"erence from European species or their
identity with them can be asserted with confidence.

Though the species of cajyptoblastic hydroids have been proved to be on the whole allocated
to definite localities, a few would appear to be widely distributed. Serlnlaria polysonias, a very
abundant British species, is also one of the most common sertularians of tiie Mediterranean, where
I have found it ranging from Naples to Nice ; while, according to the observations and identi-
fications of Agassiz, Busk, Hincks, and others, it occurs upon the Atlantic shores of the United
States, and on the shores of Greenland, Norway, Madeira, South Africa, the Red Sea, and the Falk-
land Islands ; and I have also determined this species among liydroids collected in New Zealand.
Flumularia ohliqita, another British species, has, as already stated, been recorded by Hincks from
Australia. South Africa and Australia are given by Busk as localities for Serlidarla ojjercidata,
another abundant British species, while Ilincks appears to have also identified it among hydroids
from Patagonia, the Falkland Islands, the Auckland Islands, New Zealand, and Kerguellan's Land.

^ The occurrence of IiyilroicI mcduscc at a great distance from tlie nearest laud is probably due
to the diiftiu'r action of winds and currents.

DISTRIBUTION IN SPACE. 1(51

A trophosome (without gonangia) collected by Dr. Hancy at the Cape of Good Hope, and now
in ]\Ir. Rusk's collection, is, as I have already said, indistinguishable from that of our common
Scrtularia puwila, while this species with its gonangia occurs in the New Zealand collection
already referred to, and— judging from the trophosome, for no gonangia are present — it also
occurs, as stated above, among some hydroids obtained by Dr. Scouler in California. Indeed,
Sertularia pumUa and Lafoea diimosa appear to be the only species yet discovered common to the
Atlantic and Pacific shores.

If these determinations be accepted, Serhilarella jjolj/zonias, Sertularia operculata, and Seriu-
Inria jnimila have quite a cosmopolitan range. Resides those now mentioned, including the
exceptionally large number of European species which occur also in South Africa, some others
have been recorded from other parts of the Southern Hemisphere, but the evidence of identity
does not appear to me sufficient for acceptance. We should bear in mind too that in many cases
the specimens brought to us from abroad are entirely destitute of their gonosome, and without
this important element of identification om* determinations, though we may fairly assume their
correctness, are scarcely otherwise than provisional, even though the trophosomes may present a
complete agreement.

The parts of the world from which the most abundant data have been obtained for the
establishment of definite provinces of hydroid distribution are found in the North Atlantic and
its great eastern offset, the Mediterranean.

When we compare the hydroids of the Rritish seas with those which have been recorded
from the Scandinavian shores as far even as the North Cape, and from the shores of Relgium
and the Atlantic shores of France and Spain, we find such a uniformity in the hydroid fauna of
the whole of this coast-range that its division into distinct hydroid provinces cannot be thought
of. The same group of forms, with but little variation, may be also traced westward across the
North Atlantic, by the Faroe islands, Iceland, Greenland, Labrador, Newfoundland, and Nova
Scotia. On tlie coast of the State of Maine a decided change begins to show itself by the
introduction of many species unknown on the eastern side of the Atlantic, and the difference
becomes still more marked as we proceed southwards along the Atlantic shores of the United
States, until in South Carolina and Florida — judging from the species enumerated by McCrady
and Agassiz — not a single hydroid has been found which can be referred to a species known to
inhabit the eastern shores of the Atlantic.^

Within the range now mentioned two distinct hydroid faunas may be distinguished. One
of these may be followed along the whole of tiie western shores of Europe round to the North
American shores by Greenland and along the coasts of the intervening islands, until it attains
its south-western hmit a little to the south of Nova Scotia.

A comparison of the hydroid dwellers in this northern, north-eastern, and north-western
Atlantic area with those which have been recorded from the Atlantic shores of North America,
between the State of ]\Iaine and the southern point of Florida, shows that though there is a con-
siderable number of species common to the two, there are yet so many forms which are special to
this more southern area, that we are justified in regarding this portion of the Atlantic shores of

^ Alexander Agassiz has given a very full list of Noitli American hvdroids, including both the
fixed trophosomes and such free hydroid medusfe as have not yet been traced to their trophosomes.
' lllust. Catal. North American Acalephae/ p. 222.

162 DISTRIBUTION.

jS'ortli America as another province of hydroid distriljution distinct from that of the northern,
north-eastern, and north-western shores of the Athmtic.

Again, the hjdroid fauna of the Mediterranean has many special forms, and differs so much
from those of the other two regions, that we are compelled to raise it also to the rank of an
independent province.

Three definite provinces of hydroid distribution may thus be regarded as well established
in the great North Atlantic area. These may be designated as the Boreo-Celtic, Northern
Atlanto-American, and Mediterranean.

The West Indies will probably constitute a fourth, the hydroids of the West Indian seas being,
so far as w^e know, very distinct ; the species, however, with which we are acquainted from this
locality are not yet sufficiently numerous to justify us in regarding a West Indian province as
fully established.

The data we have from other parts of the world are also very imperfect. Among these the
Pacific shores of North America, New Zealand, Australia, and South Africa are the regions from
which we have obtained the most abundant information ; and though we are as yet far from
knowing as much of their hydroid faunas as could be desired, I believe we are justified in regarding
all but South Africa, whose jjresent hydroid fauna would appear to be a spurious one, originating
largely in artificial immigration — as so many distinct provinces of hydroid distribution. In the
North Pacific Ocean especially, we would seem to have a singularly well-marked area, for we have
seen that, with the exception of Serfularia jmuiila and Lafoiia dumosa — species which may have
been easily carried on the bottoms of ships — not a single species inhabiting this region has yet
been proved to occur in any other part of the world-

The exact limits of this North Pacific Province cannot yet be assigned. It would seem,
judging from the species which I find common to the two, to extend at least from the North of
Vancouver's Island to the South of California. We know nothing of the hydroids of the Asiatic
shores of the Pacific, but, arguing from the greater continuity of the coast line by wiiich the
North Pacific Ocean is encircled, we may well believe that there is a greater agreement in their
hydroid faunas between the two sides of the North Pacific than between the two sides of the
North Atlantic.

Besides the three Provinces enumerated above as determined with sufficient certainty in the
North Atlantic, four others may thus with good reason be indicated as probable, namely, the
West Indian, the North Pacific, the Australian, and the New Zealand.

The generalisations now given are the utmost that we are justified in founding on the facts
before us. Wherever in other parts of the world hydroid trophosomes have been met with, it is
only one or two species that have been collected or recorded from any one region — an amount of
material quite inadequate for the indication of definite provinces.

h. Veriicnl Dutiihidion.

General covsidera1ions.—1\\Q vertical or bathymetric distribution of the IIydroida has, like
their horizontal distribution, been very imperfectly investigated, and beyond the North Atlantic
area we have scarcely any available data.

Until quite recently we should have deemed it sufficient to express the vertical distribution

DISTRIBUTION IN SPACE. 163

of the rivDROinA by referring every species to some one or more of the bathymctrical zones of
Forbes ; and for all that regards the tidal shores of our own latitudes, down to a depth of about
one hundred fathoms, the zones of depth, as laid down by Forbes, will be found convenient.

The recent deep-sea explorations, however, have shown how entirely Forbes's views must be
modified in all that regards the deeper regions of the sea. His zero of submarine life has not yet
been found, and has probably no existence, for there is no distance from the surface to which the
sounding line or dredge has reached — even though the enormous depth of more than two thousand
four hundred fathoms has been explored by them — where living beings are not now known to
dwell.

It is plain, too, that exactly equivalent regions are not necessarily demonstrable in all seas,
and that zones situated at the same depth from the surface in different seas may present physical
characters so very different as to determine in each an entirely different fauna and flora.

If this difference could be discovered from known data ; if, as in the case of the zones of
altitude in the distribution of terrestrial life, we could take the latitude as a factor, which, with
the distance from the sea level, would express the conditions which determine the peopling of
each zone, we should find the marine zones as full of interest and significance for the laws of dis-
tribution as we know to be the case with the zones of altitude. In the case of the sea, however,
so many disturbing operations come into play that general assertions can scarcely be ven-
tured on.

The higher regions of the sea, as may be expected, participate largely on the effects of super-
ficial currents, and in local cliriiatic and tidal influences. We are not, however, to suppose that
the deeper regions are withdrawn from distiu'bing influences ; we now know that the distribution
of heat in the deeper sea regions is very irregular, being in many cases under the influence of local
conditions which cannot be determined by any a priori reasoning ; such, for instance, as the sub-
marine currents, which have determined the deep cold area demonstrated by the "Lightning"
and "Porcupine" explorers in the North Atlantic.

It is such facts as these which greatly take away from the value of definite bathymctrical
zones as a form in which the submarine distribution of animals may be expressed. When applied
to locally limited faunas they are very convenient ; but in comparing the faunas of seas far apart,
we must be careful not to give them too much value, or regard as equivalent what have really
little relation with one another.

Indeed, the recent investigations in the animal Kfe of the deep sea have only shown how very
far we yet are from a knowledge of the laws of the bathymctrical distribution of animals. A
number of facts of great interest have been accumulated, but these, so far from being in favour of
a definite distribution in depth of marine animals, only extend the known range of forms to wiiich
our earlier imperfect knowledge had assigned a much more confined and definite limit.

There can be no doubt, however, that while many animal species have been now proved to
extend through almost all depths hitherto explored, there are some whose bathymctrical distribu-
tion is more limited, and every explorer of marine life knows that there are both shallow water
and deep water forms. Some species, indeed, may be allocated to very definite zones, and when
we assign to each zone the species which have hitherto been found in it, determining the few
which have not yet been met with beyond its limits, and indicate moreover certain relations
between the animal and vegetable life of the sea, we shall have said almost as much as our
knowledge of the distribution of animals in depth will justify.

23

164 DISTRIBUTION.

In the jilienoiucna connected with sul^uiarine vegetation, indeed, we shall find conditions
which exert an obvions inflnence on the distribution of animals. It is certain that in the range
of marine life a much more direct and intimate relation exists between depth and vegetation than
between depth and the presence of animal life. With the deeper regions vegetation has nothing
to do. We know that the higher forms of marine plant life entirely cease at a comparatively
slight depth, and even the lower forms disappear long before the deepest regions yet explored have
been reached, while no limit has yet been found to the extension downwards of living animals ;
even the BiatomacecB and other protophytal forms giving place in the deeper regions to their
protozoal representatives.

Now, one of the most marked elements in determining the animal life of submarine zones
will naturally be sought for in the abundance or paucity of their vegetation ; for this affords food
to the vegetable feeders which in their turn supply the carnivorous tribes, while both find in its
more or less luxuriant development mechanical support and shelter.

Though the Hydroida, less probably than many other groups of marine animals, are
dependent on the surrounding vegetation, it must yet be admitted that in the region of submarine
vegetation, and in that which stretches down beyond it, we have thus the limits of two natural
zones of depth, and there can be no doubt that these present us with two sets of conditions which
go far to determine the bathymetrical distribution of animals.

But in order to express with sufficient accuracy the distribution of animals in depth,
a more minute division is needed, and it will be found convenient to divide the entire
depth into several definite zones. Of these the higher ones will be each characterised not
only by its faima, but by the form of vegetation which is special to it ; wdiile below we have a
vast plantless region stretching downwards into depths which until lately were regarded as
unfathomable.

The zones which lie between high water mark and the lowest level of spring tides, and con-
stitute the Literal and upper part of the Laminarian zone of Forbes, are, at least along the tidal
shores of our own latitudes, where they are loaded with a profuse marine vegetation, very rich in
hydroid life. In this region two distinct physical habitats must be distinguished, each exerting its
special influence on literal life, and each characterised by a more or less definite hydroid
fauna.

One of these consists of the ground which during the ebb becomes exposed to the air
retaining only so much moisture as may be prevented from evaporating by the clothing of sea-
weed and the projecting ridges of rock. The other is formed l)y the rock pools, — reservoirs of
water of greater or less extent which are left behind by the retreat of the tide.

Again, in the deeper regions the physical condition of the bottom, whether rocky or covered
with large stones, or sandy or shelly, or overspread wiih ooze or mud, will exercise an important
influence on the distribution of the Hydroida and of other marine animals.

Among the positive facts which the scientific exploration of the deep sea has established with
regard to the distribution of the Hydroida, one of the most important is that the range in depth
of hydroid life is not surpassed by the known range of any other group of the animal kingdom.
For while some species occur rooted to the rocks and seaweeds close to high-water mark — not to
mention the pelagic planoblasts whose life is spent on the very surface of the open sea amid all
the influences of the atmosphere and of the light and heat of the sun — there is evidence of the
existence of others at a depth of nearly three miles from the surface, the " Porcupine" explorers

DISTRIBUTION IN SPACE. 165

Iiaving brouglit up fragments of a li3(lr()i(l from a depth of 2 135 fathoms, tlic greatest at wliieh
dredging has ever been accomplished.'

Even single species of hydrnids arc distinguished by the great extent of their bathymetrical
range. Thus SerlidnrcUa j)oIj/~oiiia.s; a species also remarkable for its very wide horizontal area,
ranges in vertical distribution from a zone between tide-marks to a dejjth of 374 fathoms, from
which specimens were brought up during the expedition of the " Porcupine." llydraUmania
falcata was obtained during the same expedition from a depth of 542 fathoms, though it is a
common species in the " Coralline zone" of Forbes, which corresponds to a depth of between fifteen
and fifty fathoms, while T/miaria articulata M'as brought up from G32 fathoms, though frequenting
a depth of less than fifty fathoms round our shores.

Many species which have not yet been obtained elsewhere were brought up from great
depths by the dredges of the "Porcupine." Among these is a Dijjhasia from a depth of G32
fathoms, while a plumularidan, which must be referred to a new geiuis, was brought up by the
same haul of the dredge. Two new sj)ecies of T/iuiaria were dredged from a depth of 640 fathoms,
and a Lafoea from 345 fathoms. A Sertularella, nearly allied to 8. Gayii, of which it may
perhaps be regarded as only a variety, ranged from 290 to 605 fathoms. It is a fact, by no means
without interest, that in every case hitherto observed, these deep-water hydroids belong to forms
which produce fixed sporosacs instead of planoblasts.^

The cold area lying between Shetland and the Faroe Isles, which is overflowed by a deep icy
current from the polar seas, and whose discovery by the " Porcupine" explorers constitutes one
of the most important additions to our knowledge of the physical geography of the North Atlantic,
is not without a deep sea hydroid fauna, although its bottom varies from the freezing point of
freshwater to nearly two and a half degrees of Fahrenheit below it. Tlie two new species of
Thuiaria already alluded to were olitained from it where the temperature of the bottom is as
low as 30° Fahr., while from the same area the new plumularian genus, with the new species of
Diphasia and Lafoiki, also referred to above, were obtained in water whose temperature varied in
different places from 30°-5 Fahr. to 29°-S Fahr.'

' The record of this fact is contained in the Report of the " Porcupine" Expedition, where, from a
depth of 2435 fathoms, in lat. 47° 38' N., long. 12° 08' W., the dredge is stated to have brought up
" two fragments of a hydroid zoophyte." (" Preliminary Report of the Scientific Exploration of the Deep
Sea in H.M. Surveying Vessel ' Porcupine,' during the Summer of 1869." — ' Proc. Roy. Soc.,' Nov.,
1S69, p. 429.) The specimens seem unfortunately to have been lost, so that no special determination
of the hydroid obtained from this great depth has been possible.

" A special report, which I have prepared on the hydroids collected during the expedition of the
" Porcupine," will appear in the general Report of the Expedition.

^ In connection with deep dredging we must here refer to the researches of Sars, who, before the
English exploration of the deep sea was undertaken, dredged round the Scandinavian shores in depths
which, however, never exceeded 450 fathoms. In these dredgings he obtained, from a depth of 300
fathoms, two hydroids, Campanularia verticellaia, Linn., and a new genus and species, Lafocina tenuis,
Sars. (See Sars in " Yidenskabs-Selskabs Forhandlinger," for 1868, p. 246. Translated in 'Ann.
Nat. Hist.' for June, 1869, p. 423.)

The late exploration of the Gulf Stream, undertaken by the United States Coast Survey, must
also be referred to. In this expedition very important additions were made to our knowledge of the
fauna of the sea bottom in the western part of the Gulf Stream, though no dredgings which can be

166 DISTRIBUTION.

The great majoiity of the Hydroida, however, belong to umch shallower water, where they
are chiefly distributed through the great belt of submarine vegetation already referred to. This
belt, as stated above, extends downwards through a region divisible into definite zones of depth.
In the systematic part of the present work I have given, so far as it was possible to determine
them, the bathymctrical areas occupied by the various species described. These I have endeavoured
to express in bathymctrical zones based on those of Forbes — a method which will apply sufficiently
well to the Atlantic shores of Europe and of North America. It is less applicable to the
Mediterranean, where the Litoral and Lammarian zones, which take so very important a part in
the hydroid distribution of our own shores, can scarcely be said to exist,^ while from other
parts of the world the data we have received are so scanty that scarcely anything can be
asserted regarding the depths inhabited by such few hydroid species as have been found there.

Batliymetrical zones and their characteristic hydroids. — In order to convey some idea of the
leading facts connected with the distribution of the Hydroida in depth, it may be well to take a
glance at the most striking features presented by each of those zones which, in the systematic por-
tion of the present work, will be referred to in recording the batliymetrical range of the species
there desci'ibed. They are, with some modifications, identical with the zones of depth, as originally
laid down by Oersted for the Danish coast, and afterwards extended to greater depths and otherwise
developed by Forbes. Our survey will embrace their leading physical and botanical characters
and the most striking features of their hydroid faunas as presented round the shores of the British
Islands. From what we have already said, however, we must avoid giving too much weight to
the influence of these zones on distribution, for few species are absolutely confined within the
limits of any one of them.

The depth-regions which we shall here distinguish are six in number, namely,
I. The Surface Zone. II. The Litoral Zone. III. The Laminarian Zone. IV. The
Coralline Zone. V. The Deepwater Zone. VI. The Abyssal Zone.

I. The Surface Zone. — This, though an exceedingly important depth-region, has been very
generally ignored as a special zone in the bathymctrical range of marine life. It is formed by the
stratum of water which reaches from the surface of the sea to a depth of two or three feet, and has
nothing to do with the subjacent ground. It is exposed to the direct action of the sun and of
the atmosphere, and is, more than any of the other zones, under the influence of latitude and

compared in depth to those of the " Porcupine" explorers were attempted. Some of the hydroids
obtained have been examined by Mr. de Pourtales, wlio describes five new species, namely, an Antennu-
laria, three species of Halecium, and a Tubularia ; none of these, however, came from a depth greater
than 270 fathoms. (See " Contributions to the Fauna of the Gulf Stream at Great Depths." By
L. F. de Pourtales, Assist. U.S. Coast Survey, Bui. IMus. Comp. Zoology, at Harvard College.)

I am myself engaged on the remainder of the hydroids collected duiing this important expedition,
and hope to be soon able to make known the results of my examination.

* It may be supposed that the species which inhabit the region between the high and low water
lines on coasts where there is a well-marked tide range are fully represented on coasts where the tide
range is but slight, with this difference only, that in the former case they are spread over a wide belt,
in the latter condensed witliin a narrow one. Experience, however, shows that this is not the case,
and that shores with a narrow tide range have but a poor litoral fauna.

DISTRIBUTION IN SPACE. 167

cliiuatc. Wliilc the explorations of the dredge have revealed to us the fauna of other zones, this
has yielded up its treasures to the towing net.

The hydroid fauna of the Surface Zone consists of the multitudes of planoblasts which have
been liberated from the fixed tropliosomes of other zones, and of such hydroid medusas as may
have been directly developed from the egg, all finding here conditions suited to their love of sun-
light and their great powers of locomotion.

A few tropliosomes, however, which root themselves to floating seaweeds, may perhaps l)e
regarded as belonging to the fauna of the Surface Zone, just as some exceptionally formed plano-
blasts would seem to frequent other zones. To the latter belong Clavatella and Eleufheria,
which are found creeping over the seaweeds and nullipores of rock-pools in the Literal Zone,
and probably also Cladouema, which, besides its power of swimming like other medusas, has also
the faculty of mooring itself to fixed bodies.^

The Surface Zone is not without its plant-life, but its flora is a floating one — the gulf-weed
of the Sargasso Sea {Sar(/assum hacci/erum), and the lower Algse — Oscillatorm and Biatomacece,
which often accumulate in such quantities, even round our own shores, as to impart their colours
to the sea for many square leagues.

With the hydroids of this zone is associated a rich pelagic fauna of Badiolaria and other
Protozoa, of discopliorous raedustc, and of siphonophores and ctenophores, as well as of pelagic
molluscs — pteropods and heteropods ; and pelagic ^^«//«/o«« — Sar/itta and Tomopteris ; and the
larval forms of echinoderms and annelidans, and of molluscs, which in their adult state
creep over the ground or moor themselves to the rocks. Certain fishes, too, must be included in
it, for many species have their habitual abode in the Surface Zone, and pursue their prey in this
highest region of the sea. It is the zone of sun.shine, the region where life becomes intensified,
and where beings whose organisation specially fits them for the enjoyment of the more exciting
influences of the external world find a congenial dwelling-place. It is the zone, too, where phos-
phorescent animals congregate in countless multitudes, and light up at night with their mysterious
fii'es the dark surface of the sea.

II. Tlie Liforal Zone. — The Litoral Zone constitutes round oui' shores a well-marked belt.
It extends through the entire space which exists between the flood and the ebb levels of ordinary
tides. Among the hydroids which occur in it are some of those species which are the most
decidedly limited in their bathymetrical range. Of the various species which constitute its
hydroid fauna some will be found rooted to the moist rocks which have been left uncovered by
the retiring tide, or may be seen spreading over the fronds of the ohve-coloured Algae — Fucus
vesiculosus, F. nodosus, and F. serratm — which form the chief features in the vegetation of
the zone ; while others have their favourite abode in the rock pools with which this zone
usually abounds, and in which they are associated with hundreds of living beings belonging to
very different groups, all findhig like the hydroids a congenial abode in the clear waters of the
rock-pool.

The greater part of the species, however, which are peculiar to this zone, are not found in the

' The curious hydroid forms, Nemopsis and Acaulis, have been taken by the towing net in this
zone; but these, as we have already seen, are probably only the detached hydranths of tropliosomes
routed to the bottom of some of the deeper zones.

168 DISTRIBUTION.

rock-pools, but in that portion of the belt which is left uncovered during the ebb, and where they
meet with conditions entirely different from those w hich are present in the deeper zones from
which the sea never retires.

Among the hydroids which contribute more especially to characterise the Literal Zone may
be mentioned the following :

Clava squamafa, Coryne puHiIla, Clavatella proliferci, Laoiiicdea Jlexuosa, GonoUiyran Loveiii,
Sertularia pumila.

Of these, Clavatella prol if era is confined to the rock pools, occurring In the smaller pools
which are situated near the level of high-water neap tides ; while the others attach themselves to
the moist rocks or to the surface of the seaweeds which are left exposed during the ebb.

III. The Laminarian Zone. — This zone embraces the average range of that portion of the
shore which is uncovered only at spring tides. It is the favorite abode of numerous hydroids,
and since the sea ebbs from it as in the Litoral Zone, it presents both rock pools which at all
times retain a supply of water, and stretches of shore from which the sea entirely retreats.

Unlike the Litoral Zone, however, it is only during spring tides that this zone becomes
exposed. Instead, thei'efore, of l)eing laid bare to the atmosphere twice in every twenty-four
hours, there are two intervals of nearly a fortnight in each month, during which the sea never
leaves it, and even then — since it forms only a narrow baud at the extreme limit of low water —
it is but for a short time at each ebb that it continues exposed. Its conditions are thus different
from those of the Litoral Zone, and as may be expected, it possesses a different fauna.

The fauna of the Laminarian Zone is a very rich one. This region is the abode of such species
as cannot endure the long and repeated exposure to the atmosphere to which those of the Litoral
Zone are subject, and which yet suffer no injury from such sjiort withdrawal of the water as they
experience after each of the long intervals which intervene between every access of spring tides.

The species which frequent the rock pools are here, as in the case of those with a similar
habitat in the Litoral Zone, less decidedly limited to one bathymetrical area than are those which
attach themselves to exposed rocks and seaweeds. The Laminaria digitata, associated with
Rhodomenias and other red and purple Algae, botanically characterises this zone round the shores
of the British Islands, and of the other more northern parts of Europe, and some of the most
distinctive hydroids of the region may be seen congregating round the roots and stems, or
spreading over the broad fronds of this grand seaweed. In the Mediterranean, with its almost
tideless shores, the place of the Laminaria is taken chiefly by beautiful Cysioscirce, whose
narrow heath-like ramifications give support and shelter to iiydroids of other species than those
of our northern laminarian fauna.

The following species may be cited as among the most characteristic hydroids of the
Laminarian Zone of Britain :

Tuhularia larynx, Tiibularia indivisa, Tuhularia bellis, Coryne vayinaia, Syncoryne eximia,
Myriotliela arciica, Ohelia geniculata, Camjxinularia Jolinstoni, Sertularia operculata, Sertularella
rugosa, Aglaoplienia pduma.

IV. The Coralline Zone. — This zone extends from the lowest level of spring tides to a
depth of about fifty fathoms from the surface, and is never exposed in any state of the tide. Its
bottom varies much, and affords appropriate habitats for correspondingly varied forms of life.

DISTRIBUTION IN SPACE. 169

In some places it is rocky, in ollicrs saiuly or niiiddy, or covered with dctaclied blocks or with
old shells. It possesses but a scanty vegetation, and this consists chiefly of the red alga,'.
Hydroid life, however, would seem to attain here its niaxiuuini of development. The sandy and
muddy bottoms possess but few species, while our knowledge of the hydroid tenaflts of the
rocky ground is still dcticient, in consequence of the difficulties here experienced in ex|)loration
by the dredge, and our acquaintance with them is chiefly derived from such as have been brought
up entangled in the lines of the fishermen. All the more minute forms which attach themselves
to the rocks must necessarily escape this mode of capture, and it is highly probable that the
rocky bottoms of the coralline and other deeper zones abound with an unknown hydroid fauna,
the source probably of many of those free-swimming planoblasts which have not yet been traced
to their trophosomes.

The stony and shelly bottoms, on the other hand, offer great facilities for the use of the
dredge, and from none of the zones has such an abundant hydroid fauna been procured.
The following species may be enumerated as among the most characteristic :
Corymorpha nufans, Hydractinia echlnata, Endendrium ramosum, Periyonimus serpens, Peri-
gonbmis mimdus, Bougainvillia ramosa, Bicoryne conferta, Obelia longissima, Laomedea vohbilis,
Laomedea verticillata, Copphiia arcfa, Lafoea dumosa, Haleciiim Jialecinum, Sertularella polyzonias,
Sertularia fdicula, SertuJaria aryoitea, Serfidaria abietina, Diphasia tamarisca, Thuiaria thuja,
Hydrallmania falcata, Antennidaria antennina, Plumularia setacea, Plumularia pinnata.

V. The Deepwater Zone. — The Deepwater Zone extends from the lower limit of the Coralline
Zone to a depth of about 100 fiUhoms from the surface. The Htdroida, as we have just seen, had
attained their maximum of development in the Coralline Zone, and we now find them begin to
decline in variety of form and in multitude of individuals. The plant-life, with the exception of
Diatomacea and some nullipores, has disappeared. There is tlie same variety of ground in this
zone as in the Coralline, but the muddy bottoms with their scanty hydroid life are more frequen*.
For our knowledge of the species frequenting the rocky bottoms of the Deepwater Zone we are
also mainly indebted to the long lines of the fishermen.

Most of the species, however, which occur in this zone have also been found in others.

The following are deepwater species, and will give a fair notion of the character of the
hydroid fauna of the zone :

Tubularla simplex, Tubularia attcnuata, Merone cornucopiee, Lafoea fruficosa, Grammaria
abietina, Haleciiim muricatum, Halecium labrosimi, Sertularella Gayii, Sertularella tricuspidata,
Diphasia pinaster, Thuiaria thuja, Thuiaria articulata, Aylaophenia myriophylltnn, Plumularia
catharina, Plumularia frutescens.

VI. The Abyssal Zone. — This zone commences at the lower limit of the Deepwater Zone, and
extends to the greatest oceanic depth yet determined. Our knowledge of its faunas is of quite
recent acquisition. We are mainly indebted for it to the deep-sea dredgings which have been
lately carried on in the North Atlantic, and which have formed so important an era in zoological
research. Its bottom would seem to vary like that of the two preceding zones, though great tracts
of fine mud loaded with Globiyerince and other protozoal forms, would seem to constitute the
most widely extended and characteristic grounds. With the exception of Diatomacea, plants
are entirely absent ; and even the Diatomacea disappear in the lowest regions of the zone.

170 DISTRIBUTION.

The lijdroicls which tenant tlie Abyssal Zone include some higlily interesting forms unknown
in the others. Among them are most of the new species discovered by the " Porcupine" explorers.
It is an interesting fact, however, in the bathymetrical distriljution of the Hydroida that several
species ^hich frequent much higher zones extend downwards into this. Thus Laomedea
verticellata has been obtained from the Aljyssal Zone at a depth of 300 fathoms, Sertidarella
polyzonias at a depth of 374 fathoms, while IlijdraUmania falcata has been brought up from a
depth of 542 fathoms, and Thuiaria articulata from a depth of 632 fathoms.

The following species have not yet been found in any zone above the Abyssal.

Lafoea halecioides, Lafoeina tenuis, Diphasia coronifera, Thuiaria Jlexilis, Thuiaria laxa,
Thuiaria salicorma, Plumularia ramulifera, Gonocladium plumosum }

Deep Lacustrine Fauna. — In connection with the bathymetrical distribiition of the
Hydroida, the recent researches of Dr. Forel on the deep fauna of the Lake of Geneva must
be mentioned. He has found the mud of this lake at a depth of even 300 metres by no means
destitute of animal life. The species occurring at this depth belong to the groups of Tnsecta
(larvaj), Arachnida, Crustacea, and worms. At a depth, however, of 75 metres (250 feet) a
much richer fauna was found, for in addition to the groups represented in the deeper zone there
occurred here Mollusca, Ccelenterata and Infusoria, the Ccelenterata being represented by a
Jli/dra.'

2. Distribution in Time.'

Among the extinct forms which have been referred to the Hydroida are the graptolites of
the Silurian period, and though there is much which is still obscure and enigmatical in these
remarkable bodies, our present knowledge of them points towards the Hydroida as their
probable allies.

Admitting the hydroid affinities of the graptolites, it is among these remarkable fossils that
we must seek for the chief evidence of the existence of Hydroida in former ages of the world,
for the association of the rugose and tabulate corals with the Hydroida, which has been insisted
on by Agassiz, has not yet received the verification which is necessary for its acceptance, while
the number of other fossils which have been on good grounds referred to the Hydroida is very
small.

We shall presently consider the well-established instances of ancient hydroid life, and shall
discuss the nature of graptolites and their probable points of contact with the Hydroida ; but
it will be best to notice first certain fossils which have, without, as I believe, sufficient evidence,
been adduced as former representatives of this order.

^ The new species here enumerated will be described in my Report on the Hydroida of the
" Porcupine" Expedition.

^ F. A. Forel, " L'Etude de la Fauna profonde du Lac Leman," ' Bull, de la See. Vaudoise de
Sci. Nat.,' vol. X, p. 217.

^ To Mr. Etheridge, Palaeontologist to the Geological Survey of Great Britain, I must here
express ray thanks for his kindness in looking over the proofs of this section before they were finally
sent to press.

DISTRIBUTION IN TIME. 171

Fossik rrferrrd fo ilie llydroida on inmifficicnt (/ronnds. — The genus Oldlunmn had heeii
founded by Edward Forbes ' for the reception of certain enigmatical inijjressions which occur as
alraost the sole evidence of life in rocks of the Cambrian age. They are not only absolutely
special to this period, but are extremely limited in their geographical di.-itribution, having been
hitherto found only in a narrow area composed of these rocks in the counties of AVickiow and
Dublin.

Tliev present two well-marked modifications of form. In one of these [Oldliaiiiia nnfiqua)
there is a distinct stem, which gives off, at short intervals on alternate sides, fan-shaped clusters
of somewhat dichotomously divided branches. In the other form {Oldhawia radiata) the stem
is absent, and the branches radiate from a common centre.

The branches present an obscurely moniliform or nodulose appearance, which has been
supposed to indicate the remains of hydrothecse, and the fossil has accordingly been referred to
the Hydroida, with the Sertularia argentea as its nearest living representative. Oval-shaped
bodies have, moreover, been described as present on the branches in some rare instances, and
in these it is believed that we have the remains of gonangia."

1 must, however, confess my inability to recognise hydroid characters in these obscure
impressions. The structure which has been taken for hydrothecae is never so well marked as to
justify this interpretation ; and though I have examined a great number of specimens, I have
never met with anything which can be regarded as gonangia.

Whatever be the nature of Oldhamia, then, 1 am not prepared to place it among the
Hydroida. Forbes believed that its affinities must be sought for among the Tolyzoa, and this
view has at least as much in its favour as the other, while that which would regard Oldhamia as
a vegetable impression seems as tenable as either.^ Indeed, there are some facts which seem to
militate even against its organic origin, such as the undoubted extension observable, in many
specimens, of the impression through numerous successive laminae of the bed in which it occurs.

Under the name of Corynoides calicularis Dr. Nicholson has described certain bodies which
he has found associated with graptolites in the Silurian shales of Dumfriesshire.* They are
small, flattened, narrow, wedge-shaped bodies, from one third to a half an inch in length, and
measuring about half a line in diameter at their broader end, from which a few short, irregular
processes diverge. Towards the opposite end they taper away to a fine point, or, according to
Dr. Nicholson, they sometimes terminate here in a double point. In the specimens examined by
myself, however, this double termination was not observed. They certainly never show any
evidence of attachment. Dr. Nicholson refers them to the tubularian hydroids, and finds tlieir
nearest ally in the living genus Corymorpha.

^ Edward Foibes, "On Oldhamia, a New Genus of Silurian Fossils," ' Jouru. Geol. Soc. of
Dublin,' 1848.

- See J. Kiunahan in ' Trans. Roy. Irish Acad.,' 1859, p. 547. Mr. Kinualian's paper contains
the best account we possess of Oldhamia, and is illustrated with excellent and cliaracteristic figures.
Under the name of O. discreta he describes a third form, which, however, is Iiardly distinct enough to
be accepted as a separate species. He strongly advocates the hydroid nature of the fossil, comparing
it to the living Scrlulaiia ari/entea.

' Schimper (' Traite de Paleontologie vegetale,' 1869) regards Oldhumia as a plant, and places it
among the KulUpores.

* Nicholson, in 'Geological ^Magazine' for 1867. vol. iv, p. 108, pi. vii, figs. 9 — 11.

•22,

172 DISTRIBUTlOxN'.

I am iinahlc, however, to participate in this view. Through the kindness of Mr. Carruthers,
of the British jMiiseum, I have been enabled to examine numerous specimens of Cori/noides, and
in none can I find evidence sufficiently clear to lead me to believe in their hydroid affinities.
Dr. Nicholson is probably right in regarding them as originally consisting of a corneous (or rather
chitinous) material, while they may also have been tubular, though we are without any of that
evidence of their tubular conformation which we find so com])letc in the graptolites. If their
affinities be with the tubularian hydroids, the wide end cannot represent a hydrotheca, as supposed
by Dr. Nicholson, while under no circumstances can they have as their representative Corymorpha,
which is entirely destitute of a proper chitinous perisarc. If the wide end represents a distal
dilatation of the pcrisnrcal tube, then the processes from it are without parallel in any known
hydroid ; while if the narrow pointed end be the distal one, there could have been no liydranth
developed in them.

Dr. Nicholson's discovery of Conjnoides forms an interesting contribution to the palaeon-
tology of the Silurian Rocks; its relation with the Hydroida may be possible, but other charac-
ters than those as 3'ct detected in these little, flattened, wedge-shaped bodies are needed before we
can assign to them, with any probability, hydroid affinities.

MM. Duncan and Jenkins have given us a very interesting memoir on a remarkable little
organism from the lower shales of the Carboniferous Limestone of Scotland.' The authors name
it Palceocorpie, and regard it as a tubularian hydroid.

It consists of a little calcareous cylindrical column, about one tenth of an inch in height,
attached by a dactylose base, and surmounted by an expansion in the form of a reversed cone,
the margin of whose wide end is extended into several radiating arms, which, like the rest of the
fossil, are entirely calcareous. Roth column and arms are beautifully ornamented by flutings and
regularly disposed puncta.

The arms are tubular, and open into the summit of the column, which is also traversed by
an axial cavity, while the base, with its root-like prolongations, presents, on section, an irregularly
chambered structure.

Two species of P«/rt>ofo;y«e are described by the authors, and from their very clear description
and the excellent figures which accompany it, there is no difficulty in arriving at an adecpiate
conception of the form and essential points of structure of these singular fossils. I regret, how-
ever, my inabiUty to recognise the hydroid affinities of Palaocorync. The completely calcified
condition of the entire fossil — both base, stem, and arms — and the certainty that it could never
have been otherwise, the peculiar ornamentation of its surface, and the chambered structure of
its base, are all directly opposed to its alleged relation with the Hydroida. 'J'he authors of the
memoir believe that they can find in the living tubularian genus B'rineria features which resemble
some of the most anomalous of the characters oiPalceocoryve. They see, especially in the extension
of the perisarc over a part of the tentacles and hypostome of Buneria a condition which has its
parallel in the calcareous arms (" tentacles") and summit of the cohunn in the fossil. Rut this partial
investment of tlie tentacles and hypostome in Bimeria is flexible and chitinous, and a consider-
able portion of the tentacle remains quite free from it, while in PaltEocoryne the thick calcareous
walls of the radiating arms are all but closed at their distal extremity, where they exhibit at most

' ?. INIiirtiii Dimcaii and II. ^M. Jenkins, "On Puhrocoryne, a Genus of Tubulaiine Hydrozoa
from tlie Carboniferous rorniation," 'Phil. Trans.,' 18G9, p. 093, ]i\. \\\\.

DISTRIBUTION IN TIME. 173

but indications of a minute foramen, quite insufficient for the transmission of a comparatively thick
tentacle like that of any known hyJroiii.

On the whole, though ihe autliors liave clone good service to Palaeontology by making us
acquainted with this remarkable little tenant of Carboniferous seas, I cannot accept their views
of its affinities, and I believe that we must seek for these with the Ithizopoda rather than with
the IIydroiua.

Fossils, exclusive of Ihc graptuHlcs, which have been correctly referred to hijdroid irnphosomes.
— Of the hydroid nature, however, of certain other fossils which have been detected in various
geological formations there can be no doubt. Instances are known of the chitinous basis of
Ilj/dracliiiia having been preserved in a fossil state. Under the name of Cellepora echinata ]M.
Michelin has described a fossil from the sub-apennine group of Asti, and from the superior
Fallunian of Bordeaux and Dax.^ M. Fischer has drawn attention to the fact that the
Cellepora echinata of i\lichelin is really a llydractinia encrusting a Murex or a Nassa, while he
has himself added another fossil Hydraclinia from the Upper Greensand of Mans.^ This he
found in the collection of jM. Ale. d'Orbigny, where it encrusted numerous specimens of Natica
tuberculafa, d'Orbig., from that formation. M. Fischer has assigned to this species the name of
llydraciinia cretacea, while to Micheliu's species he has given that of Hi/dractinia Michelini.

To the two examples thus noticed by Michelin and Fischer I am enabled to add a third
from the Coralline Crag of Norfolk and Suffolk. It occurs among some Coralline Crag fossils in
the collection of the British Musemn, where it encrusts two specimens of Purpura lapillus, one
from Orford, in Norfolk, and the otiier from Redgrave, in Suffolk. It covers with a continuous
crust the shell over which it spreads, and has a minutely alveolar structure, w'ith its surface
thickly set with short blunt spines. The original chitine is entirely replaced by carbonate of
lime. There cannot be the slightest doubt of its being a true Hi/dractinia, and, indeed, it is
impossible to find any characters which can separate ^it from the living Hydractinia echinata.
From the mere fossilised basis, however, which is, of course, all that has come down to us, we
should not be justified in absolutely asserting its identity with the living hydroid.

M. Fischer gives no specific description of the specimens to which he assigns two dis-
tinguishing names, and it is probable that no characters of diagnostic value can be detected
sufficient to distinguish them from one another or from the Coralline Crag specimens, or even
from our living Hydractinia echinata. It is by far most likely that such characters would be found
if an opportunity existed of examining in the fossils the soft parts now entirely lost, more
especially when we bear in mind that they lived at such distant intervals of time as those which
separated from one another the cretaceous, meiocene, pleiocene, and existing epochs.

If, under these circumstances, we should be justified in assigning to the Coralline Crag
species a distinguishing name when no available characters can be found on which a diagnosis
can be based, this may be derived from its geological position, and the purely provisional
designation of Hydractinia pleiocena may be given to it.

Among fossil hydroids Serfularella polyzonaas, one of the most abundant and most widely
distributed species of the present seas, has been cited from the later Pleistocene deposits of
Ayrshire.'

^ jMiclieliu, ' Icou. Zoopli.,' p. 7^>, pi. xv, fig. 4.

" Fischer, ' Bull, de la Soc. Geol. de France,' 2 ser., tome xxiv, p. 689.

' ^Morris, ' Caliil. Brit. Fossils,' p. 63.

174 DISTRIBUTION.

I have sought in vain, however, for the evidence on which this citation rests, tlioiigh there
is no a priori reason why the case should not be accepted as authentic.^

Fossil Hydroid MeduscB. — If we except the graptoHtes, which will be afterwards considered,
the instances now mentioned are the only ones yet known of any portion of an undoubted
hydriforra trophosome having been preserved in a fossil state. Among the most interesting
evidence, however, of ancient hydroid life, is that afforded by the discovery of fossil hydroid
medusae.

Several medusae — some true hydroid or gymnophthalmic forms, and others belonging to the
Discophora or steganophthalmic group- — are now known to exist as fossils. These are all of
Jurassic age, having been found only in the lithographic slates of Solenhofen and Eichstadt, and
it is to Haeckel that we are indebted for most of our knowledge regarding them.

In 1845 a fossil medusa was exhibited by Frischmann, at a meeting of the Nuremberg
Association of Naturalists, and was soon afterwards briefly described by Beyrich under the name
of JcalepJia deperdita." HaeckeP has since subjected this fossil to a careful examination, and,
under the name of Craspedites deperdita, has given a figure and full description of it. He regards
it as a hydroid medusa, and refers it to the family of the Traclii/tiemida. He afterwards^ alters
the generic name from Craspedites to Trachynemites.

The fossil belongs to the white or upper Jurassic formation of Eichstadt, in Bavaria, and is
one of the most perfect impressions of medusae known. Four specimens of it have been found,
and are distributed among the museums of Munich, Carlsrhue, and Berlin. The largest specimen
has a diameter of two inches seven lines, and presents the form of a disc bounded by a circular
furrow, and showing another furrow internal to this, concentric with it, and separated from it by
a distance of five lines. From eight equally distant points of the inner circle there run eight
straight radial furrows towards the centre, without, however, reaching it. Each radial furrow is
deeper and wider in the middle of its length than at either end, so that the convex cast on the
obverse slab has each radiating line projecting almost in the form of a lanceolate leaf.

The obvious interpretation of these appearances leads Haeckel to view the central area into
Avhich the radiating furrows do not extend as corresponding to the place of the manubrium and

' It is quite possible that the Websleria crisioides of Milne-Edwards is a hydroid. Under this
name M. Milue-Edwaids (' British Fossil Corals/ part i, p. 43, tab. vii, figs. 5, 5a) has described a
fragment of a fossil discovered by Mr. Frederick Edwards in the Loudon Clay of Hampstead. He
regards it, though with some doubt, as an actinozoon, referring it to the family of the Gorgonidce.

Judging from M. Milne-Edwards's figure and description, the resemblance of Websteria crisioides
to a sertularian hydroid, notwithstanding obscure indications of a central axis, is considerable, certainly
greater than to a gorgonidan. I have endeavoured, in vain, however, to obtain a sight of the original
and only specimen yet found. This seems to have been lost, and without an actual inspection I
should scarcely feel justified in expressing an opinion as to its affinities.

M. Pictet (' Traite de Paleontologie ') is also disposed to believe in the possible hydroid nature of
Websteria.

' It would seem, however, that the first notice of this fossil was given in 1835 by Fr. S.
Leuckart, who correctly referred it to the impression of a medusa. See Rud. Leuckart, in ' Wieg.
Arch.,' 1870, Band, ii, Seit 279.

' 'Zeit. f. Wissens. Zool.,' 1865, vol. xv, p. 506, tab. 39, fig. 1.

^ Id., vol. .\ix, p. 560.

DISTRIBUTION IN TLME. 175

nioutli, and the eiglit radiating furrows as the remains of the radiating canals.' Tiie inner of the
two concentric circles is regarded by Ilaeckel as the circular canal, and the outer as the periphery
of the umbrella, which has become compressed by the conditions to which it' had been exposed
after death. The dilatations at the middle points of the radiating canals would represent the
generative appendages of these canals. The uud)rella margin is even, showing no trace of a
division into lobes.

The living genus to which Ilaeckel regards the fossil as most nearly allied is Bhopalunrma.
This, when viewed from the summit, presents quite the same essential relations of form as the
fossil, having eight radiating canals, each carrying at its middle point a sexual sac ; and this view
is further strengthened by the fact that in Rhopalonema the radiating canals are, according to
Haeckel, supported by a double cartilaginous band, specially fitted to leave behind a well-marked
impression, while the umbrella itself in the Trachpiemida, to which Rhopalonema belongs,
possesses an almost cartilaginous consistence, and is thus peculiarly adapted for preservation in a
fossil state.

The large size of Traclii/nemiies dcperditus, when compared with any living Traehynemidan,
is remarkable, the largest known living Traehynemidan measuring little more than a quarter of
an inch in diameter, while the fossil has a diameter of more than two and a half inches.

The only other hydroid medusa which has been determined with sufficient certainty from
its fossil imjjression has been referred by Haeckel, to whom we are indebted for our knowledge
of it, to the family of the ^Ei/inidcs.^ Tiie impression, which is very distinct and sharp, occurs
on a slab of the lithographic slate of Solenhofen, forming part of the collection in the Palaeonto-
logical Museum at Munich. It exhibits both umbrella and marginal tentacles. The umbrella is
one inch and a half in diameter, with eight marginal lobes. The marginal tentacles are admirably
preserved. These are eight in number, and can be traced to the intervals between the marginal
lobes of the umbrella. They are eight inches in length, cylindrical, and about a quarter of an
inch thick for the greater part of their length, when they begin to taper to a point at their distal
extremity. They are traversed in their entire length by a reddish-brown line, which there is
every reason to interpret as the remains of an axial canal. From the peculiar way in which the
tentacles lie in the fossil, and the stiff curves into which they are thrown, it appears that in the living
animal they must have possessed considerable rigidity, a condition which is very characteristic of
the marginal tentacles in the living jEfinidte. Indeed, both umbrella and tentacles in the
^pniddB possess for the most part a cartilaginous consistence, which renders them particularly
well fitted to come down to us in a fossil state.

If we follow in the fossil the direction of the axes of the tentacles, when prolonged towards
the summit of the umbrella, we may distinguish eight small pit-like depressions at a distance of a
little more than a quarter of an inch from the margin. These appear to be the impressions of
the eight sexual sacs. No trace exists of mouth or stomach, or of the lateral gastric pouches
which characterise the living yE^inida.

Ilaeckel assigns to this fossil the name oiFalcpgina (jigantea. In referring it to the family of

' Rud. Leuckart, however (' ■Wiegra.-uin's Archiv,' 1870, Band ii, Seit 280), asserts, from an actual
inspection of the Carlsriihe specimen, that the eight radiating furrows regarded by Haeckel as
impressions of the radiating canals, are in this specimen approximated to one another in pairs, and thus
throw doubt upon the validity of Haeckel's determination.

" Ilaeckel, ' Zeit. f. Wissensch. Zool.,' vol. xix, p. 540, tab. 40.

176 DISTRIBUTION.

the JEginidce he appears to have good grotinds. Its enonrious size, however, when compared
with living representatives of this family, constitutes a striking feature in its physiognomy.

Several other fossil impressions of medusae have been described by Ilaeckel, though none of
these can with anything like certainty be referred to the Hydroida. Some of them are pre-
served with great distinctness, and are undoubtedly referable to the Ducophora or steganoph-
thalmic medusas, while some others are so imperfectly preserved as to I'ender it impossible to
determine their systematic position, though in some, at least, the evidence is in favour of their
hydroid rather than their discophorous affinities.

It wiU be seen that the only two fossil hydroid medusae which have been satisfactorily deter-
mined appear to belong, one to the Truchjnemidce, and the other to the ^E(jinid(B, families charac-
terised by certain anomalous features which distinguish them from the ordinary hydroid medusae.

GrapioVdes. — Among the extinct forms of life few possess more interest than these remark-
able fossils, absolutely confined, as they are, to one great section of the palaeozoic rocks, where
their vast abundance, wide geographical distribution, and easy recognition, render them of special
value to the practical geologist.

The graptolites are now by most palaeontologists refen-ed to the Hydroida, and their
living representatives are sought for among the calyptoblastic genera of this order. While,
however, I am unable to recognise their hydroid relations from the point of view from which
palaeontologists have generally agreed to regard them, I believe that their affinities with the
Hydroida are too decided to justify their omission from any complete exposition of the palaeonto-
logical history of this group of the animal kingdom.

The typical form of a graptolite is that of a narrow tube, straight or more or less curved,
emitting from one side a series of hollow denticles, through which the cavity of the tube opens
externally, and having a solid slender rod imbedded in the walls of the opposite side. This
type form (" monoprionidian ") is represented by the genus Graptolites proper, where the
denticles or tubular offsets from the common canal are in contact with one another at theii' bases
and usually for a greater or less extent of their length, and by the genus JRasfrites, where
they are separated from one another by considerable intervals.

But we may conceive of two such graptolites being united back to back, and the resulting
form will then present two series of tubular offsets, one on one side of the main tube and the
other on the side diametrically opposite, while the solid rod will now occupy the axis, holding
just such a position as it would do if it had been formed by the union of the two rods of the
component halves.

This form (" diprionidian") is represented by such genera as Dijjiofjraptiis, where the tubular
offsets stand out more or less free from the sides of the main tube, and by CUmacograptus, where
they are adnate to one another, so as to appear entirely immersed in its walls.

Some other forms also exist, such as Dicrano(jraptm, in which the graptolite, with a double
row of denticles, after continuing its course for a time, divides into its component halves, which
then diverge from the basal portion as two branches, constructed each on the single-rowed type.
Branched single-rowed forms {CIado(/raptus, Dichoi/rapfu-s) also occur. In Bkhoyraptus priuiary
branches radiate from a common point at the proximal end, where they are connected by a web-
like disc, apparently com[)osed of a double membrane of the same nature as that which forms
the walls of the branches.'

' See Hull, 'Graptolites of tlie Quebec Group.'

DISTRIBUTION IN TIME. 177

There are also some anomalous forms [Retiolifes, PIij/lIo_<jraptus), whose structure has not
yet been detcrniined with sufficient certainty to admit of a satisfactory association with tlic true
graptolites ; but the essential features in the morphology of the graptolites, as well as their more
important modifications, are e.\})resscd in the genera already cited.

There is sufficient evidence to show that the graptolites were flexible, and that the solid
parts, which are all that have come down to us, were of a horny or chitinous consistence.
There is also evidence to show that, though some obscure forms {Dendrograjjtus), associated
on insufficient grounds with the graptolites, were apparently rooted, the true graptolites were
never directly attached to any other bodies, as we know to be the case with the hydroid
trophosomes and most of the corals and Folyzoa of the present day.

We are absolutely ignorant of the original contents of the main tube and of its lateral
offsets, and we know just as little of other soft parts which may have accompanied the chitinous
skeleton, so that in attempting to assign to the grai)tolites their position in the system of Nature
we are driven to analogy by no means close as our sole guide.

The resemblance of the forms just described to the trophosome of a calyptoblastic hydroid
- — scrtnlarian or plumularian— after the disappearance of all the soft parts, is sufficiently obvious.
And it is this rescndilance between the fossil graptolite and the recent chitinous skeletons of the
scrtularian and plunudarian hydroids, which has induced modern pakuontologists to refer the
fossil to the order Hydroida, regarding the lateral offsets as hydrothecae, and the main tube as
the chitinous perisarc of the liydrocaulus.^

We shall presently consider whether the exact points of contact between the graptolites and
hydroids have been indicated in this comparison.

The fact which most obviously opposes itself to an acce])tance of the hydroid affinities of
graptolites is found in the presence of the rod or " solid axis," which constitutes an essential
feature in the structure of the graptolite. This rod was apparently of the same chitinous
material as that which formed the rest of the firm skeleton of the graptolite. It is frequently
continued for some distance beyond the distal or growing end, while its opposite or proximal end
usually terminates in a minute spine {" radicle " of Hall), occasionally continued into a long slender
filament, like that of the distal end. It grows with the growth of the graptolite, as can be easily
proved by following the progress of the graptolite from its younger stages, and it is difficult to
explain its increase of length and thickness without regarding it, like the proper perisarc, as an

' Tiie sei-tularian affinities of tlie graptolites have been strongly insisted on by Hall, wlio has
greatly advanced our knowledge of these fossils in his classical work, ' Graptolites of the Quebec
Group,' which forms one of the memoirs of the Geological Survey of Canada. On the structure and
principal modifications of graptolites, the works of Barrande (' Graptolites de Boheme ') and of Geinitz
should also be consulted. The sertularian affinities of graptolites have also been defended by Mr.
AVm. Carruthers, of the Botanical Department, British Museum, and I know of no one who has
worked out this question with so much care and completeness. See especially his " Revision of the
British Graptolites," in the 'Geological Magazine,' vol. v. The same view is advocated by Dr. J.
Alleyu Nicholson, who has also paid special attention to the British graptolites, and has contributed
some valuable facts to our knowledge of them.

I must here express my thanks to Mr. Carruthers for the liberal way in which he has placed at
my disposal his large collection of graptolites, and for the aid ^^hich I Ijavc derived from his extensive
acquaintance with the literature of the subject.

178 DISTRIBUTION.

excretion from the cocnosarc. It is sometimes fouiicl in the single-rowed graptolites to have
become detached from the test or chitinous perisarc, leaving behind it a furrow in which it had lain ;
this furrow being in the more perfect state of tiie fossil converted into a tube by a thin extension
over it of the test.

Though the rod would thus form an extremely excejjtional structure, its presence can hardly
be regarded as offering an insurmountable obstacle to the admission of the graptolites into
immediate relation with the Hydroida. Until lately a similar structure would have quite as
justly excluded from the Polyzoa any animal which possessed it. Tlie discovery, however, of the
living polyzoal genus HhahilopJeura shows that a rod cpiite like that of the graptolite in all points,
except in its not being continued beyond the cell-bearing portion, might be developed in an
animal possessing in all other respects a typical polyzoal structure.'

It is true that the extension of the rod in the fossil beyond the limits of the proper
hydrocaulus appears to increase the difficulty of reconciling its presence with the hydroid aifinities
of the graptolite. I believe, however, that this is, after all, not so anomalous a fact as at first
sight it may appear, and that there is reason to believe that the coenosarc invested by a proper
perisarc was originally continued along what now appears as a free extension of the rod. Its
distal extension would then correspond to what had been the young growing portion of the
graptolite, as yet destitute of denticles, and with its periderm so delicate as to be incapable of
preservation in the fossil, so that the thin perisarc has perished along with the soft coenosarc it
included, its thicker rod-like portion being the only part preserved.

This view is borne out by the fact that in the very young stage of the graptolite a distal
extension of the body along the rudimental rod, and beyond the incipient denticles, may be
noticed ; while it is further confirmed by an observation by Dr. Nicholson," who tells us that in
some specimens of JDiphi/raplm pristis he has seen the common canal without denticles continued
on each side of the prolonged rod.

The continuation of the rod beyond the denticle-bearing portion at the proximal end of the
graptolite may also have been accompanied by an extension of the ca-nosarc and its enveloping
perisarc in this direction, the rod alone remaining in the fossil. To this view an observation of
Mr. Carruthers ' gives support, for he has noticed the prolongation of the rod at the proximal end
of CUmacoyraptus scalaris frequently invested for a short distance by a sheath.

If this explanation be accepted, the continuation of the rod as a naked filament beyond the
denticle-bearing portion of the graptolite need no longer surprise us. A comparison of the rod
to the chitinous spines which bristle over the surface of Hydradinia may also here suggest
itself; but these spines are not only invested by a coenosarcal layer, but are permeated by canals
which are lined by coenosarc, while in other respects the approximation of the graptolites to
Ilydractiina offers too many difficulties to allow of its being attempted.

The lateral spines often present at the proximal end of the graptolite seem to be of a
different nature fr'om that of the rod, and would rather appear to be referable to the same group
of structures as the chitinous spines and variously formed processes by which the hydrotlieca; and
other parts of the perisarc of living hydroids are not unfrequently ornamented.

' AUman, " On Rhabdopkura," in ' Quarterly Journal of Micros. Science/ Jan., 1869, p. 57,
pi. viii.

" ' Geological Mag.,' vol. iv, 18G7, p. 261, note.

' Carruthers, in 'Intellectual Observer' for June, 1867, p. 370.

DISTRIBUTION IN TIME. 179

It has been already said that the advocates of the hydroid nature of graptolitcs regard their
denticles or hollow lateral offsets as hydrothecac. If this be really the nature of the denticles, the
mode in which their cavity opens into that of the main tube is exceptional, for in the living hydroids
the point of communication between the hydrotheca and tube of the hydrocaulus is more or less
constricted, or even provided with an imperfect diaphragm, so that the hydrothecae become proper
chambers, completely differentiated from the common perisarcal tube. Now, the denticles of the
graptolite have their cavity uninterruptedly continuous with that of the main tube, there being
no diaphragm or constriction of any kind at the point where the one passes into the other.'

There is, however, another view of the denticles which will meet this difficulty. The re-
markable bodies known as nematophores, and which are characteristic of the Plamularklm, have
been already described. Among these nematophores there is one form which consists of simple
chitinous offsets from the main tube of the hydroid filled with the protoplasmic matter which con-
stitutes the characteristic contents of all the nematophores. The mesial and lateral nematophores
of Aglaoplienia are of this nature, and a comparison of them with the denticles of a graptolite will
show how complete is the resemblance, as may be seen by a reference to woodcut fig. 51, p.
117, which represents the nematophores connected with the hydm[\\eca of J///aop/ienia jjl/fma.
It has been already shown that the tooth-like processes which project from the edges of the hollow
leaflets, which form the walls of the corbula in A(jJaoplienia, are bodies of an entirely similar kind,
and the resemblance between these and the tooth-like processes of many graptolites is complete
(see woodcut, fig. 30/ p. 60).

Now, it is not alone in general form that the nematophores of Aijlaophenia resemble the
denticles of a graptolite. The mode in which their chitinous sheaths are seen to open into the
common canal of the perisarc after the destruction of all the soft parts, is entirely similar to the
mode of comminiication between the denticles and the common canal in the fossil — in those cases
at least in which the graptolite has afforded facilities for examination such as to leave no doubt as
to the structure of the parts in question — and quite different from that in which the proximal
extremity of the hydrotheca is connected with the common tube of the chitinous perisarc in the
existing hydroid.

I cannot help believing that this is the true view to take of the morphology of graptolites.
If so, the graptolites would admit of an approximation through an luiexpected channel with the
Plumularida. They would then be morphologically plumularidans in which the development of
hydrothecae had been suppressed by the great development of the nematophores, probably the
mesial ones ; while, on the other hand, the existing plumularidan, with well-developed hydrocecae,
would present in its nematophores the last traces of the structure of its ancient representative,
the graptolite.^

' ]\rCoy (' Brit. Pal. Foss.') speaks of a septum at the base of the deulicles in certain graptolites,
but subsequent observations have not tended to confirm this statement.

^ Among the undescribed species in the collections of the United States Coast Survey is a
plumularidan in which that part of the stem which lies at the proximal side of the pinna-bearing
portion — and is accordingly destitute of hydrothecae — carries along its length a single row of fixed
nematophores separated from one another by regular intervals, and appearing to take the place of
hydrothecse. This part of the hydroid, if detached from the pinnate portion, might — except for the
much greater thickness of the common canal in the living thnn in the extinct form — almost be taken
for a recent Rastriics.

24

180 DISTRIBUTION.

Whether the denticles of the graptolites gave lodgment to hydranths, or were filled with simple
protoplasm, as we know to be the case with the nematophores of the living Plumularida, it is of
course impossible to say. If we give analogy its full weight, and extend the resemblance between
the denticles of the graptolites and the nematophores of the plumularidans to the nature of their
contents, we should then have lodged in the graptolite-denticles, not hydranths, but simple masses
of protoplasm, capable of emitting pseudopodial prolongations on which would devolve the duties
of conveying nutriment to the colony. The graptolites would thus manifest relations, not merely
to the IIydroida, but would exhibit others at least as strong to the Rhizopoda. Indeed, but a
step would be needed to convert such an organism into a true rhizopod, which might then be
compared to an association of such rhizopodal foruis as Gromia, united into a composite colony by
a common tube filled with a common mass of protoplasm.

A very general feature in the mode of growth of graptolites is found in the fact that while
the entire graptolite continues to increase in length, the denticles which are situated towards the
proximal end remain of smaller size than those which succeed them, while after thus increasing in
size they again often diminish towards the distal end, the broadest part of the graptolite being
consequently in this case towards the middle.

Now, I know of nothing like this among the living Hydroida, while, on the other hand, the
nematophores vary in form in one and the same colony, and are sometimes found more or less
arrested or otherwise modified towards the proximal end of the branch.

In support of the hydroid nature of graptolites, the occurrence of generative capsules in
these fossils has been recently adduced, and as this is a matter of great importance in the
present question we shall here consider the evidence on which it rests.

Hall has described and figured in one of the double-rowed graptolites [Diplof/rapfus) certain
appendages of an irregularly triangular shape, having one angle continued into a narrow band
by which they become attached to the body of the graptolite. They are arranged with considerable
regularity in two opposite rows, which extend for some length along the sides of the graptolite.
These appendages are compared by Hall to the gonangia of a calyptoblastic hydroid.'

I am indebted to Mr. Etheridge for an opportunity of examining a British specimen of a
Bijjiograptus which carries bodies of undoubtedly the same nature as those of Hall, and to Mr. J.
Hopkinson, who had previously examined this specimen and determined its nature, for the inspec-
tion of an excellent enlarged drawing of it intended for publication." Now, after a full considera-
tion of Hall's description and a careful examination of Mr. Etheridge's specimen, while I admit
the probability of the appendages in question belonging to the generative system, I am unable to
satisfy myself that they are the remains of gonangia. Indeed, they do not appear to me to be
capsular bodies at all, but rather hollow laminae, though the way in which they are occasionally
folded over on themselves may give them the deceptive appearance of having been capsules.

The regularity of their disposition, and the close resemblance between those of the American
specimens and those of the British, will not allow us to regard them as mere parasitical or accidental
growths, and I believe that their connection with the gonosomal system of the graptolite may be
considered as probable. If so, then it remains for us to determine the parts which represent them

' Hall, 'Graptolites of the Quebec Group,' p. 33, pi. is, figs. G — 11.

° Mr. Hopkiusou's description of the specimen hiis since buen publislicd. See ' Ann. Nat. Hist.'
for .Mi.y, 1871.

DISTRIBUTION IN TIME. 181

ill the living liydroid, and tliese I believe will he found in the leaflets which compose the corbidii;
in Jf/laophenia. (Sec woodcut, fig. 30, p. 60.)

The two rows, then, of appendages in the graptolite would, according to this view, represent
a corbula, and the gonangia, if such had existed, would have been borne upon the front of the
graptolite along the bases of the appendages. We should hardly, however, expect to find any
remains of gonangia in the fossil, for in all living hydroids which have their gonangia protected
by corbulse these gonangia are as delicate and perishable as the naked generative sacs in the
Gymnohladea.

The corbulff! of the graptolites were probably open ones, like those of the living Aglaojihenia
viyriophyllum, and of several species from extra-European seas, a condition which indicates a low
stage of diflPerentiation, and represents a form through which the closed corbula of Aglaophenia
pluma, &c., passes in the course of its development. (See woodcut, fig. 30, a, b, c, d. p. CO.)

The view here adopted of the nature of these supposed generative capsules in the graptolite
receives important support from the fact that in every case where they have been satisfactorily
observed the denticles of the graptolite become suppressed in that part of the fossil which carries
the appendages,^ a fact quite in accordance with what we know of the corbulse in the living
hydroids, for in these the hydrothecae with their accompanying iiematophores are replaced by
the leaflets of the corbula, while the naked gonangia of other hydroids are never accompanied
by an atrophy or other alteration of the hydrothecae or neighbouring parts.

In both the American and British specimens the appendages in question seem to have been
supported by a framework of branched chitinous filaments which remain behind after the destruc-
tion of the intervening membrane. The existence of these filaments probably depends on the same
morphological conditions as those which determine the presence of the chitinous axial rod, and it
must be admitted that we have no known analogy for them in any living hydroid, unless the
internal narrow chitinous lamina which passes like a midrib through the corbula-leaflet (woodcut,
fig. 30, A, e) admits of being compared with them.

This comparison of the appendages of Hall with the corbula-leaflets of an Aglaoplienia is in
harmony with the view here advocated as to the nature of the denticles of the graptolite which
we have compared to the nematophores of an Aglaophenia. I believe the corbulae of the living
AglnopUenice to consist essentially of a special and excessive development of the nematophores, so
that the graptolite not only in its trophosorae but also in its gonosome would thus present us
with an instance of the great development of the nematophoral system at the expense of the
hydranthal.^

' Hall notices a case {loc. cit., p. 33, pi. b, fig. 9) which he regards as one in wliich the appendages
are present in a graptolite which still retains its denticles. Tiiis, however, is by no means a well-marked
instance, and one might be permitted to doubt the identity of the structures here figured with the
appendages previously described by him.

' This view of the morphology of the corbulse seems placed beyond doubt by their formation in an
undescribed Aglaophenia from the deep-sea dredgings of the United States Coast Survey. The leaflets
which form the walls of the large and beautiful open corbulie of this hydroid are mainly composed of
the greatly enlarged and transformed nematophores which in the unaltered ramulus lie in front of the
hydrothecse. The hydrotheca; of the parts which become transformed into corbul» are not here actually
suppressed, but remain of somewhat smaller size, affording the clue to the morphology of the entire organ ;
and it can be plainly seen that it is the mesial nematophore of each of these arrested hydrothecae, which

182 DISTRIBUTION.

If these views be accepted we shall have nearly the entire graptolite in those instances
in which the appendages of Ilall have been noticed converted into a corbula, a state of things
which naturally follows from the simple unbranched form of the fossil. The graptolite has, in
fact, become greatly changed in form, and modified for a special reproductive function in a
way which reminds us of the so-called fertile fronds of certain ferns as distinguished from the
so-called sterile fronds.

It is true that the great rarity of these peculiarly modified graptolites is opposed to what
we know of living hydroids, for among these we are not acquainted with a single trophosome
which we are not justified in believing destined at some period of its life to develope a gono-
some. A case, however, bearing some analogy to those of the graptolites would be afforded
by fossil ferns, for we know how rare a thing it is to find among the vast midtitudes of indi-
viduals with which the coal-measures abound specimens bearing fructification.

While the graptohtes would thus seem to contrast with living hydroids in theii' rarely
developing a gonosome, it is interesting to see them contrasting also in another respect,
namely, in their free if not floating habit. And here we are reminded of the gulf-weed of the
Sargasso Sea, for throughout the thousands of square miles over which the floating meadows
of this remarkable plant extends no one has yet succeeded in finding a single specimen in
fructification, though the fructification of closely allied species, which grow attached to rocks like
ordinai'y seaweeds and like the rooted trophosomes of the hydroids, is w^ell known.

Certain bodies found associated with graptolites in the Silurian shales of Dumfriesshire have
been described by Dr. Nicholson, who regards them as the " ovarian vesicles" of the graptolites,
and as proving the hydroid nature of these fossils.' He describes them as " oval, bell-shaped,
pyriform, or rounded, provided with a mucro at one extremity, and surrounded entirely by a fili-
ibrm border, resembling in texture the axis of a graptolite." They attain a length of nearly half
an inch.

He has found them not only free, but in many cases attached to the graptolite, not, however,
to any constant point, for some spring " from the common canal, others from the apex of a
cellule, and others from the under surface of a cellule, the last two modes being the most
frequent."

The largest of these capsules which he has seen attached did not measure more than a
tenth of an inch in diameter, and Dr. Nicholson believes that at this stage they become detached.

has become enormously developed and flattened out so as to form the leaflet of the corbula-walls, while
at tlie same time it becomes complicated by carrying along one edge a row of small tooth-like
nematophores, as in the corbula-leaflet of Aylaophenia pluma, &c. The hydrothecse, with their
neniatophores, which in the untransformed ramnlus constitute a single seiies along the front of the
raiiiulus, are, in order to form the walls of the corbula, thrown alternately from side to side.

In the description of the corbula of Aglaoplienia pluma given above (p. 61), I regarded the
gonangia as representing the hydrotheca; of the untransformed branch, and as taking the place of these
hydrothecffi in the coibnla. I am now compelled to modify this view, and to regard the gonangia as
independent structures, just as in those plumularidaus which are destitute of corbulse. If any represen-
tative of the hydrolhecic had existed it would have been found, not ou the floor of the corbula, but in
its walls.

^ Nicholson, in ' Geological ^Magazine,' vol. iv, ISC", p. 259, pi. ii.

DISTRIBUTION IN TIME. 183

and then attain the large size he has observed in the specimens found free in the shale, for he has
there found tlieiu in all stages of growth, from small rounded bodies, not larger than a pin's head,
to bodies of nearly half an ineh in length.

Whatever these bodies may be, it is plain that Dr. Nicholson's account of them is irre-
concilable with the supposition that they represent either the gonangia or the gonophores of a
hydroid ; for, apart from their supposed development after detachment from the colony, their
origin fi-Qui the walls of the denticle is alone decisive on this point. Indeed, their connection
with the graptolite appears to be purely accidental.

Ilall has called attention to the occurrence in the same beds which contain the graptolilcs
of minute free bodies, which he regards as the young or "germs" of the graptolites.' In their
earliest form they would appear to consist of a little chitinous oblong sac traversed longitudinally
by a slender chitinous filament, which is continued for a little way at both ends beyond the sac,
while at one end it is accompanied by two minute lateral spine-like processes.

This early form has been traced through more advanced stages, in which it has been seen to
become more and more elongated, to develop denticles along its length, and finally to attain a
form in all essential points identical with that of an adult graptolite.

Others slightly differing in shape from those described by Hall have been also obtained.
Indeed, these young graptolites— for there is no doubt that Hall is right in his interpretation of
their nature — are now well known. They are by no means uncommon in graptolitic shales, in
some examples of which I have seen them abounding in coimtless multitudes.

Hall believes that he finds evidence of their having been contained within the so-called
reproductive vesicles of the graptolite. From his account of their relation to these, however, I
can recognise nothing but accidental proximity ; while if we admit that he has grounds for this
belief, we should then have, in the advancement of the embryo to a stage in which it has become
covered by a chitinous perisarc previous to liberation, a state of things quite at variance with all
we know of the reproductive phenomena of living hydroids.

But little remains to be said regarding other views which have been from time to time
advanced as to the affinities of graptolites. These fossils have been compared to the chambered
shell of a cephalopod, a view, however, which in the present state of our knowledge of them need
not now detain us for a moment. They have also been compared to the selerobasic Actinozoa,
and their nearest living allies have been sought for in Virgularia and Fennatula. It is plain,
however, that the resemblance here is of the most superficial kind, and that the internal solid
smooth axis of a Virgularia or a Pennatula has nothing in common with the membranous test
of the graptolite, whose axial canal and tubular offsets point to totally different affinities.

They have also been compared to the internal solid skeleton of larval echinoderms, an
approximation which has nothing in its favour, and which the occurrence of irregularly-branched
graptolites which have been continuously followed in the shale even for several feet will alone
render impossible.

Their alleged polyzoal affinities have much more claim on our acceptance. Indeed, were it
not for the discovery of the graptolite gonosome (corbulas ?) we should have nearly as much to
say for this view as for that which would refer them to the IIvdroida, more especially as the

' Ilall, op. cit., p. 33, pi. B, fii;s. 12 — 19.

184

DISTRIBUTION.

discovery of BliaMoplcurn renders us acquainted witli a polyzoon in whose test is developed a
cbitinous rod in almost all respects like that of the graptolites.^

The graptolites are exclusively of Silurian age. They occur first in Lower Silurian beds,
and unless we include among them the anomalous genus I)cndro(/raptus, which has been obtained
from the Potsdam formation of North America — the probable equivalent of our Lingida Flags —
their lower limit will be the summit of the Lingula Flags, from which they range into the Ludlow
of the Upper Silurian.^

The diprionidian, or double-rowed forms occur in great abundance in the Lower Silurian
Rocks, where they are associated with monoprionidian or single-rowed forms. Graptolites
become much more rare in the Upper Silurian, where they are represented exclusively by some
monoprionidian species.^

The following table will show the geological formations which have afforded evidence of the
former existence of the Hydroida.

LOWEE
SlLUalAN.

Upper

SlLDEIAN.

JUEASSIC.

Ceetaceods.

Miocene.

Tliocene. Pleistocene.

LrviNG.

Gymnoblastic tro-

phosome

(Hydractinia) .

X

X X

X

Calyptoblastic tro-

phosorae

(Sertularella) .

X?

X

Hydroid Medusae

(Trachyiiemidse

and jEginidse)

X

X

Monoprionidian

Graptolites .

X

X

1

Diprionidian

Graptolites*

X

^ The comparison of the rod of Rhabdopleura to tliat of a graptolite has already been made by
Dr. Nicholson (' Manual of Zoology'), though he adopts the more generally accepted view which sees
in the denticles of a graptolite the hydrothecc'e of a hydroid.

^ The curious Fenestella-Wke genus Bictyonema has been referred to the Grapiolitida by Hall.
It ranges from the Lower Silurian into the Middle Devonian ; but whatever may be the nature of this
remarliable fossil, it is almost certainly not a graptolite. Schiraper (' Traite de Paleontologie vegetale ')
places Dicii/onema along with Oldhamia among the nullipores,

^ If Reiiolites, Bar., be a true graptolite we must then carry the Diprionidian forms into the
Upper Silurian, for this genus occurs in the Llandovery and Upper Wenlock formations. RetioH/es,
however, is an exceedingly anomalous form, and departs in so many respects from the typical Grapto-
lites as to render the propriety of associating it with those fossils extremely doubtful in the present
state of our knowledge.

* See preceding note.

DISTRIBUTION IN TIME. 185

This table will, of course, not be regarded as representing the aetual distribution in time of
the IIydkoida. It is only an expression of the prineipal facts in the present state of our
knowledge of liydroid pala.'ontology, and it will be at once seen how very few these are. From
by far the greater number of formations we have no evidence of the existence of hydroid
life, while, if we except the graptolites, the examples which have come to us from any formation
whatever are very few, the vast majority of hydroid families being as yet unable to claim any
fossil representative. This, however, we can scarcely hesitate to refer to the defects of the record
rather than to the absence of hydroid life in past epochs of the globe. ^

' xVfter the above pages on the palreontological distriliution of the Hydhoida Lad been printed I
became acquainted with 11. Richter's paper " Aus den Thiiriugischen Schiefergebirge," published in the
' Zeitschrift der Deutschen Geologischen Gesellschaft/ xxiii Band, 1 Heft, 1871.

The author here describes the graptolites of the Upper Silurian strata of Thuriugia, but the paper
is chiefly occupied with a dissertation on the structure and affinities of graptolites in general. It
contains some valuable original observations, and, as the views advanced differ in certain points from
those of other German and of the British and American observers, I have thought it necessary to refer
to them here.

He describes the test as composed of two laminae, an inner thick one — usually marked by oblique
striae — and an outer thinner one, the latter being itself composed of two thin lamellae, which repeat
the striated condition of the former when such is present. If this be really the nature of the graptolite
test, the condition in which the Thuringian specimens have come down to us must be exceptionally
fitted for the detection of structure, as no evidence in favour of this view seems obtainable from any
British or American specimens.

He describes as an " organ of attachment " or "foot" the part named " basalstiick " by Geinitz,
and " radicle" by Hall, and he gives a figure of it as it has appeared to him in Graptolites priodon. It
is here, as in other monoprionidian species, described as a hollow elongated cone with rounded base, and
like the rest of the fossil composed of two membranous laminae. The common canal of the graptolite
springs from one side of it at a point not far removed from its thicker end, the pointed end of the foot
being, as a rule, thrown back on the dorsal side of the canal. Here, again, the views of the author
find no support in any British specimens I have examined.

Having noticed that certain monoprionidian graptolites occur sometimes curved towards the side
of the rod, or " dorsal " side, sometimes towards that of the denticles, or " ventral " side, he concludes
that these species must have possessed the faculty of spontaneously bending themselves from side to
side.

He has never witnessed anything which can be compared with the structures referred by the
American and British observers to the generative system of the graptolite, but he has seen in the
same rocks with the graptolites great quantities of spherical corpuscles of from 0"1 to 0'3 ram. in
diameter, and apparently surrounded by a double membrane. He does not offer any opinion as to the
import of these bodies, but along with them he finds also, in great abundance, slender conical bodies,
resembling in all respects, except that the pointed end was continued into an extremely fine whip-like
filament, that part of the graptolite which he had already distinguished as the " foot." He regards
these as the earliest stage of the graptolite. They appear to be the bodies referred to in the text as
first described by Hall under the designation of graptolite " germs ;" Richter believes that he has seen
the stem of the future graptolite originating as a bud from a point near their thicker end.

In such forms as Rastrites, in which the long distant spine-like processes are usually regarded as
the denticles with terminal orifice, he considers these processes as a mere armature of the true

186 DISTRIBUTION.

denticles, ivhich, according to llichter, scarcely project beyond the common canal, and have the orifice
situated iu the axil of the spine.

As to the mode of life of the graptolites, he believes that by the aid of the " foot," which he
regards as possessing a certain amount of spontaneous mobility, they were enabled to bore into the sea-
bed of clay or mud, and there maintain an erect position.

After comparing them with tlie Pennatulidce, Hi/droida, and Pobjzoa, he concludes that they are
most nearly related to these last, and refers to Leuckart as advocating in a letter to himself the polyzoal
nature of the graptolites. He adduces the rod in Bhabdop/eura as affording evidence of this view, and
(though the reserablauce seems to have escaped him) he might also have compared the peculiar striie
which he has described as usually present in the graptolites with those which exist in the adherent
portion of the tube in Rhabdopleura (see ' Mic. Journ.,' vol. xi, New Series, pi. viii, fig. 4), to which
it must be admitted they present a very striking similarity. I am unable, however, to find in Richter's
arguments any grounds for accepting the polyzoal affinities of graptolites, while in his comparison of
the two laminae of the graptolite test to the ectocyst and endocyst of a Pohjzoon it is plain that he has
formed an erroneous conception of the nature of the fleshy organized endocyst which constitutes the
essential part of the body- walls of a Pohjzoon.

The paper concludes with au enumeration of the graptolites which occur in the Upper Silurian
strata (Nereitenschichten und Tentaculitenschiefern) of Thuringia, with descriptions of the new species.
Among the forms described as new by far the most remarkable is one which carries three vertical rows
of cells upon the common canal. He makes it the type of a new genus, Triplogruptvs. The propriety,
however, of associating this remarkable fossil with the true graptolites appears to me very doubtful.

SUPPLEMENTARY.

There still remain for discussion in connection with the general zoology of the Hydroida
some subjects which, in strict sequence, ought to be included under the head of Morphology.
It has been deemed, however, more convenient to defer their consideration, and they are
accordingly here brought together in a separate section supplementary to the first part of the
Monograph.

CLASSIPICATION.

An important reform in zoological classification was introduced when Frey and Leuckart
separated as a distinct sub-kingdom, under the name of Ccelenterata, certain animal forms which
had been previously combined with the Ecliinodcrmata in the construction of the Cnvierian primary
group of the RADiArA.' In the establishment of the Sub-kingdom Coslenterata the German
zoologists recognised the importance of a fundamental type of organization in which the general
somatic cavity communicates freely with the outer world through the mouth.

The Ccelenterata admit of subdivision into two classes, in accordance with the presence or
absence of a stomach-sac differentiated from the general body-cavity though freely opening into it.
For these two classes the names of Hydrozoa and Actinozoa assigned to them by Huxley
may be accepted as sufficiently convenient. Each of these classes is further divisible into orders,
and one of the orders of the Hydrozoa is constituted by the Hydroida, the group to which
the present Monograph is devoted.

The characters which I regard as essentially distinguishing the Hydroida are expressed in
tlie following table, from which the relations of the Hydroida to the other ccelentcrate orders will
be at once apparent :

^ Frev uud Leuckart, " Beitrage zur Kenntniss wirbelloser Tliierc," 1847.

25

188

CLASSIFICATION.

r Generative ele-
ments discharg-
ing themselves
externally

9.

No stomach-sac^
difl'erentiated |
from the gene- ^ HYDROZOA
ral body-cavity j

Locomotion never
by bands of vi- ^
bratile lamellae

Generative ele-
ments discharg-
ing themselves
into the body-
cavity

Never with a hydriform"^
trophosome united with
the gonosome into a nata-
tory colony (woodcuts,
figs. 6-1-G7) J

Always with a hydriform ^

trophosome united with j

the gonosome into a nata- I

tory colony (woodcut, fig. |

I 68) J

f Generative elements in sym-~)
metrically disposed, longi-
tudinal, band-like projec-
tions of the inner surface
of the somatic cavity. Body
stalked, fitted for attach-
ment (woodcuts, figs, 71,
7-2)

Generative elements formed ^
in symmetrically disposed,
pouch-like dilatations of
the somatic cavityofabody ^
which is not stalked and
not fitted for attachment
(woodcuts, figs. fi9, 70)

A stomacl>-s.ic ^

differentiated I

from the gene- f

i ralbody-cavity J

ACTINOZOA

Locomotion always ^
by bauds of vi-
bratile lamellae
(woodcuts, figs.
73, 74)

Tentacles subuliform ; symmetry for
the most part hexameral (woodcuts,
figs, e-i, 03)

Tentacles leaf-shaped, with the raar-
'-- gins cirrated ; symmetry tetrameral

HVDKOIUA.

SiPHOXOPHOK.V.

LUCERN.\RI.E.

Ctexophora.

ZOANTHAKIA.

When the various living forms included under the order Hydroida are compared with one
another, it will be found that they present among them four principal modifications, and these afford
grounds for the subdivision of the living representatives of the order into four sub-orders; while, if we
include the graptolites, a fifth sub-order must be established for the reception of these extinct organisms.

Confininf' ourselves for the present to the living hydroids, we find that three out of the four
sub-orders under which they are included possess a hydriform trophosome, which, in the develop-
ment of the hydrosoma, intervenes between the ovum and the generative buds, whose assemblage
constitutes the gonosome. In the fourth sub-order there is no hydriform gonosome, and tlie ovum
becomes developed by direct metamorphosis into a medusifonn hydrosoma, to which is assigned
the performance of generative as well as nutritive functions.

CLASSIFICATION. 189

Of the three sub-orders vvliicli tlius ])ossess a hydriform tropliosome, there is one in which
no perisarc is ever excreted, while the hydrosoma is never permanently attached, and the nutritive
buds on attaining a certain degree of maturity disengage themselves from the parent stock and
continue their growth independently as free organisms. The designation of ELEUTnEiiOBLASTEA
may accordingly be assigned to the order tlius characterised.

In another sub-order a more or less obvious perisarc is always present, while the hydrosoma
is always attached, and the zooids of the trophosome never become separated as free, independently
developing organisms.' In this sub-order there is never developed either hydrotheca or gonan-
giura. So that both nutritive and generative buds are destitute of the characteristic receptacles
which afford protection to these parts in the following sub-order. The name of Gymnoblastea
suggests itself as a designation sufficiently expressive of this distinction.

In the last of the three sub-orders of living hydroids characterised by the possession of a
hydriform trophosome we have the same perisarcal covering of chitine, and the same permanent
attachment of the hydrosoma and nutritive buds which we meet with in the Gymnoblastea ; but
the hydranths are always protected by a hydrotheca, and the generative buds are always included
in a gouangiura. These characters afford the grounds for distinguishing a third sub-order, and
suggest for it the name of Calyptoblastea."

The last of the four sub-orders of the living IIydroida is distinguished by the absence of a
hydriform trophosome, the ovum becoming developed through direct metamorphosis into a medu-
siform body, just as in the other orders it is developed into a hydriform body. To the sub-order
characterised by this very exceptional condition I shall assign the name of Monopsea.^

' The detachment of the hydranths in certain Tubulariae (see above, p. 69) indicates the com-
mencemeut of decadence and death in these bodies, which never undergo further growth or development
after their separation from the parent stem. I take it for granted, also, that the free hydranths
described by Stimpson and by M'Crady under the names of AcavUs and Nemopsis have been detached
in the same way from some fixed tubularian stem. None of these cases, therefore, can be confoiuidcd
with the separation of the buds in the Eleutheroblastea.

^ The chitinous covering, which often invests the gonophore in tlie gymnoblastic hydroids, and
which, in some cases, as in Cordylophora lacustris, attains considerable thickness, must not be con-
founded with a gonaugiunn. The true gonangium is always developed round a blastostyle, though the
latter may occasionally be forced back, and become atrophied by tlie pressure of the growing gonophore.

So also the thin expansion of the perisarc, which may be seen extending over the base of the
hydranth in certain Gymnoblastea, as in Cordylophora lacustris, and more especially in various species
of BougainvUlia, is entirely different from a hydrotheca. The hydranth is never to any extent
retractable within it, as it is within the true liydrotheca, and during the contraction of the hydranth
this chitinous sheath merely adapts itself to the contraction, being then thrown into more or less well
defined transverse folds.

' I am not disposed to regard a non-.sexual trophosome as necessarily absent from the jMonopsea
any more than from the other three sub-orders. As yet none but Geryonidan and yEgiuidan medusae
have been proved to possess the characters on which the sub-order Monopsea has been founded ; for
the case of Lizzia, which has been adduced as an instance of direct development of the medusa from the
egg, is probably based on some error of ol)servatioa (see above, p. 100). Now, in both Geryonidan
and .^ginidan medusae the generative elements are produced in sac-like offsets of the radiating canals,
and reasons have already been given for regarding these generative sacs — at least in the Geryonidan
forms — as true zooids ; and the medusic from wliose radiating canals they are emitted would then belong

190 CLASSIFICATION.

All the living Hydroida are thus either destitute of a ehitiiious perisarc — Eleutheroblastea
nnd MoNOPSEA — or else are more or less extensively invested by this protective covering — Grjixo-
BLASTEA and Calyptoblastea. In no case, however, does there exist the additional support of a
solid chitinous rod. This is found in the extinct group of the graptolites, whose tubular offsets,
moreover, we have already compared, not to hydrothecoD, lint to nematophores.

Though w(; are still entirely ignorant of many important points in the structure of grapto-
lites, and know nothing beyond mere surmise of the functions of their parts, or of the phenomena
of their lives, we find in them nevertheless a group of characters which — accepting the view
advocated above of their hydroid atfinities — necessitates the establishment for them of a separate
sub-order of IIyduoida. And even though we may regard the nature here attributed to the
tubular offsets or denticles as too hypothetical to form the basis of such a separation, we shall
have in the solid rod alone a character sufficient for this purpose.^ The name of Rhabdophora
may be assigned to the sub-order th\is constituted.

Among the leading groups into which the Hydroida are thus divisible the sub-order of the
Eleutheroblastea is that in which we meet with the greatest morphological uniformity, the
entire group being represented only by the single genus Ilijdra.

The MoNOPSEA are more diversified, for we find among them the various types of the
Gcryonidan and Jj]ginidan medusae. Our knowledge, however, of these medusae is still imperfect ;
and though much light has recently been thrown on them, especially by the researches of Haeckel,
who has shown close and hitherto unsuspected relations between the Geryonidan and ..^ginidan
forms, we are by no means sure that they are all developed on the same plan, and the group
MoNOPSEA can scarcely be yet regarded as possessing more than a provisional value.

In the Gymnoblastea there is still more morphological diversity, this sub-order being repre-
sented by numerous fauiilies and genera.

In the Calyptoblastea the special morphological modifications are greater than in any of the
others, with the exception, possibly, of the RHABnopHORA. In the Calyptoblastea, indeed,
these modifications necessitate a primary division of this sub-order into two subordinate gi'oups,
each of which includes families and genera." The characters here assumed as the basis of hydroid
classification may be tabulated in the following scheme :

to the properly non-sexual form or " blastocherae." If this view be correct the Geryonidans must be
regarded as composed of a hydrosoma, consisting, as in other hydroids, of a trophosome and a gonosome,
but in which the trophosome is raedusiform instead of being, as in all the other sub-orders, hydriform.

It is only since the earlier portion of the present Monograph was printed that the development of
the medusa from the egg without the intervention of a hydriform trophosome has been established on
sufficient evidence. Certain views, founded on the absence of evidence in favour of this phenomenon,
have been there expressed, and must accordingly be now received with some modification.

^ Though the solid rod is present in all the typical graptobtes, there are some very aberrant forms
[Retiolites) in which it has not yet been demonstrated with certainty. It is a matter for consideration
whether these do not constitute a group witii affinities pointing in quite a different direction from
those of the true graptolites.

" Hincks (' Hist. Brit. Zooph.') divides the HynRoiDA into three suborders, to which he assigns
the names of Gvunochroa, Athecata, and Thecophora. These groups are equivalent respectively to
the Eleutheroblastea, Gymnoblastea, and Calyptobl.\stea of the present Monograph. Had not
tliese last names been used by me in the earlier part of the ^Monograph, before the publication of Mr.
Hiiicks's work, I should have here hesitated to add to the existing heavy nomenclature of zoology.

CLASSIFICATION.

191

. a '^

c3 -^ ^

-2 |i

a p

"1 -^

a. w

t o S i 5

'5 E

E = S
^ o .5

vaioaaAH

192

HOMOLOGIES.

HOMOLOGICAL RELATIONS OF THE CCELENTERATA.

Independently of the general agreement which necessitates the association of the Hijdra,
Actinia, and other Coelenterate animals into one primary group of the animal kingdom, we
must also expect a special morphological correspondence between the various forms of animals
thus associated. In other words, a homological agreement ought to be determinable between
the parts of animals included in any one subordinate section of the Ccelenterata with the parts
of animals included in any other.

Fig. G2.

Diagramatic longitudinal s ection ofActinra.

a. Radiating inter=eptal space; a', ten-
tacle; h, diflerentiated stomach-sac; 6',
somatic cavity; c, aperture in distal end of
radiating lamella; d, genetalia borne by
radiating lamellte.

In this figure, as well as in figs. 64, 66,
69, 71, and 73, a bristle is represented as
having been passed from the main canity
of the body into a gastrovascular canal or
its homologue.

Fig. 63.

Biagramntic transverse section
of Actinia.

a, o, Interseptal spaces; I,
differentiated stomacli-sac.

A comparison of the two primary sections of the Ccelenterata {Actinozoa and Hyurozoa),
and of the various orders of tliese with one another, will show that such an agreement really
exists, and that it is possible, by easily luiderstood and thoroughly consistent modifications, to
convert each special type into any of the others.

With the view of rendering apparent these relations, we shall compare an Actinozoon
{Actinia, woodcuts, figs. 02 and 63) with a Hydrozoon {Hydra, woodcuts, figs. 64 and 05), and
shall further compare with one another the various orders of the Hydrozoa.

Agassiz was, I believe, the first to compare the radiating chambers which, in an Actinozoon,
intervene between the stomach-sac and the outer walls, with the radiating canals of a medusa.'
He seems to have thus struck upon the true homologies of these parts ; but when he maintains
further that the differentiated stomach of an Actinozoon is only the proboscis (hypostome) of a

Contr. Nat. Hist. U.S.,' vol. iv, p. 377.

HOMOLOGIES.

193

Hydrozoon inverted into its body-cavity, he suggests a view wliicli is insufTicient to explain tlie
actual structure, for the radiating lamellae still remain unaccounted for.

We possess very few observations on the early stages in the development of Actinia, but it
seems to be generally taken for granted that the first appearance of the stomach-sac is to be
explained by sujjposiiig a l)ending inwards of the margin of the oral aperture, and if this be
admitted the conception of Agassiz will so far be borne out.' But then the septa which are found

Fig. G4.

Fig. 65.

Diagramatic lon*;itudinal section of

Hydra,
a. Tentacles; i, hypostome; />', so-
matic cavity.

© o©

Dia^rramatic transverse section of
Kijdra through hypostome and
tentacles,
a, Tentacles ; i, hypostome.

surrounding the stomach-sac must in this case be formed as independent structiu-es between the
outer walls and the inverted mouth-margin, and the intervening chambers must have an entirely
different origin from that of the radiating canals of a medusa, and can hardly be regarded as their
strict homologues.

This introversion of the mouth-margin, however, is not a necessary interpretation of the
successive forms presented by the embryo in its early stages, for an exactly similar appearance

' See Cobbold ('Ann. Nat. Hist.,' 2nd series, vol. xi), and more especially Kowalevsky (Nach-
richten der K. Gesellschaft der Wissensch. zu Gottingen,' 1868, s. 157). Kowalevsky gives no figure,
and Lis communication is in such a condensed form, stating the results of his observations rather than
the steps by which these results are obtained, that it is not always easy to discover his exact meaning,
especially with regard to the mode of formation of the septa and the radiating chambers. The forma-
tion of the stomach-sac, liowever, by the turning in of the mouth-margin is clearly maintained.

It may be here mentioned that Kowalevsky denies the correctness of parallelising the common
cavity of the Cmlenterata with the true body-cavity of other groups of the animal kingdom, and main-
tains that the so-called body-cavity of the Cwknterata is rather the homologue of the intestine of other
animals, being, according to him, a secondary formation supervening on that of the proper body-cavity
in which it has been formed by a process of introversion of the walls, the proper body-cavity
itself becoming more or less obliterated.

A view agreeing in most respects with that of Kowalevsky is also advocated by Semper {' Reise
im Archipel der Philippinen,' Zweiter Theil, Erster Band, p. 131), who maintains that the endoderm o

194

HOMOLOGIES.

would he produced hy tlie prolongation of the oral margin outwardly, accompanied by an
extension around it of the general body-cavity.'

According to the latter interpretation, the so-called stomach-sac would represent the hypo-
stome of Hydra uninverted, and surrounded by a continuation of the body-cavity. But this distal
prolongation of the body-cavity is farther seen to be divided by radiating septa into chambers
which open freely into the simple primitive body-cavity below. The exact meaning of these
septa and of their intervening chambers remains for determination, and is at once suggested by
a comparison of Actinia with Hydra.

In order to form a correct notion of the horaological relations between an Actinia and
a Hydra, Vie have to imagine tlie tentacles of a Hydra (woodcuts, figs. 04 and G5, a) for a greater
or less extent connate with the sides of the hypostome {Ij) and with one another. The hypostome

Fig. 66.

Diagramatic loiia;itudliial section of a
H^dro'id Medusa.

a, Radiating canals; a\ marginal ten-
tacles; h, manubrium; h\ atrium; c, lumen
of circular canal ; d, geuf rative elements ;
r, atrial region of umbrella ; r', manubrial
region of umbrella ; v, velum.

Fig. 67.

Dia^jramatlc transverse section
of a Hydroid Medusa througll
the manubrial region of the
umbrella.

a. Radiating canals; &, manu-
brium ; (?, generative elements;
r', manubrial region of umbrella.

of the Hydra, while retaining its normal position, will thus become the stomach-sac of the Actinia
(woodcuts, figs. 62 and 63, U), and this will necessarily become connected with the outer walls by
a series of radiating laniellc-c— the connate tentacle-walls — separated from one another by radiating

tlie Calcnterata is formed by an introversion of the ectoderm. He asserts that the Coslcnterata have no
body-cavity, but only an analogue of it in a connective tissue lying between the endoderm and
ectoderm, and to which he assigns tlie name of Coenenchyma. He maintains that all the canals
which in every direction permeate the zooids and the coenosarc, are only appendages of the digestive
cavity, and that, consequently, the generally received view, which represents the alimentary canal of
the Ccelenterata as opening into the body-cavity, must be given up.

That the views of Kowalevsky and Semper do not hold good for all the Ccelenterata there can be
no do\ibt. They certainly are not applicable to the Hydroida,

^ The account of the development of the Red Coral given by Professor Lacaze Duthiers (' Hist.
Nat. du Corail ') bears out this interpretation, rather than that which would sec in the stomach-sac
an inverted hypostome.

HOMOLOGIES.

195

Fig. GS.

chambers («)— the cavities of the tentacles ; while such portions of tlie tentacles of the ITjjdra
us still continue free will be represented by a single circle of the tentacles of the Actinia (woodcut,
fig. 02, a).

Having thus established a fundamental identity between the regions of an Actinia and of
a Hydra, there will be no difficulty in recognising the relations between an Actinia and a
hydroid medusa (woodcuts, figs GG and G7) ; for as we have attempted to prove in a former
part of this Monograph (sec p. 40) the tentacles of a Hydra are represented by the radiating
canals and those extensions of them which form the primary marginal tentacles of the medusa.

The distal ends of the radiating lamellae in Actinia are perforated each by an opening
through which the radiating chambers communicate with one another (woodcut, fig. 62, c).
Agassiz has compared these openings to the circular canal of a medusa, and I believe that in this
view he has correctly expressed the relations in question.

If we further add that the generative apparatus (woodcut, fig. C2, d) is borne by the
radiating partitions, we shall have all the leading points in the morphology of an Actinozoon.

Acomparison of thevarious orders of the HTDROZOAwith one another will result in the detection
of close homological cori'espondencies, and will throw important light on the morphology of each.

Between a siphonophore and a hydroid the homology is so obvious as to be instantly
recognisable. The siphonophore (woodcut, fig. 68) as well as the hydroid presents us with a
colony of zooids kept in organic union with one another by
means of a common connecting basis or ca?nosarc ; but this
ccenosarc, instead of being fixed, as in the IIydkoida, is in the
SiPHONOPHORA invariably free and provided with a special
apparatus for natation.

In consequence of the great extent to which heteromor-
phism is carried among the zooids composing a siphonophoral
colony, we can scarcely institute a satisfactory comparison between
the two orders without determining the homologies of each kind
of zooid in the siphonophore. Beginning with the polypites or
alimentaiy zooids (e) of the siphonophore, and comparing these
with the hydranths of a hydroid, we shall find the two forms to
agree in almost every point, except in the number and position
of the tentacles, which in the siphonophore are reduced to a single
one (/), springing, in all the typical siphonophores, from the base
or proximal end of the polypite. The branched condition of the
tentacle in the siphonophore is in no respect inconsistent with
this comparison ; and even if it were necessary to find a parallel
to it among the IItdroida, wc should have this in the branching
tentacles of Cladoconjne.

The hydrocysts {g) of the siphonophore are plainly
arrested polypites in which the mouth has never become deve-
loped.

Again, the generative zooids (/) are exactly paralleled by
those of the Htdroida, and are, like them, referable to two
types, expressed in the Hydroida by the phanerocodouic and

Diagram of a Siphonophore.

e, Polypite;/, tentacle of polypite;
/', bruuches given off by the tentacle ;
y, liydrocyst ; h, tentacle of hydrocyst ;
i, frenerative zooid representing tlie
jiliiinerocodonic gonophore of a hydroid ;
k, k, nectocalices ; I, bract ; m, m, cceno-
sarc ; n, pneuuiatocyst.

26

196

HOMOLOGIES.

the adclocodonic gonophores, the situation of the generative elements being precisely similar in
the two orders. So also the nectocalices or locomotor zooids {k, k) are essentially hydroid medusae
with specially developed umbrella, but with the manubrium suppressed and the somatic cavity
reduced to the atrium, from which spring radiating canals which, exactly as in the hydroid
medusae, open round the margin into a cu'cular canal.

The bracts or hydrophyleia (/) of the siphonophore are essentially cscal offsets from the
common canal of the coenosarc, but with the ectoderm greatly developed and modified, as in the
umbrella of a medusa, so as to fit them to become organs of protection for the other zooids
They have thus essentially the same morphological foundation as the nectocalices, but with a
different functional destination diverge widely from these, and constitute an a])paratus of protec-
tion instead of locomotion.

All these zooids are kept in union with one another by a coenosarc {m, m), which in all the
typical SiPHONOPHORA is filiform, with an axial canal in free communication with the cavity of
each of its appended zooids, thus corresponding essentially with the filiform tubular ccenosarc of a
hydroid colony ; while in the somewhat aberrant forms with fusiform or discoidal coenosarc
{PliysalidcB, VeleUidce) an obvious comparison is suggested with the appressed, expanded coenosarc
of Hydractinia.

From the hydroid ccenosarc, indeed, that of the Siphonophora mainly differs in the absence

Fig. 69.

Fig. 70.

Diagramatic longitudinal section of a Discophorous
Medusa.

a. Radiating canal; h, manubrium; J', somatic cavity;
d, generative pouches ; o, o, o, o, pillar-like offsets from the
oral side of the umbrella ; z, tentacula-like appendages
of the inner surface of the somatic cavity.

Diagramatic transverse section of a
Discophorous Medusa,
a, a, a. Radiating canals ; 6, manu-
brium; rf, rf, generative pouches; o, o, um-
brello-manubrial pillars.

of an external chitinous sheath and in its free mode of existence, the siphonophore dwelling at
large in the open sea, through which, in the great majority of the order, it is propelled by the
contractions of the nectocalices. In the section PhysophoridcB the proximal extremity of the
ca3nosarc, instead of forming, as in the Hydroida, a hydrorhiza for fixation, is modified by an
inversion of its walls, so as to constitute an air -filled chamber or pneumatocyst («), which acts as
a float.^

^ In the above comparison of the Siphgnophoka with the Hyduoida, I have adopted generally
the views of Huxley, who was the first to bring out in a complete form the liomological relations

HOMOLOGIES.

197

Continuing to take the Hydroida as a standard of comparison, the other liydrozoal orders
may be now contrasted with them. If tlie atrium (woodcut, fig. OG, b'), or tliat portion of tlic
somatic cavity which hes at the base of the manubrium in a hydroid medusa, be expanded
laterally (woodcut, fig. 09, b'), and the ectoderm of its floor be projected along four or eight
symmetrically disposed radiating lines into as many thick pillars (woodcuts, figs. 69 and 70 o, o)
which converge towards the axis and there meet the manubrial extension of the cavity, while the
thin intervening portions of the floor between the pillars become developed into generative
pouches ((/), and the velum or perforated diaphragm which stretches across the codonostome in
the hydroid disappears, we shall have the hydroid medusa converted, in the more essential
jjoints of its structure, into a discophorous medusa (woodcuts, figs. 69 and 70).

Again, a Li/cernaria (woodcuts, figs. 71 and 72) may be conceived of by imagining a Hydra
with four tentacles to have these tentacles expanded laterally until their sides meet and coalesce,
while the hypostome {b) still remains free, the proximal portion of the body continuing to form
a peduncle of attachment {p), and generative sacs {d) becoming developed on each side of the
partitions formed by the coalescent sides of the tentacles.

Fig. 71.

Dli

ramiitic loni^itudinal sectiou of Lucernarla.

a, Ciroumoral disc ; a', marginal tentacle ; i, hypostome ; b\
somatic cavity ; c, aperture by wliich the chambers of the circum-
oral disc communicate with one another across the distal eud
of the partition ; d, generative bands ; p, peduncle ; z, teutacula-
like processes of the inner surface of the somatic carity.

Fig. 72.

Diaijramatic transverse section of a
Lucernaria across the oircuiuoral disc
and hypostome.

a, a, Chambers of the disc ; b, hypo-
stome ; d, generative bands.

Lastly, the Ctenophora (woodcuts, figs. 73 and 74) admit of an obvious comparison with a
hydroid medusa. In order to understand this we must keep in mind the presence in the hydroid
medusa of an atrial segment of the somatic cavity. This is formed by that portion of the somatic
cavity which is immersed in the substance of the umbrella at the base of the manubrium, and
from which the radiating canals proceed (see woodcut, fig. 66, b'). The hydroid medusa thus
admits of a division by a transverse plane into two regions — an alrini region (r), which corre-
sponds to the solid summit of the umbrella with the parts therein contained, and a mmmhriaJ

between the two groups. In referring to the parts of the Siphonophora I have also employed the
terminology proposed by Huxley for this order. See bis 'Oceanic Ilydrozoa,' page 8, &c.

198

HOMOLOGIES.

rcnion (/),wliicli corresponds to the manubrium with that portion of the umbrella which, with its
associated structures, is projected round the manubrium in the form of a bell.

Now, in ?kBeroe (woodcuts, figs. 73 and 74) the manubrial region is never developed, and the
body is represented by the atrial region alone. From the atrium (3', b') contained within this
ret^ion two radiating canals {a, a) are given off. These immediately divide and subdivide so as to
become ultimately eight, which are, moreover, united at their distal extremities by a circular
canal, which corresponds to that of the medusa, though here thrown back by the non-development
of the manubrial region of the umbrella. Besides the eight longitudinal canals (.?•, x) into which
the two radiating canals ultimately subdivide, these two canals give off, each immediately after its
orio-in, an accessory canal [x, x) which runs without division close to the main body-cavity
towards the oral orifice, and opens like the others into the circular canal.

The generative sacs {d, d') are developed as diverticula along the course of the radiating
canals, whence they extend into the gelatinous substance of the body.

Fig. 73.

Fig. 74.

Diagramatic longitudinal section of
Beroe in a plane at right angles to
that of the compressed somatic cavity.
In order to give a sufficiently com-
prehensive view of the structure, a
few parts are here represented which
are in reality somewhat removed from
the plane of the section.
a, a. Transverse portion of the radiat-
ing canal system, two of the primary
tranches being shown as if cut off close
to their origin ; x, x, meridional portion
of this system ; x\ x\ deep or accessory
canals, their distal terminations in the
circular canal cut off ; c, lumen of cir-
cular canal ; 6', h' , somatic cavity ; t, one
of the aboral outlets of the somatic
cavity ; *, valve-like processes of the
inner surface of the somatic cavity;
d, d', generative sacs, male and female.

Diagramatic transverse section of
Beroe.

b' Somatic cavity; x, x, meridional
portion of the radiating canal system ;
x' x', deep or accessory canals; d, d\
generative sacs, male and female; y,
vibratile lamella;.

Leuckart^ insisted on the association of the Ctenophora with the Actinozoa rather than
with the Hydrozoa, and the same view of their affinities has been advocated by Huxley.
According to this conception of ctenophoral homologies, the ctenophora must be provided with a
stomach-sac differentiated, as in the actinozoon, from the general body-cavity. Now, though the

^ Frey und Leuckart, * Beitriige/ p. 35.

Huxley, " Lectures," in ' Med. Times.*

HOMOLOGIES. 199

somatic cavity in Beroe suddenly diminishes in width towards the aboral end, and is there
provided with a pair of valve-Hke folds (s), so that the entire tract admits of being distinguished
into two regions, it is nevertheless as continuous as in Hydra.

The advocates of the actinozoal nature of the Ctenophora see in the canal-system of a Beroe
or a Cydippe the radiating chambers of an Actinia separated from one another by partitions of
relatively enormous thickness. I do not desire to dispute the correctness of this view ; we
have already compared a hydroid with an actinozoon, and have seen in the radiating canals of
a hydroid medusa the homologues of the radiating chambers of an Actinia ; so that even though
the Ctenophora be truly HyDROzoA, we must expect to find in them the same points of agree-
ment with the AcTiNOZOA which we have endeavoured to demonstrate for the other hydrozoal
orders.

Now, the fact of the radiating canals being widely separated from the axial cavity instead of
being adnate to it, is exactly the point which essentially distinguishes a hydrozoon from an
actinozoon, and the fact of the intervening space being in the Ctenophore obliterated by the
interposition of a voluminous gelatiniform mass does not alter this relation, for it is exactly wdiat
we find in the atrial region of an ordinary hydroid medusa, while it is distinctly expressed in the
gonopbore of Clavatella (see below, page 216), where the free or manubrial portion of the
umbrella is rudimental, and the whole gonophore, apart from the marginal tentacles, becomes
comparable to the atrial region of an ordinary hydroid medusa.

The two accessory canals of Beroe run, it is true, close upon the walls of the axial cavity until
they leave these to throw themselves into the circular canal ; but this fact cannot, in opposition to
the greatly preponderating hydrozoal features, be used as an argument for the actinozoal nature
of the Ctenophora.

These accessoiy canals are not represented in the hydroid, while the Beroe further differs from
the hydroid in the presence of the two short aboral canals by which the aboral end of its somatic
cavity communicates with the outer world {t), as well as in the disposition of its so-called nervous
system and sense organs, and in its characteristic bands of vibratile lamellge (y), all which features
are among the special characteristics of the order, and in no way justify the absorption of the
Ctenophora into the Actinozoa.

In this attempt to determine the true affinities of the Ctenophora, I have taken Beroe
instead of Cydippe or other ctenophorous genus as the subject of comparison, not only because
Beroe is a typical ctenophorous form and of comparatively simple structure, but because I have
myself made its anatomy and development a subject of special study.'

' " Ou the Structure and Development of Beroe," ' Proc. Koy. Soc. Ediub.,' 1862.

200 TERATOLOGY AND PATHOLOGY.

TERATOLOGY AND PATHOLOGY.

Under this head must be inchided certain phenomena which, notwithstanding their abnormal
character, cannot be omitted in any complete survey of the order, for they are by no means
without interest in tlieir bearing on questions connected with specific form, as well as in their
relation to the general morphology and physiology of the Htdroida.

Modifications remlting from Parasitism. — One of tlie most remarkable of the abnormal
conditions of the Hydroida is connected with a case of parasitism to which certain species
are subject. Specimens of Coryne pusilla and of Syncoryne exiinia may be occasionally met
with having some of their braifches, which under ordinary circumstances would have carried
hydranths, converted into piriform sacs, in whose thickened walls endoderm and ectoderm may
be still distinguished, while the whole is invested by a continuation of the external chitinous
perisarc of the colony (woodcut, fig. 75).

The sac {a, h) communicates freely with the somatic cavity of the hydroid, and admits into it
the general contents of this cavity. Li every case there is also included in it a living animal,
which, instead of having any direct relation with the hydroid, belongs to an entirely different and
far more elevated group of the animal kingdom.

The transformed branch of the hydroid has, in fact, become the abode of a pychnogonidan,
which lives in it at the expense of the fostering animal, and which may be followed through
various stages of its development, from the embryo as it leaves the egg to that stage in which
it has almost attained its adult form.^

Two stages of the development are represented in the accompanying figure (woodcut, fig. 75).
In the earlier (4) a pair of pincer-bearing cephalic appendages (" pates-machoires," Milne-Edwards) is
greatly developed. These appendages stand out free from the body, on which no true legs are at first
visible. A closer inspection, however, shows three pairs of legs bent upon themselves and closely
appressed to the body, and all included within a common external membrane. Offsets have already
begun to bulge out from the stomach ; three of these on each side have penetrated the bases of the
legs, while two others extend towards the bases of the cephalic appendages. The rostrum is as yet but
slightly developed.

In the more advanced stage (a) the three pairs of legs have freed themselves from the investing
membrane and have become fully developed, while the offsets from the stomach may be followed into
the penultimate joints. The anterior offsets have not advanced beyond the bases of the cephalic
appendages. The rudiments of the fourth pair of legs may be seen in the form of unjointed hollow
processes, one on either side of the rudimentary abdomen. Both the stomach and its oflsets are of a
deep red colour.

It is in this stage that the pychnogonidan leaves the hydroid, by making its way through
the walls of the sac. Its further development takes place during its free life in the surrounding

TERATOLOGY AND PATHOLOGY.

201

Tlie p3clinogonidaii which thus infests the Syncorync c.vimir/, and probably also that of other
Corynidec, appears to be the PhovichUiduim coccinetim of Johnston. There can, I think, be
little doubt that the eggs of the parent Phocivhi-
lid'uim arc swallowed by some of the hydranths

of the colony, and thus reach the somatic cavity, pj^ yj

from which they gain access to the interior of
a young bud, which, under ordinary circum-
stances, would have become a hydranth-bearing
branch, but which now becomes arrested by the
developing parasite, and changed into a recep-
tacle for its protection, in a way which strongly
reminds us of the production of galls in the
vegetable kingdom. In no case were more than
a single individual found tenanting the infested
branch.'

It is not alone, however, species of Coryne
or of Syncoryne which are thus subject to the
attacks of a parasitical pychnogonidan. Gegen-
baur, as we have just said (see note to preceding
paragraph) has noticed it in Eudendrium ramo-
sum, while Ilydfaclima eckinata may occasionally

' The occurrence of this remarkable form of
parasitism, which presents the unusual feature
of the parasite belonging to a higher group
than the animal infested, was first noticed by
me, in 185/, at the meeting of the British Associa-
tion for the Advancement of Science, held in that
year.

A similar case has since been described, in
considerable detail, by Mr. Hodge, who has found
Syncoryne eximia infested in this way, and has re-
cognised in the parasite the Phoxichilidhim coccineum
(see ' Ann. Nat. Hist.,' 3rd ser. vol. ix, p. 33, pi. iv
and v).

The first, however, to notice the liability of a
hydroid to become the abode of a parasitical
pychnogonidan was Gegenbaur, who observed
the hydranth.s of Eudendrium ramosum fre-
quently changed in form, by becoming greatly
enlarged, and by the reduction of their tentacles
to short tubercles. Here the transformed hydranth
was found to be tenanted by a brood of deve-
loping pychnogonidans (see Gegenbaur, ' Genera-
tionswechsel,' p. 38, note).

A portion of a colony of Syncortfne eximia infested by
Phoxichilidium coccineum.

a, A lumulusof the hydroid transformed into a sac by the
presence of the parasite, which has liere reached the stage at
which it is ready to escape into tlie surrounding water; A, a
ramulus, with the contained parasite at a much easier
stage ; c, a normal hydranth of the Syncoryne,

202

TERATOLOGY AND PATHOLOGY.

Tig. 7G.

be seen to be similarly affected. In this liydroid the infested hydranths become converted into
laro-e oval sacs, each enclosing numerous individuals of a pychnogonidan in various stages of
development. The parasite appears to be here of a different species from that of the Spicorpie,
but I have not follovt'ed it to a sufficiently mature stage to enable me to determine it.

The infested and transformed hydranths of llydractinia may be seen associated with other
perfectly normal ones in the same colony. Their transfonnation will be found in various stages
of completeness, some still possessing short stunted tentacles, while in others the tentacles will
have entirely disappeared. The parasite is here, as in the case already described, most likely
introduced in the state of egg by being swallowed cither by the infested hydranth or by some
other hydranth of the colony.

Dr. Strethill Wright was the first to notice the liability of Ilydradinia to the parasitical
attack of a pychnogonidan,^ and his statements are fully borne out by ray own observations.

The misshapen club-like branches which Van Beneden figm-es in his Eudendrium {Bouf/ain-
villia) ramosuni- originate most probably in a similar case of parasitism.

Other abnormal conditions. — Besides the changes of form dependent on parasitism there are
others by no means destitute of morphological and physiological interest.

While examining the medussB which had been thrown off from a colony of Spicori/ne jiiil-

chella in one of my jars, I was struck by
ob.serving two of these medusae united to
one another by a small space on the convex
surface of their umbrellas, at a short dis-
tance from the summit (woodcut, fig. 7G).
One of the united medusse was a little
smaller than the other; but otherwise they
were both equally developed, and presented
the ordinary condition of these zooids at the
time of their liberation from the trophosome.
The cavities of the two rmibrellse freely com-
municated with one another through the
surface of junction.

That neither of the medusa; thus so in-
timately united had been produced by a bud
from the other was evident ; for the original
point of union with the trophosome, as
well as the remains of the canal, by which
the cavity of the manubrium had at one
time communicated with the somatic cavity
of the trophosome, were still distinct in each,
while these facts are also opposed to the
view which would regard the twin raedusa3 as representing a single one in the process of self-
division.

Conjoined tuin nu'dusa-, developed from a specimen of
Si/ncort/ne j>ulchella.

' Proc. Roy. Plivsical Soc. Edinburgh,' November, 1861.
' I;'Eml)ryogeiiie des Tubulaires.'

TERATOLOGY AND PATHOLOGY. 203

The only explanation which it seems possible to suggest is, that in the twin medusae
we have a case of accidental adhesion contracted between two neighbouring buds while still con-
nected with the trophosome, this adhesion having been followed by a free communication between
the two umbrella cavities. I never met with more than a single example, and, whatever ex-
planation we may be disposed to offer as to its origin, it seems evident that it cannot be regarded
as otherwise than abnormal.

The other case to which I would here wish to draw attention consists in a change of form
observed also in medusas of Syncorijne ^ndchella, whicli had been thrown off from trophosomes con-
fined in my jars. Li none of these medustp, though they had remained under observation for
nearly a month, had any development of generative elements taken place, but they had all under-
gone a very remarkable change.

The commencement of this change might have been noticed a few days after their liberation.
The umbrella became everted, and then began to diminish in size, contracting from its margin
towards its summit, until in a few days it had almost entirely disappeared, being then merely
represented by a thick disc of a somewhat quadrangular form, which projected round the base of
the manubrium. Each of the four angles of this disc was continued into one of the marginal
tentacles, whose proximal end, following the contraction of the umljrclla, had been thus brought
upon a level with the base of the manubrium. The interior of the disc was occupied by a cavity
which communicated freely with that of the manubrium, and with that of each of the four tentacles
which extended from its angles. With the contraction of the umbrella the circular canal and
velum had disappeared, and the radiating canals were now represented solely by the short
channels by which the interior of the hollow disc communicated through the thickness of its walls
with the tubes of the tentacles. Neither tentacles nor manubrium had undergone any material
change ; the former retained their full power of extension and retraction, and the latter all its
original irritability — moving from side to side, lengthening and shortening itself, opening and
closing its mouth with at least as much vigour as before the disappearance of the umbrella. The
medusa in this condition reminded us strongly of the gonophore of Clmatella, though the degrada-
tion of the umbrella was more complete than in the latter. The medusa had, in fact, become
changed by a retrograde metamorphosis into a hydranth.

Changes had been noticed also by Dujardin in the medusa of his Syncoryne deci-
piens ; but he had not followed them beyond an eversion of the umbrella, which is pro-
bably the commencement of the changes residting in the disappearance of this part of the
structure.

Notwithstanding the very striking character of the changes now described, and their
resemblance to a normal metamorphosis, I cannot see in them anything more than a degradation
of structure resulting from imperfect nutrition — a mere forerunner of complete disintegration
and death. They are, however, most instructive in their bearing upon the homologies between
the medusa and the hydranth, and completely support the view that the radiating canals of the
medusae are the homologies of the channels by whicli the gastric cavity of the hydranth is con-
tinued through the thickness of its walls into the interior of the tentacles, which will then repre-
sent those marginal tentacles of the medusa which constitute the continuations of its radiating
canals.

Colonies of Ilydradinia echinata may occasionally be met with in which certain hydranths
have become bifurcated at some distance from the base, each branch of the bifui-cation carrying

27

204

TERATOLOGY AND PATHOLOGY.

Fig. 77.

a mouth and tentacles like those of the ordinary hydranths. Agassiz has described an occasional

bifurcation of the spiral zooids of his Hydractinia
polyclina, and Hincks has figured a bifurcating
zooid of Hydractinia ecJiinafa in which one Ijranch
of the bifurcation carries tentacles and possesses all
the characters of an ordinary hydroid, while the
other is indistinguishable from a blastostyle though
no gonophores are developed from it. I have
myself seen the spiral zooids of Ilydractinina
ecJdnata depart from their normal condition by
developing tentacles on their distal extremities.

Li Cordylopfiora lacudris an abnormal pheno-
menon, full of interest, may sometimes be seen.
This consists in the conversion of the spadi.x, after
the discharge of the generative products from the
gonophore, into a true hydranth (woodcut, fig. 77).
The spadix in this case, after it has performed its
function as a part of the gonosome, elongates itself
{b), differentiates an ectoderm, and developes upon

A portion of a colony of Cordylopjmra ?a««<W«, showing J^g extremity a mouth and teutacles.
tbe transformation of a spadix into a hydranth.
o, A normal hydranth ; 4, the spadix of a gonophore
transformed into a hydranth after the discharge of the
generative elements from the gonophore ; c, the remains
of the external thin chitinous investment of the gono-
phore ; (Z, main branch.

The conversion of the entire gonophore into a
hydi'anth in Mhizogeton firsiformis is described by
Agassiz,^ who regards it as a normal feature in the
hydroid. In this view, however, I cannot concur.

I am quite of opinion that the phenomenon which Agassiz regards as normal is not so, and that

it is one of the same class as that just described in Cordylophora.

Cent. Nat. Hist. U.S.,' vol. iv, p. 12G, pi. xx, fig. 22.

TUBULARIA INDIVISA. 205

ANATOMY OF SPECIAL FORMS.

The following are some special studies of hydroid anatomy which I have deemed it right
to bring together in this place, in order to render more complete the general exposition of the
Hydroida attempted in the present work. They have been selected so as to afford examples
of the more important morphological variations met with among the gymnoblastic forms — such,
at least, as I have had an opportunity of studying. Each includes the anatomy of a species
which may be taken as the type of some generic group.

TuBULARIA INDIVISA.

Plates XX and XXIII.

This beautiful and by no means uncommon hydroid, one of the largest and most interesting
of the entire order, presents clusters of unbranched stems which spring from a creeping, tubular,
branching and anastomosing stolon, attaining a height of many inches and a thickness of about a
tenth of an inch, and carrying each a hydranth on its summit.

Ilie Ihjdrophjton. — The entire hydrophyton is invested by a firm chitinous perisarc, in
the older parts of which layers of deposition may be detected. The coenosarc is closely embraced
by the perisarc except in that part of the stem which lies immediately below the hydranth,
thus differing from the majority of the Hydroida, in which the coenosarc for the greater part of
its course is separated by a considerable interval from the surrounding perisarc, with which its
connection is maintained only by irregular transverse processes of the ectoderm, which here and
there stretch across the interval from the one to the other. In the present species, however, as
well as in the Tubularice generally, it is only for a short distance below the hydranth that the
coenosarc has withdrawn itself from contact with the surrounding perisarc which here forms a
thin membranous sheath with large irregular annulations (PI. XX, fig. 2).

By means of a carefully made transverse section of the stem the structure of the coenosarc
may be rendered apparent. If the section be made in any part of the course of the stem excpet
immediately below the hydranth, a very remarkable stnicture will be brought into view
(PI. XXIII, fig. 7). The endoderui will be here seen to consist of two distinct portions, a peri-
pheral and an axial. The axial portion (d) consists of large cells with clear colourless contents,
and occupies as if with a sort of pith the entire axis of the section, which is cpiite destitute of an
axial somatic cavity. The peripheral portion (c) is composed of small spherical cells filled with
coloured granules. In this portion numerous empty spaces may be seen ; they are somewhat

20G ANATOMY OF SPECIAL FORMS.

wedge-shaped in section, and are arranged at ratlier regular intervals in a zone concentric with the
circumference of the stem. They are richly ciliated on their sides with very distinct, long, actively
vibrating cilia. They are of unequal size, one of them especially being in almost every instance
considerably larger than any of the others. They are simple lacuncB, and represent the transverse
section of a system of longitudinal canals which are excavated in the peripheral portion of the
endodenn, and which take the place of the axial somatic cavity of other hydroids. Their presence
is indicated even to the naked eye by longitudinal bands, which are visible through the trans-
parent perisarc in the uninjured condition of the stem.

By following these canals through the stem they will be seen to communicate occasionally
with one another by short transverse channels, while towards the summit of the stem they
coalesce into a common chamber which occupies the axis.

Surrounding the peripheral portion of the endoderm is the zone of ectoderm {b). It lies in
immediate contact with it, there being here no intervening zone of fibrillated tissue. In the
ectoderm no distinctly cellular structure can be demonstrated ; only some granules and nucleus-
like bodies may be seen scattered through a common plasma in which numerous thread-cells
are also imbedded — a structure, however, which indicates a true cellular composition at an
earlier period of development.

On the outside of the ectoderm and in close contact with it is the chitinous perisarc («),
showing no structure beyond that of concentric laminae of deposition.

When a living and uninjured specimen of Tubularia indivisa is examined under a moderate
magnifying power active currents may often be witnessed in the channels of the endoderm.
These take a course sometimes from the proximal towards the distal end of the stem, sometimes
from the distal towards the proximal, while two currents may often be seen running in opposite
directions in two diflferent channels at the same time. They are plainly under the influence of
the vibratile cilia which clothe the walls of the canals.

The Hydranth. — The hydranth (PI. XX, fig. 2) is flask-shaped, with two circlets of tentacles,
a proximal and a distal. The tentacles of the distal circlet surround the base of a conical hypo-
stome ; they are much smaller than those of the proximal ; they slightly alternate with one another
and appear to consist of two closely approximated verticils. Their ectoderm (PI. XXIII, fig. 4) is
loaded with thread-cells, but no distinct cellular structure can be detected in it. It is separated
from the endoderm by an intervening fibrillated layer. The endoderm consists of large cells,
each of which stretches transversely across the tentacle, leaving no pervious canal in the axis,
and conferring on the tentacle the characteristic septate condition of this part. These endodermal
cells contain a clear colourless fluid and present along the line of the axis accumulations of pro-
toplasm with coloured granules. The tentacle is eminently contractile.

The tentacles of the proximal set are much larger and thicker than those of the distal, and
their contractility is much less marked. They are disposed in a single continuous verticil.
Their ectoderm is similar to that of the distal tentacles, but their endoderm, instead of consisting
of large transversely arranged cells, is formed of smaller cells rendered polygonal by mutual
pressure, and presenting the condition of a pith-like tissue by which the tentacle is filled to the
complete exclusion of all trace of an axile tube. These cells contain a clear colourless fluid, with
protoplasm enveloping some more opaque corpuscles and accumulated in a layer on the cell-wall.

Between the ectoderm and endoderm of the tentacle there exists a fibrillated tissue
(PI. XXIII, fig. 5). I have succeeded in isolating this tissue and its component fibrillar in the

TUBULARIA INDIVISA. 207

tentacles of a specimen which hnd been long preserved in spirits. Two sets of striae, a longitu-
dinal and a transverse were visible in the interval between ectoderm and endoderra. Of these
the longitudinal striae were the more distinct and were manifestly the expression of longitudinal
fibrillac. On tearing the tentacle these fihrillae might sometimes be isolated, when they were
seen to be tubular, with a diameter of ai)out ;,;,,3 of an inch. They are perfectly smooth, but in
many cases a very distinct oval nucleus, having a shorter diameter of about ^„^ "f a" inch and
with a brilliant nucleohis, was visible in them. They might occasionally be seen to taper away
to a point, and I have little doubt of their being greatly elongated fusiform cells (fig. G).
The transverse strias may also re[)resent a fibrillated layer, but I am by no means so certain of
this, and yet I can scarcely regard them as rugae resulting from the contraction of the tentacle.

In longitudinal section (PI. XXIII, fig. 1) the hydranth is seen to contain a capacious cavity.
A little within the mouth the endoderm is thrown into prominent patches of a bright carmine
colour, with intervening paler furrows. Passing downwards from the mouth the carmine-coloured
patches are replaced by smaller spots ; still further down these become resolved into minute
puncta, which at the base of the hydranth-cavity are arranged in radiating striae.

A little above the origin of the posterior tentacles the endoderm sends off a zone of
pendulous fusiform lobes (PI. XXIII, fig. 1, c, and figs. 2 and 3). These lobes are composed of
large cells containing carmine-coloured granules, among which may be seen several clear spherules,
apparently oil-drops (fig. 3).

T/ie Gonophores. — The gonophores form pendulous raceme-like clusters, which spring from
the body of the hydranth immediately within the proximal zone of tentacles (PI. XX, figs. 2 and
3). They are adelocodonic, and are in the form of oval sacs borne on the summits of short
peduncles, which are given oft' alternately from a common tubular stalk whose cavity directly
communicates with that of the hydranth. In each raceme the gonophores increase in maturity
as they recede from the base of the common peduncle. The racemes thus differ from the inflor-
escence, which the botanist designates by this term, in being centrifugal instead of centripetal in
their evolution (see above, p. lOS).

In the walls of the gono[)hores three layers may be distinguished — an external (ectotheca,
PI. XXIII, figs. 8 and 11, d), a middle (mesotheca, b), and an internal (endotheca, e). The
ectotheca is perforated by an aperture in a point which is more or less diametrically opposite to
the point of attachment of the peduncle ; abundance of thread-cells are ind^edded in it. The
mesotheca is rendered obvious by the presence in it of four radiating canals (c), which extend from
the proximal towards the distal end of the gonophore, where they become united by short trans-
verse branches, forming a circular canal, which surrounds an aperture [d) in the mesotheca exactly
corresponding to that of the ectotheca ; where the radiating canals enter the circular canal they
become dilated into small bulb-like expansions, containing accumulations of coloured granules.
The endotheca in the female gonophore (fig. 11) disappears at an early period, apparently under
the pressure of the increasing mass of the generative plasma which is formed within it, though
it continues longer as a distinct membrane in the male (fig. 8) .

Occupying the axis of the gonophore is the large club-shaped spadix surrounded by the
generative mass (/). It consists of a hollow process of endoderm, whose lining cells are filled
with carmine-coloured granules and have their free surfaces clothed with vibratile cilia (fig. 10).
The radiating canals spring from the base of the spadix, where they comnmnicate with its cavity.

The spadix in a very early stage is invested by a layer of ectoderm, but the generative

208 ANATOMY OF SPECIAL FORMS.

elements soon make their appearance Ijetween the endodermal and ectodermal layers, which thus
l)econie separated from one another, and the ectoderm becomes an endotheca.

The generative mass in the female (fig. 11,/) constitutes a cellular plasma, into which the
spadix is plunged and from which portions are successively detached. These detached portions
{(j) lie loose in the cavity of the gonophore after the disappearance of the endotheca. They consist
of a mass of clear spherical cells, many having within them a brood of secondary cells (fig. 12).
They become developed into embryos, and though I have never succeeded in detecting in them
a germinal vesicle or witnessed in them the phenomenon of yolk-cleavage, they must, for reasons
stated in a former part of this work (see p. 90, &c.), be regarded as true ova in various stages of
development.

As already mentioned, the ovum becomes developed, not into a ciliated planula, but into an
actinula. The various stages of this development and the structure of the actinula are repre-
sented in figs. 13 — 10, and have been already fully described (see above, p. 90, &c.).

In the male the spermatogenous tissue also forms a mass (fig. 8, /) into the midst of which
the spadix is plunged. It is pi'oduced, like the generative mass in the female, between the
endoderm and ectoderm of the central process, so that the ectoderm becomes separated as an
endotheca (e), which continues longer apparent than in the female. As the spermatogenous mass
approaches maturity it presents the appearance of radiating striae, and after the rupture of the
endotheca the free spermatozoa escape as active caudate corpuscles (fig. 9) through the aperture of
the gonophore.

CORYMORPHA NUTANS.

Plate XIX.

This beautiful hydroid is eminently conspicuous by the great size of the solitary hydranth
with its dense clusters of medusae, and by the thick fleshy stem destitute of that firm chitinous
])erisarc wjiich forms a strong protective covering for others, while in its internal structure it is
further distinguished by well-marked and ])ecuhar features. Whether om- attention be directed
to its trophosome or to its gonosome, we cannot fail to recognise in Corpnorjjha a special confor-
mation which marks it out as a very distinct hydroid type among its co-ordinal associates.

The Hydrophjton. — The hydrocaulus (figs. 1, 2), which is always solitary, attains a height of
from two to three inches, while it varies in thickness from its distal to its proximal end, usually
attaining, in its thickest part, which is situated a little above its attached extremity, a diameter
of about two lines, and thence rapidly tapering downwards to a blunt point, which is plunged into
tlie sandy sea-bottom. The hydrocaulus is here a little above its proximal extremity bent nearly
at right angles on itself. The place of the perisarc is taken on it by an exceedingly thin colourless
and structureless film, which without careful examination would quite escape detection, except
towards the proximal extremity, where it is separated by a considerable interval from the ectoderm
of the stem, and thus becomes at this part sufficiently obvious.

Towards the proximal extremity fleshy conical processes arc developed from the sides of the
stem in longitudinal rows (fig. 1, a), while still lower down the stem gives off all round great
numbers of very much attenuated and elongated filaments {h, b).

Even by the naked eye the stem will be seen to be traversed from end to end by longitudinal

CORY.MORPHA NUTANS. 209

bands. Under a low magnifying power these bands are found to inosculate with one another
here and there, while towards the base of the stem they become fewer and broader by coalescence.
They represent canals excavated in the endoderm of the stem, and in a transverse section of tlie
stem (fig. 5) these canals may be seen to lie just within the ectoderm. The peripheral portion
(d) of the endodeiiu in which they are excavated is composed of small spherical ceils filled with
reddish granules, which give to the stem its pale reddish tinge, while the whole of the axis (e) of
the stem is occupied by a sort of pith, composed of large cells with colourless contents. Very
distinct currents may occasionally, by means of the microscope, be witnessed in the canals, so
that the structure and phenomena presented by these parts closely correspond to what we meet
with in the stems of Tabiilarla iiuKvim.

Under a high power of the microscope delicate parallel longitudinal strioe may be detected
lying external to the canals. They may be traced upwards for some distance on the body of the
hydranth. They lie between the endoderm and ectoderm, and represent a fibrillated muscular
tissue.

The peculiar short conical papillae (fig. 1, a) which arc given off from the stem near its proximal
end are arranged in regular longitudinal series, which follow the course of the canals, the stem
immediately over each canal bearing two alternate rows. They are tubular, with the cavity of the
tube apparently communicating with the canal over which they lie, and witii their free extremities
blunt and imperforate (figs. 10, 11). They appear to be extensile, and one or more of them n)ay
occasionally be seen to be much elongated, and then with their extremities slightly clavate
(fig. 10).

These tubular processes have never been seen to act as organs of adhesion, nor have we yet
any evidence of the office they may serve in the economy of the animal, but it is impossible not
to recognise in them structures having a close relation to the filaments of attachment which are
given off from the stem a little lower down.

These filaments (fig. \,b,b) constitute a remarkable feature in the structure of the Corymorpim.
When the hydroid is examined shortly after its detachment from the sea-bottom no trace of them
may be discoverable ; but after resting for a few hours in our aquaria the filaments become developed
in great numbers from the whole of that part of the stem which lies below the papilliform pro-
cesses, passing outwards through the fine structureless pellicle which here loosely invests the
extremity of the stem. They now rapidly extend themselves, adhering closely to the sides of the
aquarium and by repeatedly crossing one another form an entangled web-like tissue, by which the
Corymorpha has become attached. It is almost certain that such filaments existed also before
the removal of the hydroid from its native bed, but that in the act of detachment they were torn
off. to be renewed when an opportunity may be afforded, as in the confinement of the aquarium.

Under the microscope the filaments are found to be tubular, terminating each in an imper-
forate clavate extremity. They consist of a granular substance, which, as it continues to elongate
itself, becomes invested with a perfectly transparent, structureless film of extreme tenuity.

Besides these capillary filaments of adhesion there is nothing which can be regarded as a
hydrorhiza, and it would seem that the chief support of the hydroid, when in its natural condition,
is afforded by the proximal end of the hydrocaulus, which is then plunged into the sand of the
sea-bed, and thus fixes the hydroid to its place.

The Hydranth. — The hydranth (fig. 1) may be described as flask-shaped. It bears, towards
its base, a zone of long, imperfectly contractile tentacles, arranged in a single series, while at a

210 ANATOMY OF SPECIAL FORMS.

consideralile interval in front of this zone, and a little below the mouth, is a bnish-Iike group of
about eighty very contractile tentacles, much smaller than the proximal ones, and arranged in six
or seven closely placed alternate verticils.

When a longitudinal section (fig. 7) is made through the hydranth from the mouth to the
stem, it will be seen that the body, as far back as the zone of proximal tentacles, presents a con-
tinuous cavity (a) lined by small spherical cells containing coloured granules. The floor of this
cavity projects into it in the form of a broad conical elevation. This conical projection consists
beneath its covering of small pigment-bearing cells, of a large-celled and clear-celled endoderm (e),
and the same tissue is continued backwards as far as the summit of the stem. A careful examina-
tion will show that this cellular mass, which frequently seems to fill up the proximal part of the
hydranth, is perforated in its axis by a tubular prolongation of the cavity of the distal part,
though, in consequence of temporary obliteration caused by the approximation of its walls, this
axial tube is usually difficult to detect. Like the cavity of the distal portion of the hydranth, it
is lined with small spherical cells {/) filled with coloured granules. It is continued towards
the summit of the stem, and then, becoming wider (3), receives the longitudinal canals (e), which
have been already described as traversing the stem in its entire length.

The tentacles forming the proximal zone are destitute of any trace of a cavity (fig. 8), and
consist of a simple extension, with little change of form, of the large-celled colourless endoderm of
the body, surrounded by a layer of ectoderm. The distal tentacles seem to admit the cavity
of the hydranth for a short distance into their interior, but the tube soon becomes obliterated, its
axis at the same time assuming the usual septate appearance (fig. 9).

The same difference of structure between the proximal and distal sets of tentacles exists
here as in Tahularia, and I regard the distal set in both cases as the true equivalents of the
tentacles in Eudendrium, Coryne, &c., while the proximal set must be viewed as superadded
structures.

In fully developed specimens the hydranth measures about one inch across the proximal or
posterior coronal of tentacles. The colour of the brownish-red contents of the cells which line its
cavity is visible through the transparent ectoderm as far back as the proximal zone of tentacles.
Similarly coloured granules exist in the stalks of the gonophores, from which they may be traced
into the manubrium of the medusa, and, while this is still young, into its radiating canals. The
whole of that portion of the hydranth which lies behind the long tentacles is nearly colourless,
the pigment-containing cells being here obscured by the thickness of the colourless layer of
endoderm which lies external to them.

The Gonophores. — The gonophores (figs. 1, 3, 4, 5, 5") are grouped in crowded clusters on the
extremities of branched tubular stalks which form their peduncles. The axis of these peduncles
is occupied by a continuous tube, which communicates freely with the cavity of the hydranth.
They are usually from fifteen to twenty in number, and are situated immediately within the zone
of proximal tentacles in two alternating series (fig. 1 and fig. 7, /)•

I have not been able to demonstrate in the gonophores any trace of an ectotheca, and must
therefore regard them as truly naked. They are planoblasts, and when arrived at that stage in
which they are about to detach themselves from the stalk and become free (figs. 3, 4), they present
a deep umbrella, having its summit continued into a short conical projection traversed by a narrow
canal, which had kept the cavity of the manubrium in connection with that of the stalk. There
are four radiating canals, each of which expands into a bulb at the point where it enters the

CORYMORPHA NUTANS. 211

circular canal. Of these bulbs one is much larger than any of the others, and is continued into
a tentacle, while none of the others present any trace of such an appendage. They all contain
brownish-red pigment-granules, but no true ocellus can be recognised. There is a broad velum.

The solitary tentacle is largely developed, and consists of a very extensile moniliform cord,
presenting, when extended, the appearance of ten or twelve little spherules distributed at equal
distances upon a cylindrical string. The last of these spherules exactly terminates the string,
and is larger than the others, while one or two situated near the proximal end are smaller and less
distinct. The spherules are composed of accunuilations of thread-cells, and the connecting cord
has its axis occupied by an rminterrupted tube directly prolonged from the cavity of the bulb at its
root. During contraction the spherules assume the form of circular discs (fig. 4), and in extreme
contraction the connecting cord disappears and the surfaces of the discs are brought into contact.

The manubrium is large and sub-cylindrical, and the mouth is without tentacula or lobes.

From the above description it will be at once apparent that the medusa of Corymorpka
nutans belongs to a form to which Edward Forbes has given the generic name of Steemtrupia}
Of this relation between Steemtrujpia and Corymorpka Forbes himself had a suspicion ; indeed,
he expresses a belief that his Steenstriqjia rubra will turn out to be the free medusa of
Corymorpka nutans. The development of the medusa from its first appearance to its liberation
has been already traced (see above, p. 77).

The medusae when they become free are about 53 inch in diameter (fig. 3), and as yet show
no trace of generative elements, and though I kept them alive for more than a week they
scarcely increased in size, and never presented any indication of ova or spermatozoa.

I oljtained, however, by means of the towing-net, in the neighbourhood of the locality which
produced the Coryinorpha, a little medu.sa (figs. 5, 5") regarding which there can be no doubt that
it is a more advanced stage of the gonophore of Corymorpha nutans. It was about four times
the size of the newly liberated medusa ; the tentacle had proportionately increased in length and
now presented upwards of forty spherules, while the radiating canals at their origin from the
manubrium curved upwards towards the summit more decidedly than in the younger form.
The generative elements— not yet, however, fully developed, and apparently male, though from
their immature condition no active spermatozoa could be detected — were very distinctly visible
as a pale yellow mass between the endoderm and ectoderm of the manubrium, which was rendered
tumid by their presence (fig. 5). In all other respects the little medusa was identical with the
younger ones, and continued to present the acuminated summit, which was even still traversed by
the canal which originally maintained a communication between the tube of the supporting stalk
and the cavity of the manubrium.

It would thus seem that the changes undergone by the medusa between the time of its
liberation from the trophosome and its attainment of sexual maturity are of little importance.
It is worthy of especial attention that there is never more than a single marginal tentacle developed.

Free Frustules. — But besides the production of medusiform sexual buds, I have also witnessed
in Corymorpha nutans another process of reproduction, very remarkable, but of whose exact sig-
nificance I was unable at the time to speak with as much confidence as I can now do. In a glass
jar containing living specimens of Corymorpha, which had been in my possession for more than a
fortnight, I observed attached here and there to the surface of the glass minute oblong bodies

^ Forbes, ' British Naked-eyed Medusae,' p. 72.

28

212 ANATOMY OF SPECIAL FORMS.

(fig. 1 2), about half a line in their longer diameter and one eighth of a line in the shorter. They
appeared to be composed of a soft, minutely granular, colourless substance, and their intei'ior was
occupied by a very distinct cavity. They were destitute of cilia, and were invested by an extremely
delicate membranous or mucous tube, quite structureless, which extended for some distance beyond
their ends and adhered for its whole length to the side of the jar.

Besides these little bodies others (fig. 13), which I do not hesitate to regard as the same bodies
in a more advanced stage, were also found attached to the sides of the jar. They consisted of a
colourless tubular filament, about four lines in length, attached to the glass by one extremity and
developed at the opposite into a minute hydranth, having a general resemblance to the hydranth
of Corymorpha, but with only six or eight tentacles in the proximal circlet, while the distal
circlet was also composed of six or eight tentacles, which were shorter than the proximal ones, with
blunt, almost capitate extremities, and, like the proximal tentacles, disposed in a single verticil.

Others (fig. 14) representing a still more advanced stage were also found attached to the
sides of the jar. They had attained a size about double that of the last, while the posterior
tentacles now presented a verticil of sixteen or twenty, and the anterior ones, though still dis-
posed in a single verticil, had become multiplied to about the same extent.

Beyond these three stages I was unable to trace the development through any further steps.
The last of them, however, manifestly requires little to convert it into the form of the advdt
Corymorpha.

If it were not that the medusa thrown off from the adult hydroids in my jars had, so far
as I could find, all perished before the formation in them of generative elements, I should have
regarded the little organisms just described as presenting three stages in the development of the
embryo from the ovum. In the absence, however, of all evidence of the presence of ova, we must
seek for some other explanation of these little bodies. The phenomena of spontaneous fission,
already described as occiu-ring in a campanularian (see above, p. 152, woodcut, fig. 61), will here
afford a clue, and the close resemblance of these enigmatical bodies to the fission fmstules of the
campanularian will suggest the probability that they represent different stages in the development
of fragments which had been spontaneously detached from some of the filamentary processes
emitted from the proximal end of the adult specimens in the jar.

Clavatella prolifera.
Plate XVIII.

Clavafdla prolifera, though of small size, is unrivalled in the elegance of its form, while its
wonderful ambulatory medusae present a combination of characters which, if we except the closely
allied Elcutheria of De Quatrefages, is quite unparalleled among hydroid planoblasts.

The Hydrophyton. — In this remarkable little hydroid the hydrocaulus is rudimental, having
for its sole representative very short simple processes (fig. 1), which are given off at intervals
from a creeping filiform hydrorhiza, and which carry each a hydranth on its summit. These
processes as well as the hydrorhiza are invested by a delicate perisarc.

Tlie Hydranth. — The hydranth is well developed, and when extended is of an elongated,
nearly cylindrical form, abruptly dilated at its base and carrying on its distal extremity a single

CLAVATELLA PROLIFERA. 213

verticil of from six to eight capitate tentacles round the base of a short conical hypostome. It is
very contractile, and in extreme contraction assumes a short, thick, flask-shaped form. During
contraction the body is seen to be marked by fine, regular, closely set transverse rugae.

A longitudinally fibrillated layer can be distinctly demonstrated beneath the ectoderm on
the body of the hydranth, through the entire length of which it may be traced (fig. 11). The
endoderm presents a peripheral layer of large cells with clear colourless contents, and an axial
layer which lines the somatic cavity and is composed of much smaller cells containing coloured
granules. In the ectoderm no distinct cellular structure can be demonstrated, but it contains
numerous thread-cells imbedded in it.

The tentacles (fig. 3) present the usual septate condition. The spaces included between the
septa have their axes occupied by an accumulation of granular protoplasm, often looking like a
continuous axile column running through the whole length of the tentacle. The capitulum in
which the tentacle terminates is very well marked, and is loaded with large thread-cells.

The Gonophores. — The gonophores (fig. 1 and figs. 4 — 10) are borne on one or two short-
branched tubular peduncles, which spring from the hydranth at the spot where the basal dilatation
is seen in the extended hydranth to pass into the narrower, more distal portion of the body. When
two of the branched peduncles are present they are situated upon opposite points of the body.
There are generally two or three gonophores, in various stages of development, in a cluster, each
being borne on the extremity of one of the branches of the peduncle, whose tube com-
municates freely on one side with the digestive cavity of the gonophore, and on the other with that
of the hydranth.

The fully formed gonophore (fig. 4) is medusiform, but yet singularly exceptional among
medusae, for its umbrella is never developed into a swimming-bell, and locomotion is performed by
peculiarh' constructed marginal tentacles, on which this strange ambulatory planoblast is capable
of creeping about from place to place.

The gonophore is dome-shaped, and when viewed from its dorsal aspect under the carefully
regulated action of the compressorium, it will be seen to be occupied by a large central cavity,
from the circumference of which six short but wide tubes radiate at equal intervals towards the
margin of the dome, where they meet a wide circular canal, which runs parallel to this margin
and close to it (figs. 4, 5).

At points corresponding to those at which the radiating canals meet the circular the margin
is extended into six long tentacles, each of which divides into two branches at some distance
beyond its middle. The two branches are situated in a radial plane, so that one is internal or
looks towards the axis of the medusa, the other external or directed away from the axis. The
external branch terminates in a capitulum of thread-cells, the internal one in a slightly swollen
extremity, which carries a minute suctorial disc (fig. 7). The endoderm of the tentacle and of its
branches is large-celled, giving to them a somewhat septate appearance; but the axis is pervious,
being traversed by a tubular prolongation of the circular marginal canal.

In the centre of the under surface of the dome-shaped disc is the quadrangidar aperture of
the mouth, sometimes scarcely raised above the general surface, and sometimes elevated on a
prominent conical projection. It is entirely destitute of lobes and of tentacular appendages.

On the outer side of the base of every tentacle is a very distinct ocellus, consisting of a heap
of red pigment-granules, in which a clear refracting spherule may sometimes be seen to be
immersed, and over which a transparent, slightly elevated extension of the ectoderm may be

214 ANATOMY OF SPECIAL FORMS.

traced (figs. 4, 6). Though I found the refracting lens-like body occasionally very conspicuous, it
was frequently impossible to discover it, and it seems to be very soft and easily broken down in the
mass of the pigment. This will probably explain the fact that while it was detected by Quatrefages,
Krohn, and Claparede in the Eleutheria- and Clavafella-medusai described by them, it eluded the
attempts of Hincks to discover it. The correctness of Quatrefages' statement of the evidence of
a cornea in the ocellus of Eleutheria is disputed by Krohn ; my own observation, however, of a
convex extension of the transparent ectoderm over the surface of the ocellus in Clavatella tends,
on the other hand, to confirm it.

Just above the base of the tentacles the medusa is encircled by a zone of thread-cells
imbedded in the ectoderm.

I have never met with the male, and, with the exception of Krohn, who saw it once, no
other observer seems to have been more fortunate. I found Clavatella jjroUfera, in autumn, on
the west coast of Scotland, with its medusas carrying eggs (fig. 8). These were contained
between the endoderm and ectoderm of the dorsal or proximal side, as had been already noticed
by Quatrefages in Eleutheria, and by Hincks and Krohn in Clavatella. They were few in
number, not more than five or six, were of comparatively large size, and by their pressure from
within gave rise to prominent bosses on the dorsal surface of the medusa. In every case in
which I noticed them they seemed to have passed the earlier stages of their development, for
their germinal vesicle was no longer visible ; but I lost my specimens before tracing the ova, as
Krohn had done, into the condition of "embryos," which he informs us they attained while still
in the dorsal brood-chamber of the medusa. From this chamber it would seem, according to
the observations of Krohn, that they ultimately escape by the rupture of the ectoderm which con-
fines them, and then swim away in the form of free ciliated planulse.

Claparede' and Filippi^ have also described the gonophore of a Clavatella, but their accounts
differ from that just given in some important points. In Claparede's medusa, which seems distinct
from that of the Clavatella prolifera, the ova lay beneath the ectoderm of the distal or lower side,
whence they projected externally. Filippi, though finding the ova in a dorsal chamber, will not
allow that this chamber is formed by the separation of the ectoderm and endoderm, as in all other
Hydroida. He regards it, on the contrary, as a special cavity, bounded on all sides by the endo-
derm— a view which is almost certain to have originated in some error of observation.

The medusa makes its appearance as a minute, closed, pyriform sac (fig. 1), consisting of a
layer of ectoderm lined by a layer of endoderm, and having its cavity in communication through
its peduncle with that of the hydranth. As development proceeds, its cavity may be seen to
extend itself at the distal side in six short, tubular, radially disposed offsets, lined by the endo-
derm, and prolonging themselves into the substance of the thickened ectoderm of the distal end
of the bud. These are the incipient radiating canals, and it is certain that about this time each
becomes united to its neighbour by a transverse communication, the commencement of the cir-
cular canal, though the actual formation of this communication has escaped me.

Up to this point tiie ectoderm is continued evenly over the whole bud ; but it now begins to
show, at the distal side of the bud, six little elevations, which correspond to the distal extremities
of the radiating canals (fig. 9). These elevations gradually increase in length, while an exten-

' Claparede, ' Beobachtungen iiber Anat. und Entwicklungesch. wirbellos. Tliiere,' p. 4.
' Filippi, ' Mem. della lleale Academia d. Scieuze di Torino,' Ser. 3, Tom. xxiii.

CLAVATELLA PROLIFERA. 215

sion from the canal system prolongs itself into their axes. They are now plainly seen to be the
rudimental marginal tentacles, as yet, however, quite simple.

We next observe a small tubercle showing itself close to the extremity and upon the inner
side of each of the developing marginal tentacles (fig. 10), so that the tentacles at this stage appear
obscurely double-headed, the outer head set with thread-cells, the inner one without any special
accumulation of these bodies. The two heads become rapidly more and more distinct from
one another, each borne on a stalk, and the tentacle thus becomes bifurcated at its extremity.
The thread-cells increase in number on the extremity of the onter branch, which now terminates
in a distinct capitulum, while a suctorial disc of attachment is developed on the extremity of the
inner branch. The ocellus is by this time quite distinct, and we soon find that the little medusa,
fitted for an independent existence, has detached itself from its stalk and creeps about on the
corallines and sea-weeds of its rock-pool, the inner branch of its tentacles serving for attachment
and locomotion, the other aiding it apparently in the prehension of its prey (fig. 4).

It will be noticed that in the development just traced an ectotheca is never present, the
medusa being exposed during the whole time to the direct contact of the surrounding water.

The important observation was made by Krohn that the C/avatelia-medusa may repeat itself
by buds from the medusa itself, and he has further shown that this gemmation may take place
simultaneously with sexual reproduction, the buds being borne by medusae which are at the same
time laden with ova.

I can fully confirm these observations of Krohn. The buds are tlevcloped from the margin
of the disc, and from the middle of the interval between two tentacles (fig. 5, a, b). They arc
almost always two in number, and are then pi-oduced in two opposite interradial spaces, from
which they project outwards between the roots of the tentacles. They originate in the form of a
small hernia-like process from the circular canal, consisting of a layer of ectoderm lined by one
of endodcrm and having its cavity freely communicating with that of the circular canal. Their
development is entirely similar to that of the medusas budded from the hydranth.

I found the buds produced both by medusae in which eggs were at the same time present
and by those in which eggs had not yet shown themselves.

A very remarkable difference from the position just described as that of the buds in
Clavatella proUfera had, however, been noticed by Claparede in the species which came under his
observation. In this the buds, while still produced from the circular canal in the interradial
spaces, are described as developing themselves under the ectoderm, so as to lie between the disc of
the medusa and its digestive cavity. It is difficult to say where the space exists for them in this
direction unless that in their growth they insinuate themselves between the ectoderm of the disc
and the endoderm of the enlarged atrium. Indeed, Claparede expresses some doubt as to their
being really buds, and thinks it possible that they may be eggs which had passed on to an
advanced stage of development in the proper egg-cavity, which in this species is described as
lying beneath the ectoderm of the lower side.

A comparison of the medusa of Clavatella with an ordinary hydroid medusa gives results of
much interest.

It will be borne in mind that in many medusae, such, for example, as those of Synconjne
and Bougainvillia, the proximal end of the manubrium or that from which the radiating canals are
given off presents a special dilatation, and has those endodermal cells which are characterised
by the secretion in them of the coloui'ed granules more abundantly developed there than elsewhere.

216 ANATOMY OF SPECIAL FORMS.

This dilatation, at least in the young stage of the medusa, is more or less deeply plunged into the
substance of the umbrella. It is the " common cavity" of Huxley, the " chyme-receptacle" of
Gegenbaur, and the " atrium" of the present Monograph.

If we suppose this atrium to become greatly developed and to extend itself laterally until it
occupies nearly the entire width of the umbrella summit or atrial region, while the concave por-
tion or manubrial region of the umbrella, as well as the free portion of the manubrium itself, attains
but slight development, we shall have as the result a medusa with the characteristic features of
that of ClavateUa} The umbrella will no longer present the free dilatable and contractile bell of
an ordinary medusa, while the radiating canals, which in the latter extend throughout nearly
the whole height of the umbrella, will in Clavalella, by the approximation of the atrium
to the circular canal, be reduced to the condition of the short tubular passages necessary to keep
up the communication between these two parts of om- singularly modified medusa. The marginal
tentacles agree with those of the natatory medusa in their position and relation to the gastro-
vascular system, but to compensate for the loss of locomotive function in the umbrella they take
this function upon themselves, and are fitted for the conveyance of the medusa from place to
place by acting as organs of reptation.

The modification undergone by the manubrium will explain the situation of the ova,
apparently so very different from that of those bodies in the natatory medusa. In consequence of
the very imperfect development of a free pendulous manubrium, the generative elements which in
a natatory medusa would be developed within the cavity of the umbrella in the walls of the
manubrium are here, as it were, forced upwards and compelled to occupy a space immediately
beneath the ectoderm of the back, through which they at last escape by rupture ; though in the
species described by Claparede they approach more to the usual position, lying beneath the
ectoderm of the lower surface, and ultimately escaping at this part of the medusa. In all cases,
however, whether the medusa be natatory or ambulatory, the generative elements have their
origin between the ectoderm and endoderm.

Cladonema radiatum.
Plate XVII.

This singular and beautiful little hydroid was first described by Dujardin. It possesses
special interest not only as a well-marked type with striking peculiarities of structui-e, but as
having been the first in which the entire life-scries was followed through the complete succession
of hydranth, medusa, and planula, back to the hydranth again.

The Hydropliyton. — The hydrophyton consists of very slender branched stems, which spring
from a creeping filiform stolon (figs. 1, 2). The whole is invested with a smooth perisarc. The
ramification, however, varies much, and specimens may frequently be seen consisting of a delicate
tubular filament adherent to the walls of the aquarium, over which it may extend for several
inches, while it sends off from distance to distance short simple hydranth-bearing ramuli. In
some of the specimens examined long slender branches, which showed no hydranths on their

' Tlie comparison of the digestive cavity of the Ctoi'a/e//a-medusa with the " chyme-receptacle"
of Gegenbaur Las already been made by Krobn.

CLADONEMA RADIATUM. 217

summits, and had the pcrisarc, a little beyond the origin of the branch, thrown for a short dis-
tance into spiral folds, were given off here and there from the hydrocaulus. I can scarcely,
however, regard these branches as normal ; and as all the cases in which the trophosome of
Cladonema radiatum has been observed were those of specimens which had shown themselves in
the confinement of the aquarium, it is not improbable that the growth of these adventitious
branches is the result of the imnntural conditions to which the hydroid had been there exposed.

The Ilydranths. — The hydranth, which is somewhat fusiform in shajie, has its tentacles dis-
posed in two verticils, with four tentacles arranged crucially in each. The two verticils differ from
one another in the form of the tentacles. In the distal verticil, which surrounds tiie base of a well-
developed hypostome, the tentacles terminate in spherical capitula, crowded with thread-cells ;
while in the proximal verticil, which is situated close to the base of the hydranth, the tentacles
are much smaller than those of the distal verticil, are less contractile, and are destitute of
capitula. The endoderm of both sets is septiform, with some opaque granules imbedded in
protoplasm which occupies the axis.

The Gonophores. — The gonophores (figs. 3 — 10) are phancrocodonic, and present us with an
interesting and remarkable type of hydroid medusas. They bud from the hydranth immediately
above the proximal verticil of tentacles.

After detachment from the trophosome these curious planoblasts easily attain to maturity
in the aquarium, differing in this respect from most other planoblasts, which in the confinement
of the aquarium soon perish, without ever attaining their mature condition.

The mature planoblast of Clado7irma (figs. 3, 4, 10) is about a hne and a half in height.
When in repose the vertical and transverse diameters of the umbrella are nearly equal to one
another, but during systole the vertical considerably exceeds the transverse. The walls are thin,
and the outer and inner surfaces nearly parallel. A small boss-like hemispherical elevation projects
from the outer surface of the summit.

The sub-umbrella is of a pale raw-sienna colour, and under the microscope may be seen to be
marked with distinct stria?, which take a transverse, somewhat wavy course between every two
radiating canals. They appear to indicate the action on the sub-umbrella of the contractile tissue
of its walls.

The velum is very wide, the diameter of its central aperture when most diminished being
scarcely more than one eighth of the entire diameter of the codonostome (fig. 8). It is marked
by very distinct concentric striae, intersected by radiating striae, which extend from the margin of its
central aperture to its outer edge. Both these sets of striae represent the contractile fibrillated
tissue of the velum.

The manubrium manifests great contractility. It is sometimes drawn together, so as to
present the appearance of an irregular globular mass in the top of the bell, and at other times
extended so as to touch with its distal end the surface of the velum, or even become protruded
beyond it. When extended it is seen to consist of two regions, a proximal and a distal (figs. 3, 4j.
The proximal region occupies about two thirds of the entire length of the manubrium. It is
somewhat fusiform when viewed from the side, but in transverse section it is seen to be penta-
hedral. It is in this portion alone that the generative elements are developed, and their presence
gives rise to five hemispherical projections, by which the manubrium becomes encircled near
its middle. The distal portion constituting the remaining third is very much thinner, and is
nearly cylindrical, though tapering slightly towards the mouth. The oral margin is continued

218 ANATOMY OF SPECIAL FORMS.

into five sliort cylindrical processes, each of which carries on its extremity two or three small
globular clusters of thread-cells.

At the proximal extremity of the manubrium its cavity is prolonged into five small narrow
pouches, which are seen at the angles of a pentangular area, which forms its fundus and which is
visible when we look down upon the medusa from its vertex (fig. 7). When the manubrium is
viewed laterally the pouches give to it the appearance of being suspended by five short roots from
the summit of the bell. It is in these pouches that the radiating canals originate. Sometimes
one or more of the pouches give off directly two canals ; sometimes a single canal arises directly
from a pouch. After a longer or shorter course the single canals will bifurcate, and as the ultimate
result ten radiating canals are formed. These pass down at exactly equal distances from one
another to the circular canal, into which they open after presenting just before their junction with
this canal a marked enlargement of their cahbre.

All the other authors who have described the Cladonema-mednsn give eight as the number
of the radiating canals and marginal tentacles. It is evident that there is in this respect a
variation which must not be regarded as pointing to a specific distinction.

The marginal tentacles correspond in number with the radiating canals, of which they are,
in fact, the direct continuation; in my specimens there were therefore ten of them. They are at
first very thick, continuing so for some distance from their origin, but as soon as they begin to
give off their branches they rapidly taper away to their extremities. A bulbous dilatation with a
very distinct ocellus exists at their junction with the circular canal. After continuing simple for
about a third of their length they begin to emit peculiar appendages. These are given off
alternately, and all direct themselves inwardly or towards the axis of the medusa. The first
three of these appendages are veiy different from the others. They are nearly cylindrical, though
tapering slightly towards their distal extremity, where each terminates in a spherical capitulum
thickly set with thread-cells (fig. 5). Their endoderm is composed of large cells, which give
them an irregularly septate appearance, while a line of coloured granules occupies their axis. The
proximal one is shorter than the other two, which are equal to one another in length.

The rest of the appendages consist of very e.xtensile offsets from the tentacle, the last
forming a bifurcation with the terminal portion of the tentacle. When extended (figs. 3, 4, 6)
they are in the form of cylindrical filaments, all thickly set with spherical clusters of thread-cells,
which give them a nodulated appearance, a single cluster, larger than the others, terminating
each appendage. When contracted (fig. .5) they are in the form of short, thick, club-shaped
offsets. Each tentacle carries four such appendages.

The tentacle from the point at which the nodulated appendages begin to be given off to its
extremity is covered in a similar way with knot-like clusters of thread-cells, and, like the
ai)pendages themselves, is terminated by a large spherical cluster.

The axis of the tentacle presents throughout a continuous tube, which extends laterally into
the nodulated appendages as far as their ends, and contains abundance of dark crimson-brown
granules, thus contrasting with the smooth capitate appendages, which merely present a few
scattered coloured granules in the axis of their large-celled endoderm. At the spot where the
last-mentioned appendages are given off the main tube of the tentacle slightly widens, and then
again becomes diminished in diameter, to taper away with the tapering tentacle to its distal
extremity.

The three smooth appendages which are given off near the base of the tentacle possess very

CLADOXEMxV RADL\TUM. 21 &

little contractility (fig. 5). They serve the medusa as organs of attachment, the terminal
caijitulum having tlic power of closely adhering snckcr-likc to any surface to which it may be
apiilied, and the little medusa may be often seen attaching itself by means of these appendages to
the sides of the glass jar in which it is confined.

The ocellns is very conspicuous. It is situated in tlic walls of the Ijulbous expansion, in
which the tentacle originates, and on tliat side of it which is turned away from the axis of the
medusa. It consists of an accunudation of pigment-granules, appearing of a deep crimson by
reflected light, and having a spherical refracting body imbedded in them (fig. 5).

In a very early stage the marginal tentacles of the medusa are quite simple (fig. 1), I)ut when
it is about to free itself from the trophosome they present each a single appendage (fig. 9). Tliis
consists of one of the smooth capitate organs of adhesion, while that portion of the tentacle which
lies beyond it is covered with knot-like groups of thread-cells. As development proceeds the
organs of adhesion increase in numljer to three, the nodulated branches being at the same time
given off in succession, from tlie more distal portion of the tentacle, until they acquire the complete
number which characterises the adult medusa. During no period of the development of the
medusa could I detect the presence of an ectotheca.

While the medusa is in the act of swimming the tentacles are contracted and curved
npwards round the margin of the umbrella. The cirri, or nodulated appendages of the tentacles,
are at the same time more strongly contracted, but the appendages destined for attachment always
remain extended (fig. 5). While floating passively in the water (fig. 3) the tentacles with their
appendages are extended to their entire length, and then hang down in a graceful tassel-like
cluster from the margin.

The most remarkable attitude, however, is assumed when the medusa has fixed itself to the
side of the aquarimn (fig. 4). It then adheres to the glass by means of the sucker-like capitula,
which terminate the organs of attachment. On these appendages the medusa is elevated as on
SI) many feet, while all the rest of tentacle is extended to the ntmost, and thrown back over the
umbrella. The cirri are at the same time fully extended, and as they are now turned outwards,
float freely in the surrounding water, spreading themselves out with their thread-celUloaded
spherules, as if in search of prey. A more fascinating object it is scarcely possible to conceive of
than this wonderful little planoblast, whether moving in rapid darts through the water with alter-
nate systole and diastole of its crystal bell, or carried along passively by some gentle current with
its long pendulous tentacles drooping from the margin, or fixed by its suckers with these same
tentacles now tin-own around it in n whole grove of long flexile stems and undulating
branches.

29

220 ANATOMY OF SPECIAL FORMS.

Hydractinia eciiinata.
Plate XV aiul Plate XVI, figs. 10 and 11.

In many respects this hydroid departs widely from the usual condition of the order. It is
full of interest, whether we regard the very exceptional structvn-c of the hydrophyton or the great
extent to which heteromorphism has been carried in the zooids.^

The species is found spreading over the surface of dead univalve shells, almost always such
as are inhabited by a hermit crab.

The Ilijdrophyion.- — There is nothing in the morphology of Hydractinia more deserving of
attention than the coenosarc with its chitinous excretion. Investing the shell of which the hydroid
had taken possession, and usually extending for some distance beyond the shell-mouth, is a
continuous layer (PI. XV, figs. 1, 2), soft and fleshy in the living animal, but replaced after the
death of the hydroid and the disappearance of its soft parts by a firm chitinous crust, thickly
covered by blunt spines. This crust was the only part of the animal with which tlie earlier
observers were acquainted, and, being very different from all other hydroid forms then known, its
real nature was entirely misunderstood, and it was sometimes regarded as a polyzoon, sometimes
as a sponge. It is the chitinous secretion of the fleshy ccenosarc which forms a common basis
for the various zooids of the polymorphous colony.

When a section of this common horizontal basis of the colony is made in the living animal
from the more superficial towards the deeper parts, the knife lays open a multitude of polyhedral
spaces (PI. XVI, fig. 10). These are the cross sections of freely anastomosing tubes, which, being
placed in different planes, form a canaliculated or sponge-like structure. These tubes consist of
the soft fleshy ccenosarc, bounded by a wall of chitine, and having their cavity surrounded by a
layer of minute cells filled with coloured granules. The fleshy mass itself I believe to consist
mainly of an endoderm, the cells with coloured granules which surround the cavity of the tulles
corresponding to a similar layer which surrounds the somatic cavity in hydroids generally. The
peripheral portion of this endodermal mass appears to be surrounded by a true but very thin ecto-
derm, on which devolves the function of secreting the chitinous investment of the canals. This
ectodermal layer is sufficiently obvious in the very young coenosarc, and in the superficial canals
of the older, but in other portions of the adult ccenosarc it is very difficult to demonstrate it.

At the free surface of the coenosarcal expansion (PL XVI, fig. 10, a a) its intercommunicating
canals (i) are only partially invested by chitine, this excretion being in the superficial layer of canals
confined to their deeper parts, thus forming open channels, in which the canals are lodged, so that
when the soft parts are removed the chitinous perisarc forms on the surface a nuiltitude of inter-
secting ridges having between them the channels which had contained the superficial coenosarcal
canals. Upon the whole of the free surface, however, the ectoderm of these canals forms a con-
tinuous and very conspicuous layer [a a), having acquired increased thickness and developed in its

' Dr. Strctbill Wright has given au excellent account of this hydroid. He was tlie first to call
attention to the occurrence in it of the spiral zooids. See ' Proc. Roy. Pliys. Soc. Edin.'

HYRACTINIA ECIIINATA. 221

substance abundance of thread-cells. The whole free surface of the connuon basal expansion of
the colony thus presents an absolutely naked layer of ectoderm.

Over the entire surface short blunt conical spines spring at short intervals (PL XV, fig. 1).
These spines consist of a chitinous framework, continued from the general chitinous system of the
common basis, and overlaid and penetrated by an extension of the ccenosarc.

The framework of each completely formed spine consists of a conical process of chitine {fig. 7),
whose surface is traversed, from the base to the apex, by longitudinal jagged ridges, which at the
base of the spine are continuous with the ridges of chitine, between which, as already said, the
superficial coenosarcal canals are lodged. A transverse section close to the base of a W'Cll-deve-
loped spine shows it to contain numerous cavities, which are here arranged in more or less regularly
concentric series (fig. 4). These cavities diminish in number and regularity towards the middle
of the spine, while close to the summit they have entirely disappeared, and the longitudinal ridges
here meet in the axis (fig. 5).

The cavities of the spine intercommunicate with one another, like those of the common
horizontal layer, and, like them, each is filled with a mass of cocnosarc, having a central cavity
surrounded by cells filled with coloured granules. The furrows occupying the intervals of the
longitudinal ridges are also filled with ccenosarc (figs. 4, h, .5, and 7, h), which is here excavated
into similar cavities, while they are closed in by a layer of distinct thread-cell-bearing naked
ectoderm, which thus invests the whole spine, precisely as in the common horizontal basis.

It is plain from this that the structm-e of the spine is in all essential points the same as that
of the horizontal basis, chiefly diS"ering from it in the greater development of its chitinous element,
which acquires much greater thickness and shows very evident layers of successive deposition.
Towards the mouth of the shell the spines are but slightly developed, and are here little more
than short, blunt, conical tubercles (fig. 1).

There can be no doubt that the whole hydrophyton of Ili/dracf'mia must be regarded as
consisting of a set of ccenosarcal, freely intercommunicating tubes, which have excreted from their
sm-face a chitinous perisarc, and have intimately coalesced with one another.

The reproduction of the hydrophyton may be watched in a mutilated specimen, kept alive in
a jar of sea- water; and it will be then seen to be at first a simple network of ccenosarcal
tubes (PI. XVI, fig. 11) invested by a chitinous perisarc, and, as Dr. Strethill Wright first
pointed out, presenting very distinct currents in the contained somatic fluid. It is obvious that
there is in this stage no essential difference between the hydrophyton of Hi/dractinia and that of
other hydroids with undeveloped hydrocaulus. The meshes of this network, ho\\'cver, are
ultimately obliterated by the thickening and coalescence of their chitinous walls.

The Ili/dranths and Spiral Zooids. — Prom the whole of the naked surface of the ccenosarc,
whether covering the horizontal layer or extending over the spines, the hydranths arise ; while
the spiral zooids, to be presently described, are confined to a narrow space along the extreme edge
of the colony. The somatic cavity of both hydranths and spiral zooids is in direct communication
with the system of canals which traverse the ccenosarc in every direction, while their ectoderm is
directly continuous with the naked superficial ectoderm of the ccenosarc, and their endoderm
with the endodermal portion of this same common basis of the colony.

The hydranths (PI. XV, fig. 1, ct, a, a) are eminently contractile, and distinct longitudinal
muscular fibres may be traced in them through their entire length. When fully extended thcj-
have the form of long, attenuated, nearly cylindrical columns, dilated towards the summit;

222 ANATOMY OF SPECIAL FORMS.

hut wlicn in various states of contraction tlicy may assume the shape of thick cluljs, or Lc more
or less fusiform, or present a trumpet-shaped figure. The nioutli may be so completely closed
that it will be difficult to find any indication of it, or it may be kept widely open with the cavity
of the hypostonie actually everted, and the contracted tentacles thrown back and in great ])art
concealed beneath the reflected lip. So very mutaljlc, indeed, are tiie liydranths that the observer
■can scarcely avoid attributing to the colony a greater amount of heteromorphisni than it really
possesses. In its most contracted state the Ijody of the hydranth is marked by very distinct
transverse rugae.

Tiie tentacles vary much in number. Tliey are usually from fifteen to twenty, but I have
counted as many as thirty-one in a full-grown hydranth. Tlicy consist of two or more alternating
series, so closely approximated as to constitute a single circlet surrounding the base of a well-
developed hypostome. They are irregular in length, those belonging to the distal series of which
the circlet is composed being longer than those of the proximal series. The axis of the tentacle
is formed by the usual large-celled endodermal tissue, but the cells are of more unequal size, and'
give a less regularly septate appearance to the axis than is the case in many other hydroids.

I have occasionally met with a hydranth which has become double-headed by a bifurcation
near its middle. Quatrefages has also noticed this condition, and Agassiz has noticed it in the
spiral zooids to be next described.

Near that portion of the margin of the colony which lies over tiie mouth of the univalve
shell invested l)y it there occurs a set of very remarkable zooids (fig. 1, c, c, c). These are long-
cylindrical bodies, usually attaining, when fully extended, about two thirds of the height of the
ordinary liydranths when similarly extended. Their distal extremity is slightly enlarged, and
provided with a coronal of spherical clusters of thread-cells. Their axis is hollowed out into a
tubular cavity like that of the ordinary liydranths, but they seem to be entirely destitute of
mouth. They are provided with a strongly developed layer of muscular fibres between ectoderm
•and endoderai. Their most remarkable feature consists in their power of rolling themselves into
a spiral, whose coils are in vertical planes, somewhat recalling the circinate vernation of a fern.
On any irritation of the neighbouring parts they throv,' themselves into energetic and violent
action, uncoiling and coiling themselves, and lashing backwards and forwards on their basis.
They are quite constant in their occurrence. Among all the specimens of IL/dradinia I have
examined I have never met with one in which they were not present ; and it is difficult to
understand how Quatrefages and Van Beneden could have overlooked them. The first published
description of them is by Dr. Strethill Wright.

T//e Gonosome. — The gonophores are always borne upon blastostyles (PI. NV, fig. 1, 5, h, h,
and fig. 3). These are scattered among the liydranths without any apparent order. They are
cyluidrical, slightly dilated at the distal extremity, and for the most part about half the height
and thickness of the hydranth. The summit is provided with globular projections, which take
the place of the tentacles in the liydranths. These projections are arranged in two or thi-ee
alternating, closely approximated verticils. Each of them is formed by a cluster of thread-cells,
thickly set in a projection of the ectoderm. I have not satisfied myself as to the com))lete
absence of a mouth in the blastostylc. If, however, it be present, it is only in the state of a
minute, scarcely perceptible puncture, which may occasionally be witnessed on the summit of the
blastostyle, and through which Wright has seen the contents of the somatic cavity voided under
pressure ; but it can hardly ser>'c the i)urpnse of ingestion of nutriment, though Agassiz has

GEMMARIA IMPLEXA. 223

iiieiitioned the existence of a well-developed nioutli in the American species described hj him.
A fibrillated layer is developed, as in the hydranths and the spiral zooids, between the ectoderm
tmd endoderm, throughont the whole length of the blastostyle.

The gonophorcs are simple sporosacs and are borne near the distal extremity of tiic
blastostyle, usually in a closely aggregated cluster. Tliey are of an oval form, with a well-
tleveloped spadix. Tlieir walls (fig. G) consist of tw-o membranes, an internal thin membrane
(endotheca), in which no evident structure can be detected, and au external (ectotheca), with a
<listinctly cellular structure and containing abundance of thread-cells. No mesotheca is developed.

The ova are invested by a delicate but distinct vitellary membrane. They contain a very
large and evident germinal vesicle, within which a nucleated germinal spot may be easily detected
(fig. G). Each colony is absolutely unisexual.

Quatrefages mentions the occurrence of bodies which he regards as eggs, and which he has
seen escaping from the ccenosarc when this had been torn muler the microscope. I have never
noticed those bodies, and am unable to form any opinion as to their nature. We may compare
the observation of Quatrefages with Cavolini's account of egg-like bodies which he also found
escaping from the ccenosarcal cavity of his Ser///laria jjarasifica.

The hydranths, blastostyles, and spiral zooids arc constant, but besides these Dr. Wright
has described certain long filiform bodies which, like the spiral zooids, occur near the margin of
the colony. He calls them " tentacular polypes," and regards them as constant, and employed
by the colony in the prehension of food. They are, however, by no means constant, nor are they
even usually present, and I cannot consider them as otherwise than abnormal. They arc probably
malformations which some of the spiral zooids or some of the hydranths situated close to the margin
have uiulergone, from having been there exposed to conditions unfavorable to their development.

Dr. AA'right also describes as an essential part of the colony certain sporosacs similar to
those which are borne on the blastostyles, but which, instead of being so borne, spring directly
from the ccenosarc. I think that there is here some error. I have never met with these cceno-
sarcal sporosacs, while in a specimen given to me as affording an example of their occurrence I
could find nothing of the kind, the sporosacs which on a superficial view had the appearance of
arising directly from the ccenosarc proving on close examination to arise from the sides of very
short blastostvles.

GeM.MARIA IMPLEXA.

Plate VII.

Tliis beautiful httle hydroid may be found on the under side of stones at extreme low-
water mark, or it may be dredged up from the Coralline Zone, where it grows over the surface of
old shells. It is well deserving of a special study, for its medusa is in many respects one of the
most remarkable in the whole range of planoblasts.

T/ie iri/droj}//?/(on.— The hydrophyton is but little developed in proportion to the hydranth,
jind consists of short stems, very rarely sending off a branch, and with a creeping filiform
hydrorhiza (figs. 1, -2)}

^ A variety witli the liydrocaulus more developed and more decidedly branched would scera
occasionallv to occur.

224 ANATOMY OF SPECIAL FORMS.

The condition of the perisarc is peculiar in that part of the hydroculaus which lies
immediately below the hydranth (fig. 1). It here I)ecomes suddenly thin and transparent, and is
separated by a wide interval from the contained coenosarc. The ccenosarc, however, is kept in
connexion with it by regularly disposed, radiating offsets of its ectoderm. It would seem to be
this condition which both Alder and Wright have described as an inner annulated jjcrisarcal
tube, surrounded at an interval by an external smooth and thinner one. In my specimens no
internal tube of perisarc was apparent.

The Hydrantlis. — The hydranths are proportionately large, and when extended are nearly
cylindrical. They are very contractile, and when in extreme contraction assume the form of a
thick club. The tentacles are very numerous and scattered over the body ; they present the
usual septate condition of their endoderm, and terminate each in a spherical capitulum loaded
with thread-cells. Those hydranths which carry gonophores appear to be in every instance
somewhat less developed than the others, though never reduced to the condition of blastostyles.

The endoderm of the hydranth body (fig. 5) presents a peripheral zone of large prismatic
cells (c), and containing a clear colourless protoplasm, having their longer diameter transverse to
the axis of the hydranth, and an axial zone {d) of spherical and much smaller cells containing
pale reddish granules. Between the endoderm and ectoderm lies a very distinctly fibvillated
tissue {b). In the ectoderm («) no distinct cellular structure can be demonstrated, though
obscure indications of cells may become apparent under a high power and carefully adjusted
compression. Numerous thread-cells may be seen imlDcdded in it. Besides thread-cells of the
ordinary form, there occurs here a large oval form of thread-cell in which the coils of the
contained filament are very regular and distinct (figs. 9, 10). After evolution the cxscrted
portion is seen to present some curious and suggestive modifications of the form more common
in the Hydroida. These have been fully described in a former part of the present work
(see p. 119).

The Planoilasfs. — It is in the planoblasfs (fig. 3) that the most striking characters are to
be found. They spring from the body of the hydranth in a group near the proximal limit of the
tentacles. They arc phanerocodonic, and among the whole of the hydroid medusae I know of
none more remarkable than those of Gemmaria.

The umbrella is deep bell-shaped, with four radiating canals and two opposite marginal
tentacles, with large bulbous bases, destitute of ocellus, while the place of two other tentacles is
taken by a small bulbous dilation of the distal extremity of each of the intervening radiating
canals. The axis of the tentacle is traversed by an uninterrupted tube from its base to its
summit (fig. 4). Lying immediately over each of the four radiating canals, and running in
the walls of the umbrella close to the outer surface, is an elongated fusiform sac. It seems to
originate by its narrow end in the circular canal, and then running exactly parallel to the
radiating canal extends over about one fourth part of a meridian of the umbrella. It is filled
with thread-cells immersed in a clear flnid (fig. G). These singular sacs admit of a comparison
wdth the superficial ridges loaded with thread-cells which Gegenbaur describes as running from
the summit of the umbrella to the codonostome in his Zanclea cost at a ; while similar l)ut much
smaller sacs, also filled with thread-cells, are developed in the umbrella-walls of the little medusas
which constitute the genus Willia of Forbes.

Imbedded in the substance of the umbrella just within its margin are some oval thread-cells
with the contained filament in distinct spiral coils. They are disposed in a line parallel to the

GEMMARIA IMPLEXA. 235

circular canal (fig 3). During contraction four meridional farrows may be seen on the outer
siu-flice of the umbrella, alternating with the radiating canals, and extending from the summit to
the codononostonic.

The mannbrium extends through aljout one half the vertical diameter of the mubrella. It
is nearly cylindrical, destitute of oral tentacles or lobes, and having the month snrroundcd l)y a
circle of minute thread-cells.

A still more remarkable feature than tliat derived from the prescnci; of the thrcad-ccll-
l)earing sacs in the umbrella is atlbrdcd by the two marginal tentacles (tig. 3, 4). Along its entire
length from the bulbous expansion at its base to its tip each tentacle carries oval capsules filled
with thread-cells. They are thickly set along the whole length of the tentacle, each arising by a
])edunclc from its outer side. The capsule itself carries a pencil of long vibratile cilia on its
summit. Along the sides of its peduncle small spherical masses containing granules are
irregularly scattered. These appendages of the marginal tentacles would seem to have their
representatives in what appear to be very similar bodies figured by Gegenbaur in Zanclea.

Whether, however, the pedunculated capsules which are carried on the marginal tentacles
of Gcmmaria resendjle in all respects the tentacular appendages of Zanclea is a question which
remains open for future observation, but it is certain that these bodies in Gemmaria present a
phenomenon to which nothing similar has yet been recorded in any other hydroid.

Some years ago I dredged oft' the Forfarshire coast a colony of Gemmaria implexa, which,
after remaining for about a fortnight in a jar of sea-w^ater, threw off' its medusje. On examining
these with a hand-lens as they floated through the water I was struck by the appearance of a
mould-like growth with which many of them seemed covered, and for a moment I thought that
they had become infested by some low parasitic vegetation. That this, however, was something
very different from what it appeared to be soon became evident, for wdien the little medusa was
touched ever so slightly with the point of a needle the whole of the flocculent mass instantly
vanished.

It needed, however, a higher power of the microscope to reveal the true nature of this
phenomenon and show that the apparently parasitical growth consisted of the enormously
elongated peduncles of the thread-cell-bearing capsules, each of which, as it now proved,
had the po\ver wdiile still carrying the capsule on its extremity of extending itself to a
length wdiich considerably surpassed that of the longer or vertical diameter of the umbrella
(see fig. 3).

AVhile the medusa continued to float undisturbed through the water the peduncle would
remain projected in a straight line from the tentacle, becoming at the same time amazingly
attenuated, but on the least disturbance it would suddenly shorten itself to less than one
twentieth part of its length when extended, drawing the capsule back with it in its contraction.
Tlie extended condition of the peduncles was seen only while the medusa was passively floating
in the water, while engaged in active locomotion the peduncles were always contracted.

The small spherical masses Avhich are attached to the sides of the peduncle become widely
separated from one another on extension, and the whole peduncle with its terminal capsule was
seen during the extended condition to be in a state of constant vibration. This was due to the
pencil of long fine vibratile cilia which the capsule carried on its summit.

I believe that the remarkable phenomenon now' described presents us with a case of true
.sarcode extension and contraction, and that the peduncles of the thread-cell-bearing capsides

2-26 ANATO:\IY OF SPECIAL FOR^IS.

consist of a simple irranulnr protophisiii, wliicli in its power of extension nntl retraction at once
recalls the pseuclopodia of a rliizo[)0(l.

DlCORTNE CONFKllTA.

Plate YIII.

This hydroid may be found investing the surface of various univalve shells in the form of a
dense moss-like growth. It aflbrds the only example we yet know of a free locomotive
sporosac.

27ie Hydrajj/zj/foii and the Jhidranllis. — The hydropliyton (fig. 1) consists of branched stems
with occasionally some simple ones, all arising from a creeping rctiform hydrorhiza, the whole
invested by a rather coarse perisarc.

The hydranths (fig. 1) are fusiform, with a single circlet of filiform tentacles surrounding the
base of a conical hypostome.

The Gonop/iorcs. — It is the gonophorcs, however, which confer upon liicorync a special
interest, and eminently distinguish it from all other known hydroids. They are borne on true
blastostyles (fig. 1), which spring either from the hydrocaulus or from the hydrorhiza, and which
represent hydranths modified by the complete suppression of both tentacles and mouth. The
gonophores are in the form of an oval sac, and are densely crowded on the sides of the blas-
tostyle. The male and female clusters entirely resemble one another in external form ; they
are always borne on separate stems, but I have occasionally met with both male and female
stems in the same colony. While still attached to the blastostyle the gonophores (fig. 5) may be
seen to consist of an external sac (ectotheca), within which is a second sac (endotheca). Within
this second sac are the ova or spernuitozoa, while a long simple spadix occupies its axis. By
'carefully adjusted compression it will also be seen that just behind the proximal end of the
spadix two tentacula-like processes are given off. These run forward between the ectothecal
and endothecal sacs until they reach a point nearly half way between the proximal and distal ends
of the gonophorc.

On attaining maturity the endothecal sac and its contents arc ready to liberate themselves,
and the sac accordingly becomes detached from the summit of its peduncle along a defined line
which lies just behind the roots of the tentacular appendages, and, carrying these appendages
along with it, breaks through the endotheca and becomes a free zooid in the open sea (figs. 3, 4,
G), leaving the torn ectotheca behind it attached to the summit of the peduncle.

The planoblast thus set at liberty is a sexual zooid of an exceedingly singular kind, and, so
far as we yet know, without any exact parallel. Immediately on acquiring its freedom it throws
back its two tentacles which had been previoiisly turned forward as they lay impacted between
the ectotheca and endotheca of the gonophorc, and which, now becoming extended to at least
twice their former length, diverge from the proximal extremity of the liberated zooid.

This zooid is thus a free sporosac. It is an oval sac with a pair of tentacles diverging from
its proximal extremity. It is completely covered with vibratile cilia, which extend even to the
extremities of the two tentacles, and by their aid it swims actively in the surrounding water. A
long spadix runs through the entire axis of the sac impinging on its sununit and there remaining .

DICORYNE CONFERTA. 227

adherent to its wall. In the male (fig. C)) the spadix is siiiTounded by a contiiuious mass of
spermatozoa. In the female (figs. 3, 4) the place of the spermatozoa is taken by ova, which are
invariably two in number, and are situated one on each side of the spadix, and in such a position
that only one is visible when the zooid is viewed at right angles to the plane of the two tentacles
(fig. 3).

The endoderm of these tentacles is composed of large transverse cells, giving the usual
septate appearance to the axis of the tentacle, and by the careful employment of the compres-
sorium this structure may be seen to be continued for some distance into the walls of the
spadix (fig. 7), where, however, thft cells become less regular, while the cavity of the s])adix is
immediately surrounded by a layer of endodermal cells filled with reddish-brown granules. The
walls of the sac contain imbedded thread-cells (fig. 7).

The ovum presents a well-defined germinal vesicle, in which a germinal spot is visible, or
maybe easily rendered so by slight compression. Each ovum is invested by a proper membrane,
which presents the remarkable and unique character of possessing considerable thickness and
being richly set with thread-cell-like bodies (fig. 7).

It is plain, then, that the plauoblast of Bicoryne is a free sporosac, consisting of a simple
endotheca traversed by a spadix. To the base of the spadix the endotheca still adheres for some
distance, the generative elements not intervening here between spadix and endotheca so as to
separate them from one another. It is from the extreme end of this part that the two tentacles
are given off.

The plauoblast of Bicoryne admits of a very instructive comparison with an ordinary
racdusiform gonophore. It is, in fact, a medusa in which the place of the umbrella and its canals
is taken by two tentacles, the manubrium of the medusa being represented by the rest of the
plauoblast. It will be recollected that the two tentacles are turned forward, while the sporosac is
still invested by its ectotheca, and that they then hold exactly the place of an umbrella
(mesotheca) between endotheca and ectotheca. They ai'c, in fact, the radiating canals of the
medusa reduced to two and developed as free tubes, instead of being immersed in the walls of
an umbrella. It will also be borne in mind that in the medusae of Ohelia, the umbrella with
its canals is frequently inverted and thrown back, so as to assume the position of the tentacles
in the plauoblast of Bicoryne, after this body has broken through the ectotheca and become
free. The free sporosac of Bicoryne is thus a medusa reduced to the condition of a manu-
brium and two opposite radiating canals. In order, indeed, to convert it into an ordinary medusa,
little more is necessary than to suppose the number of the tentacles increased to four by
the STOimetrlcal development of two others, their extremities connected by a circular canal, and
their sides by a continuous muscular membrane (umbrella) inflected at its free margin so as to
form a velum.

It will be noted that the planoblast of Bicoryne is one of the three forms of locomotive
zooids which occur among the Hydroida, these three forms being the natatory medusa {Syncoryne,
Cladonema, Corymorplia, &c.), the ambulatory medusa {Eleitfheria, Clavatella), and the natatory
sporosac {Bicoryne).

The planoblast of Bicoryne conferta always swims with its body in a vertical position,
carrying the posterior or tentacular extremity uppermost, and maintaining all the time a constant
rotation on its longer or vertical axis.

30

228

SUPPLE MENTAEY NOTE.

STRUCTURE OF CORDYLOPHOKA.

Since the earlier part of the present ^Monograph was printed there lias appeared a valuable
Memoir by Dr. Schulze, of Rostock, on Cordylophora lacustris {' Ueber den Bau und die Entwicklung
von Cordylophora lacuslris nebst Bomerkungen Uber Vorkommen und Lebensweise dieses Thieres/
von Dr. Franz Eilhard Schulze, Leipzig, 1871).

Dr. Schulze obtained the Cordijlophora in brackish water near Rostock, and has made its minute
structure the subject of a laborious aud exhaustive examination. By the aid of osmic acid, and
steeping in Miiller's .solution and in iodized serum, he has succeeded in making sections eminently
fitted for examination under high powers of the microscope.

Perhaps the most important puint contained in the Memoir is his demonstration of a structure-
less hyaline membrane which lies between the ectoderm and the endoderm, and which, in those parts
where a muscular layer exists, is found on the inner or endodermal side of this layer. It is the
supporting lamina, " Stiitzlamelle" of Reichert, who recognised its presence in the same position in other
hydroids, though he confounded with it the muscular layer whose existence in the Hydroida he denies.

Though I have had no opportunity of verifying by new examinations the results at which
Dr. Schulze has arrived, I am quite ready to accept his confirmation of the presence of the " Stiitz-
lamelle'' of Reichert. I have been long aware of the appearance of a very narrow clear space between
the etidoderm and ectoderm, as visible in various sectional views of hydroids (see especially PI. IV,
fig. 3; PI. VII, fig. G ; PI. XXIII, fig. 9 ; and woodcut, fig. 48, p. 12, of the present work) ; but I
could never convince myself that this appearance ought to be regarded as the expression of an inde-
pendent membrane. Schulze's very careful investigation, however, appears to me now to set this
question pretty nearly at rest, and to justify us iu regarding the " Stiitzlamelle" as a second element
lying between ectoderm and endoderm, the muscular layer being the other.

Schulze informs us of an internal annular process of the "Stiitzlamelle," which runs transversely
across the base of the tentacle iu Cordylophora, forming a septum with a central aperture. I have
not seen anything of this kind.

He can in no case find either in Cordylophora or in Hydra the secondary cells which I have
described as existing in the interior of the large endodermal cells of the somatic cavity, and to which
I believe myself justified in attributing a secreting function. Notwithstanding, however, the trust-
worthy means of observation adopted by Schulze, 1 am not prepared to relinquish my belief in the
reality of these cells. It is not impossible that their existence may depend on certain states of
nutrition, and that they may be present at one time and absent at another.

The author is inclined to regard the free surface of the endodermal cells as destitute of membrane,
and believes that the protoplasm which is accumulated at this spot has its surface freely exposed to the
somatic cavity — an important and interesting fact if confirmed by subsequent observations. He has
also succeeded in clearly making out the presence of vibratile cilia over the whole of the endoderm of
the somatic cavity, not only in Cordylophora, but in Hydra, each of the cells carrying a single long
fine cilium on its free surface. Hydra, therefore, affords no exception to the ciliated condition of the
endoderm.

The occurrence of longitudinal rugae which I have described in the endoderm of the stomach in
Cordylophora and Hydra is attributed by Schulze to the mere contraction of the walls. There is
no doubt that the distinctness of these rugre depends to a certain extent on the state of contraction of
the hydranth, but their existence is not on that account the less real. la many marine hydroids the
endoderm of the gastric walls presents well-marked lobes and rugse.

The author maintains the ectodernud origin of the generative elements. In this I cannot agree
with him. I believe, on the other hand, that I have satisfactory evidence that they are products of
the endoderm (see above, p. 148). Schulze, it is true, affirms the existence of a continuation of the
hyaline " Stiitzlamelle" over the spadix, so that the generative elements would be thus separated from
the endoderm of the gonophore. If this be the case it would go far to prove the view taken by

SUPPLEMENTARY NOTE.

229

him, but with the strong evidence ou tlie other side I must suspend my belief in the correctness of
his interpretation of the appearances here presented by liis dissections.

Accepting, indeed, the statement of Schulzc, that the " Sliitzlamelle" can be traced over the
spadix, it will still, I believe, be found that the generative elements originate at the cndodermal side
of this membrane. In some instances, as in Laoinedca ca/iaiJaia, a hydroid with a branching spadix,
an extremely delicate, structureless membrane may, with careful illumination, be traced over the free
surface of the ova while these are still in contact with the spadix, and I now believe that it is the same
membrane wliich may, in some cases, be seen to be carried out, hernia-like, before the ova when these
are expelled under pressure from the gonojihore, and which I formerly regarded as probably representing
a condition of the vitellary membrane {see p. 64). Is not this the membrane which Schulze has traced
over the spadix, and which he regards as a continuation of the " Stiifzlamelle" ? If so, the ova must
have been produced between the " Stiitzlaraelle" and tlie spadix, and not upon the outer side of this
membrane, as maintained by Schulze.

THE GONANGIUM.

Among the hydroids in Mr. Busk's collection, which he placed in my hands for examination, is
a calyptoblastic species from New Zealand. Its trophosome, so far as its state of preservation in the
dried specimen allows of comparison, differs in no respect from that of a typical Sertularian, but its
gonosome presents the hitherto entirely unprecedented character of having its gonangia borne upon
peduncles, which spring from ivithin the hydrothec<g.

The gonangia are oval, opening at their distal extremity
by a tubular orifice, and ornamented by curved ridges which
terminate at each side in a zigzag line, which runs down the
middle of the gonangium walls from the summit to the base.
Each gonangium is borne on a long cylindrical peduncle,
which springs from the bottom of a hydrotheca, in which it
occupies the position of a hydranth. TIjc peduncle extends
through the whole length of the hydrotheca, and as it nearly
equals it in diameter it almost fills its cavity. It is covered
by a delicate chitinous perisarc, and immediately, on emerging
from the cavity of the hydrotheca, carries the gonangium ou
its summit.

Whether those hydrotheca; from which the peduncles of
the gonangia emerge ever carried hydranths, which subse-
quently became replaced by the gonosome, or whether they
have been all along exclusively devoted to the gonosome, it is
impossible to determine from dead and desiccated specimens. At all events, it is certain that there is
not a single point, either in position or in form, by which these gonangia-bearing hydrothecEe differ
from the others.

The hydroid thus so very exceptionally constructed must constitute the type of a new genus and
species, to which the name of Synthecium elegans may be given.

Sr/iifheciam elegans.
Ordinary hydrotheca destined for the reception
of a hydranth. a'. Hydrotheca in which the
liydrauth is replaced hy the peduncle of the
gonangium. 6. Gonangium.

FUNCTION OF THREAD-CELLS.

In the account given above (p. 108) of the alleged stinging property of the thread-cells, reference
ought to have been made to the researches of Mr. G. H. Lewes (' Sea-side Studies'), who has investi-
gated, both by observation and experiment, the thread-cells of various species of Actinia, with the
view of determining how far we are justified in attributing an urticating property to those bodies.

Mr. Lewes decides against their alleged powers of urtication. He supports his view on the
admitted fact that thread-cells are often situated in places where it is impossible for them to exert such
powers, as well as on the further observation that in their number and development they present no

230 SUPPLEMENTARY NOTE.

ratio to the stingiug powers of the animal. He believes, moreover, tliat experiment leads to the same
results, for he has allowed small liviug annelids and crustaceans to come in contact with the filaments
connected «itli the radiating septa of Actinia, and which are loaded with thread-cells, without finding
that the little animals so treated had thereby sufl'ered any harm. So also animalcules touched by the
filament ejected from the thread-cells in the tentacles of an Aciinia were not killed nor disabled by it.
Against the conclusions of Mr. Lewes there may be adduced the statements of other observers,
who describe tlie sudden arrest of all motion in the prey when touched by the thread-cell-bearing sur-
face ; and while it is true that the soft bodies of the annelids fit them for the experiment, it may be
objected to the employment of crustaceans that the resisting nature of their surface would enable them
to bear the exposure with impunity. Mr. Lewes, it is true, explains the arrest of motion in the cap-
tured prey by referring it to the habit of feigning death so well known in many insects and spiders,
but this can apply to only a very small number of the animals affected. That an actual penetration of
the thread-cell to a slight extent takes place I believe there is sufficient evidence, and the probability
that this is accompanied by the injection of a poison has been already maintained (see p. 129). Still,
however, with the investigations of so careful an observer as Mr. Lewes tending towards an opposite
conclusion, and supported as they are by those of jMobius and others, we cannot yet regard the alleged
stinging powers of the thread-cell as beyond doubt.

EVOLUTION OF THE HYDROTDA— ANCESTRAL TYPE.

Haeckel (' Generelle Morphologic,' vol. ii, p. li) regards the whole of the Ccehnterata as
derived fiom a hypothetical aboriginal form, to which he assigns the name of Archydra, and which
has continued itself with but little change down to our own times in Hydra and Cordylophora, the only
freshwater representatives of the Cwleiiterata. He supposes that these animals, in consequence of
the simple nature of the influences to which their fresh water habitat has exposed them in the struggle
for existence, have, like many other dwellers in fresh water, retained nearly their original simple
structure. With these persistent archydral forms he further unites those marine hydroids which come
nearest to Hydra in never giving origin to planoblasts, and in developing fixed sporosacs in place of
free locomotive buds. He regards Hydra as having for the Cmlenterata a significance similar to that
which Amphioxys has for the Vertebrata.

Under the name of Protohydra Dr. Greef, of Bonn (' Zeitschr. f. wis. Zool.,' vol. xx, p. 37, pi. iv),
descrii)es a remarkable little organism which he obtained among diatom-mud and algaj from an oyster
bed at Ostend. It is almost of microscopic size, but has a very close relation with Hydra, from
which it chiefly differs in never developing a trace of tentacles, and in multiplying itself solely by
spontaneous fission. It is very contractile, varying in form from that of a sphere to that of an
elongated club. It attaches itself to algae and other bodies by one extremity, in the manner of the
Hydra, and is perforated at the opposite end by a mouth which leads into the cavity of the body.

Its minute structure appears to be in all essential points similar to that of Hydra, and its body
is composed of a distinct endoderm and ectoderm with interposed muscular layer ; both endoderm and
ectoderm consisting of undoubted cells with nucleolated nuclei. The ectoderm contains abundance of
thread-cells, and the endoderm scattered reddish-brown pigment-granules.

The endoderm may be broken down under the microscope, and the protoplasmic contents of the
cells liberated by the rupture of tiie cell-walls. The liberated protoplasm will then often show amoebi-
form movements, and emit pseudopodial prolongations.

The spontaneous fission of the body takes place transversely, dividing the animal into quite
similar halves, each of which may again present successive fissions.

Dr. Greef calls attention to a very interesting phenomenon which presents itself duiing the
division. While the constriction which is to result in the complete separation of the two segments is
as yet slight, the movements of the dividing animal are, as may be expected, quite simple, and such
as might be imagined to follow from the volition of a single individual ; but as soon as the constriction
has attained a certain depth the form of the segments has become changed, and each now carries out
in itself its own movements, but yet without the mutual dependence of the still united segments being
destroyed, for the movements of each segment are perfectly conformable to those of the other, and
buite svuchronous with them. If the one stretches itself out or contracts itself, so does the other, in

SUPPLEMENTARY NOTE. 231

precisely the same way and during tlie same time, the shape of the one being always exactly that of
the other, until, finally, the complete separation of the two puts an end to their dependence.

Dr. Greef regards his Protohijdra as representing the lowest type of a coelcutcrate animal, and,
in accordance with the descent hypothesis, he assumes it as the true ancestral form of the whole of the
Coelenterata.

The unsuitableness of the Hydroida for preservation in a fossil state leaves us almost entirely
without direct evidence as to the forms which may have been presented by their remote ancestry.
Reasons, however, have been already given which would seem to justify our going as low as the
Protozoa for the ancestral form of the IIybroida, whose relations to the K/Uzopocla would seem to be
traceable through the graptolites. See above, p. 179.

PAET 11.

THE GENERA AND SPECIES OF THE
GYMNOBLASTEA.

THE GENERA AND SPECIES OE THE

GYMNOBLASTEA.

Genkkal Principles of Hydrotd Zoogrmmiy.

WniLE the first part of this Monograph is devoted to those general questions of hydroid
organization and Hfe whieh concern the IIydroida as a whole, I purpose in this second part to
deal with the zoography or descriptive and systematic zoology of the various genera and species
which are eud)raced under the Gymnoblastea, that great section of the Hydroida to which the
special part of the iMonograph must be confined.*

It is but quite lately that those princijjles of classification which are acknowledged as the
only sound ones, and which have been our guide in the study of every other group of the
animal kingdom, had begun to be recognised in the systematic treatment of the Hydroida.

The cause of this long exclusion of the Hydroida from tiie douiain of a philosophic
classification is sufficiently obvious. The individual hydroid, as we now know, frequently
presents itself in disconnected parts which are veiy different from one another, and it is only
recently that zoologists have shown the mutual relations of these parts, and have demonstrated
that organisms totally different from one another in form, and now enjoying an independent life,
may have been at one time united into a single individual of which they formed essential
constituents, and are at all times necessary for an adequate conception of it.

So long, however, had these component elements of the zoological individual been regarded
as entirely independent of one another, that even after their true relations became known
zoologists continued to find it more convenient to treat them as independent organisms, to
assign to them separate places in their systems, and to record them under distinct generic and
specific names.

A practice, however, so totally at variance with the first principles of a philosophic
classification and of a scientific nomenclature could not last. With the possible exception of
the MoNOPSEA, in which no hydriform trophosorae exists," the individual hydroid can only be

' See Part I, pages 188 — 191, for the systematic position of the Hydroida, and the leading
diagnoses of the larger groups.

^ I say, " with the possible exception," for even in the Monopsea we ought probably, as already
pointed out (see p. lOG), to recognise the existence of the two zooidal elements, tro[)hosome and
gonosome ; here, however, the medusiform condition of the trophosome and its permanent uuiou with
the gonosome will be sure to secure to both a full consideration iu any diagnosis.

Victor Carus, iu his excellent treatise on the Hvdkoida (Carus und Gerstaecker, ' Handbuch der
Zoologie,' vol. ii), has proposed the name of Haplomorpha for the group to which I have in the
former part of this Monograpli applied that of Monopsea. There is no possible objection to the name of
Haplomorpha; and were it not that I had accidentally overlooked its existence, I should not liave
ventured to substitute another for it.

31

236 PRINCIPLES OF HYDROID ZOOGRAPHY.

understood by regarding it as the product of two factors — one of them finding its expression in
the trophosome and the other in the gonosome ; and whether the gonosome remains permanently
attached to the trophosome, or becomes in whole or in part free, attaining thereby an independent
existence, it is equally necessary that it should take its place in our diagnoses of families,
genera, and species.

We accordingly now find that our leading descriptive treatises recognise the necessity of
combining in their diagnoses, not only those parts which are destined for the nutrition of the
colony, but those also which are destined for the sexual perpetuation of the species, whether
these last- mentioned parts be in the form of fixed sacs or iu that of free locomotive planoblasts.'

If all trophosomcs which presented no difference sufficiently great to justify generic
separation had their gonosomes just as closely resembling one another, while, on the other hand,
all closely resembling gonosomes had closely resembling trophosomes, the classification of the
Hydroida would be freed from one of its great difficulties.

This, however, is far from being the case, and the study of the Hvdroida renders us
acquainted with two sets of phenomena, which signally break down the uniformity assumed
in the above suppositions. These are — 1. The association of similar trophosomes icith dissiinilar
(gonosomes; and 2, the association of dissimilar trophosomes with similar gonosomes. The difficulty
which these phenomena throw in the way of a natm-al classification of the Hydroida may be
compared to that which the mineralogist meets with when he finds isomorphism and dimorphism
interfering with the uniformity of his mineralogical system.

But the great ditficulty, after all, in the application of the method here advocated is found in
the fact that the planoblast at the time of its liberation is still in an immature state, and may be
destined to undergo important changes before arriving at its adult condition. In such cases,
unless we have succeeded in following it to its ultimate form, our determination of its type must
be regarded as only approximate. Analogy, however, will greatly aid us in this determination,
by pointing out what are the parts most liable to change, and what the direction in which this
change is likely to take place.

From these considerations we learn that the number and form of marginal and oral tentacles
in the recently liberated medusa must be accepted with great caution, as aftbrding valid systematic
characters, these organs being especially liable to an increase in number, and often to an alteration
in form as the medusa advances towards maturity, while the form of the umbrella is also subject
to a certain amount of change.

In some cases, however, especially as regards the number of the marginal tentacles, we may
fairly assume the condition presented by the young medusa as representing its permanent
characters, as, for example, in the case of the single long tentacle in the planoblast of Corymorpha,
where we find, by going back to the early stages of the development of this planoblast, that the
peculiar asymmetrical form which in a later stage finds its expression so decidedly in the great
development of a single tentacle is quite apparent before any trace of a tentacle can be detected.

^ This principle has been recognised in the following publications :

Agassiz, ' Coiitr. Nat. Hist.' U. S. ; M'Crady, ' Gymnopbtbalmata of Cbarleston Harbour ;'
Victor Carus in ' Carus u. Gerstaecker Handbucb der Zoologie,' vol. ii ; Allman, " On the Construction
and Limitation of Genera among the Hydroida,'' in ' An. Kat. Hist.' for May, 18Gi ; Alex. Agassiz,
' lUiistr. Catal. of North American Acalcpbie;' Hincks, 'Brit. Hydr. Zoophytes.'

PRINCIPLES OF HYDROID ZOOGRAPHY. 237

In assigning their proper limits to the chissificatory groups under vvliich the gymnoblastic
hydroids are here distributed, I have accordingly availed myself of both tropliosome and
gonosome in the selection of the characters. As to the gonosome, I have regarded the condition
of the gonophores — whether phanerocodonic or adelocodonic, as well as certain important
differences in the form of the planoblast, as of primary importance, and have employed these
characters in the limitation of families. It is true that there are certain cases which may be
viewed as transitional from the adelocodonic to the phanerocodonic gonophore, yet the difference
between the medusa and the sporosac is, on the whole, so strongly marked as to afford valual)le
characters for the limitation of the higher gronps in any philosophical classification of the

HVDROIDA.

It may be objected that to regard the differences between phanerocodonic and adelocodonic
gonophores as of so high an order as to justify our employing them in the definitions of families
would involve the necessity of widely separating species in all other respects closely resembling
one another; that among the calyptoblastic hydroids, for example, the Campanularias, with their
phanerocodonic gonophores, and the Laomedeas, with their adelocodonic gonophores, would have
to take their places in separate families.' 1 am, nevertheless, prepared to defend such a
separation as would be here involved, for I do not, admit the principle that any agreement of
the trophosome can justify the depreciation of so very important a difference as is involved in
these two conditions of the gonosome. It has been already fully recognised in the comparison
of genera, and I believe we are quite justified in employing this well-defined and excellent
character as a limiting element in the diagnoses of families.

While the gonosome will thus afford characters for the diagnoses of the higher groups,
characters of corresponding value will also be yielded by the trophosome. These will be derived
chiefly from the condition of the tentacles as showing itself in their disposition, whether verticillate
or scattered ; and in their form, wliether filiform or terminating in capitula.

Characters of subordinate importance available for the diagnosis of Genera are also derivable
from both gonosome and trophosome. The gonosome will yield them in certain subordinate
differences of form in the planoblast, and in the presence or absence of a blastostyle ; while in
the trophosome we shall find them in such features as are afforded by the presence or absence of
a developed hydrocaulus and by certain minor differences in the disposition of the tentacles of
the hydranth.

Characters which are merely Specific will be found in still more subordinate differences,
which are also presented both by gonosome and trophosome, and will easily suggest themselves
to the observer.

With regard to nomenclature, I am convinced that except in cases where a manifestly
incorrect determination has been made, and in a few other special cases, we must retain for our
hydroid the name under which it was first described, whether this original description refers to
the gonosome or to the trophosome. The fact of our assigning to the complete hydroid the
generic name by which the planoblast alone had been previously known, needs not prevent our
continuing to employ the same name for all those plauoblasts, which do not differ from this in
characters of generic importance, but whose trophosomes have not yet been discovered ; we

^ I refer to the genera Campanularia and Laomedea in the sense in \yhich I have already
defined them, and which I see no reason to alter. ' On the Construction and Limitation of Genera
among the Hvuhoida,' 'Ann. Nat. Hist.' for May, 1861.

238 PRINCIPLES OF HYDROID ZOOGRAPHY.

must keep in mind, however, that the name, when used in this sense, is purely provisional and
liable to be clianged when the discovery of the trophosome shall determine the true geims of
our, then no longer incomplete hydroid. So also no name which has been given to a tropho-
some whose gonosome is unknown can be regarded as otherwise than provisional.

It is upon these principles that I have based the diagnoses, nomenclature and arrangement,
of the families, genera and species, of hydroids, to whose description the second part of the
Monograph is devoted.

Keeping in mind that it ought to be the aim of the systematist, in framing his diagnosis, to
render it al^solutely exclusive of every object or group of ol)jects to which it is not intended to
apply, I have endeavoured in every case to carry out this principle. With the view also of secur-
ing against difFuseness, it has been my aim in the definitions of the more special groups to avoid
a repetition of those characters on which the more general groups are based. In other words, I
have endeavoured to confine each diagnosis to its proper " differentia." It is true that the con-
ception of any group involves, not only the " differentia '' of the group, but also the characters of
those more general groups imder which it is included. These characters, however, instead of
being repeated at length, may in our technical diagnosis be expressed by the use of a single word —
that which has been adopted as the name of such higher group as immediately includes the
group under definition. This word would, then, like an algebraical expression, become a short
symbolic representation of a multitude of distinct facts, the facts which constitute the essential
characters of the group of which it is the accepted name, and which is supposed to have been
previously defined ; and though it may not be deemed necessary actually to express this word in
the definition, it or rather the facts it represents must in every case be understood.

In order to facilitate comparison I have adopted as far as possible a uniformity in the selec-
tion of characters, and in the order in which these characters are noted, while in every diagnosis
I have given one paragraph to the trophosome and another to the gonosome.

When the colour of the hydroids composing a species is known I have always given it. Colour
affords a character which cannot be neglected in our specific descriptions, but as it is in many
cases too variable to be relied on with certainty I have not ventm-ed to introduce it into the
proper diagnosis.

The technical terms employed in the descriptions are not numerous, and, if these be not
immediately understood, there will be no difficulty in rendering them so by a reference to the
section on Hydroid Glossology (see part I, pages xiii, &c.), where every term is defined. By the
use of such terms I have hoped to avoid tedious circumlocution, and to condense the descriptions
without sacrificing their precision ; and if in some cases the definitions may appear more diffuse
than could be desired, they do so because I believed that condensation in such cases could not
have been carried further consistently with exactness. The beautiful simplicity and terseness of
the Linnsean definitions is scarcely possible any longer ; for the vast increase of known species
since the days of the great systematist renders necessary the employment of a very much larger
number of characters for exact diagnosis than sufficed at a time when comparatively few species
had to be distinguished from one another.

The follow ing table will show the mode in which I have distributed the genera of gynnioblastic
hydroids under their respective families : —

SYNOPSIS OF FAMILIES AND GENERA. 239

SYNOPSIS OF THE FAMILIES AND GENERA OF THE
GYMNOBLASTEA.

HYDROIDA GYMNOBLASTEA.

Frm. I.— CLAVID.E.
Clava, Rhizogeton, Cordylophora, Tdbiclava, Merona.

Fam. II.— TURRID/E.

TURRIS, CaMPANICLAVA, CORYDENDRIUM.

Fam. III.— CORYNID^.

CORYNE, ACTINOGONIUJI.

Fam. IV.— SYNCORYNID.E.
Syncoryne, Corynitis, Gymnocoryne, Gemmaria.

Fam. v.— DICORYNID.f].
Dicoryne.

Fam. VI.— BBIERID.E.

Garveia, Bimeria, Wrightia, IIydrantiiea, Stylactis, Heterocordyle, Cionistes.

Fam. VII.— BOUGAINVILLID.E.

BOUGAINVILLIA, DiPLURA, PeRIGONIMUS.

Fam. VIII.— EUDENDRID^.

EUDENDRIUM.

240 SYNOPSIS OF FAMILIES AND GENERA.

Fam. IX.— HYDRACTINID^.

HiDRACTINIA.

Fam. X.— PODOCORYNIDJ^..

PODOCORYNE, CORYNOPSIS.

Fam. XL— CLADONEMIDiE.
Cladonema.

Fam. XII.— NEMOPSID.^.

Nemopsis.

Fam. XIII.— PENNARIDiE.
Pennaria, Halocordyle, Staurididm, Vorticlava, Heterostephanus, Acharadria.

Fam. XIV.— CLADOCORYNIDyE.
Cladocorynk.

Fam. XV.— ]\IYRIOTHELID/E.

Myriothela.

Fam. XVI.— CLAVATELLID/E.

Clavatella.

Fam. XVIL— CORYMORPHID^.

CORYMORPHA, HaLATRACTUS, AmALTH^A, AcAUIJS.

Fam. XVIIL— MONOCAULID^.

MoNOCACLCS.

SYNOPSIS OF FAMILIES AND GENERA. 241

Fam. XIX.— TUBULARID/E.
Tdbdlaria.

Fam. XX.— HYBOCODONID^.
Htbocodon, Ectopleura.

Provisional— LKMDJE.
Lar.

Ill the following descriptive zoology of the Gymnoblastea the species which have not yet
iiecn foiuul voiuid the shores of the British Isles are indicated by a tripple asterisk.

242 CLAVID^.

CLAFIB^.

TROPHOSOME. — HiDROCAULUS rudimental or developed. Hydraxtus, with
scattered iiliform tentacles.

GONOSOME. — GoNOPHOUES fixed Sporosacs.

CLAVA, Gmelin.
Name. — From Clava, a club, in allusion to the form of the hydranths.

CoRYNE, Lamarck.

TROPHOSOME. — Hydrocaulus rudimental, and consisting of very sliort simple
tubular processes from the free surface of a HYDRORniZA which is composed of
creeping tubes, either distinct or adnata to one another by their sides, and invested, as
well as the rudimental hydrocaulus, by an obvious perisarc. Hydranths claviform.

GONOSOME. — Sporosacs springing from the body of the hydranth at the proximal
side of the tentacles.

In the thirteenth edition of the ' Systema Natura?,' published in 1788, the genus Clava is
constituted by Gmelin, the editor of that edition, for a hydroid which had been described by
O. F. Miiller, in a paper published in 1775, in the ' Beschaftigungen der Berlinischen
Gesellschaft naturforschender Freunde.' Miiller gave no name to the hydroid there described,
but under that of 7/y(/r« s^««>««/a he described and figured in the ' Zoologia Danica,' 1777, a
nearly alUed species.

Lamarck, in his 'Animaux sans Vertebres,' omits the genus Clava altogether, and refers
Midler's hydroid to a genus to which it has no claim — the Corj/iie of Gartner.

The genus Coryne, with the extended limits thus assigned to it by Lamarck, was subse-
quently subjected to a revision by Sars, who, however, instead of restoring the genus Clava to its
rightful place, retains the name of Coryne for the forms properly included under Clava, and
institutes a new genus of his own, under the name of Stipula, for the true Corynes of Gartner ;
while Ehrenberg follows Sars, objecting only to Sars' name of Stipula on the grounds that it
belongs to the botanist, and substituting for it one of his own, namely, Si/ncoryne.

CLAVA SQUAMATA. 243

The first to restore the name of Clava to its legitimate phice, after its long banishment from
the nomenclature of the IIydroiua, was Johnston, who has thus not only performed an act
of justice to the memory of Gmelin, but has removed no slight element of confusion from the
classification of the HvDRomA.

1, Clava squamata, 0. F. Milller.
Plate I.

ZoOPHYTON MINUTUM CoRYN^ SIMILLIMUM, Pcillas, Spicil., fasc. X, p. 36, pi. Iv, fig. 9.

Hydra squamata, — Otho Fred. Mliller, Zool. Dan. Prod., 278G. Zool. Dan. Icon., tab. iv.

Otho Fabricius, Fauna Groenlandica, p. 347.
CoRYNE SQUAMATA, — Lamarck, An. s. vert., 1816, II, 63. Sleenstrup, Hermaphr. i Nat.,

pi. i, figs. 17 — 31.
CoRYNA MULTicoRNis, — Ehrcnberg, Corallenthiere, Abliandl. Berl. Akad., 1832, p. 293.
Clava multicornis [in part], — Johnston, Brit. Zooph., 1847, p. 30.
Clava membranacea, — Strethill Wright, Proc. Roy. Phys. Soc. Ediub., vol. i, p. 228,

pi. X, figs. 2 and 3.
Clava cornea, — Strethill Wright, Proc. Roy. Phys. Soc. Edinb., vol. i, p. 228, pi. xi, fig. 4.
Clava squamata, — Hincks, Brit. Hydr. Zooph., p. 4, pi. i, fig. 2.

TROPHOSOME. — Hydrocaulus, about one twentieth of an inch in height, consist-
ing of minute, simple, closely aggregated, tubular offsets from the surface of the hydro-
rhiza. Htdeorhiza formed of closely approximated inosculating tubes, united to one
another along their sides by an extension of their perisarc, so as to form a continuous
basal expansion. Htdranths very much elongated, somewhat fusiform between the
rudimental hydrocaulus and the club-shaped head, when fully extended attaining a
height of about one inch, closely approximated at tlieir base, so as to form a tassel-
like cluster ; tentacles about twenty.

GONOSOME. — GoNOPiioRES in clusters springing from the body of the hydranth
immediately behind the proximal tentacles, each cluster carried upon a very short
peduncle.

Colour. — A clear yellowish-red, with pale hyaline tentacles.

Develojiment of Gonosome. — April to September.

Habitat. — On Fuci, especially Fitciis serratus, Fucus vesiculosus, and F/fcus nodovus, on timber-
piles, &c., in estuaries and sheltered bays.

Bathymetrical dislrihution. — Literal zone.

Localities. — Shores of Denmark and Norway, 0. F. Miiller ; coast of Greenland, Otho
Fabricius ; sea-coast near Norwich, Pallas. Generally distributed round the shores — especially
the more northern ones — of the British Isles.

32

244 CLAVA SQUAMATA.

The figures given l)y 0. F. iMiiller, on the fourth plate of his ' Zoologia Danica/
are so characteristic as to leave no doubt regarding the animal to which the celebrated
Danish zoologist has assigned the name of Hydra squamata, and must remove all hesitation as
to the identity of jMuller's species with that described above, though a nearly allied species
described a few years previously by Forskal, under the name of H^dra muUicornis, has been con-
founded with the present form, not only by Miiller himself, but by most subsequent writers.

The first, however, who gave a description of the present species, as has been pointed
out by Johnston, was the celebrated Pallas. He discovered it in England, on the shore near
Harwich, and under the designation of " zoopliyton minutum corynse simillimum," has recorded
it in tiie tenth fasciculus of his ' Spicilegia Zoologica,' published in 1774, accompanying
his description with an indifferent figure.

That the species which Pallas had in view is identical with the Clava squamata of the
present Monograph is evident from his description of it : — " Pedunculus hujus zoophyti mollis
est intestiniformis, subannulatus eo(\we c/rec/alim dence pluresve simid, passim fucis adherens."

Pew hydroids equal in beauty and interest this fine species, and most students of the lower
forms of marine life will readily accept the estimate of it by the famous author of the ' Zoologia
Danica :' — " Aninialium qua3 zoophyta dicuntur, nullum elegantius, observatorique majus gratum
esse potuit.'' Attached to the fronds of the olive-coloured seaweeds, it may be seen at half-
tides with its clear Venetian-red hydranths, forming soft, flexuous, and pensile clusters, which
float passively in the surrounding water as they yield to every motion of the ebbing or flowing
current.

Its proper station being from half-tide to near low-water mark, it may remain for several
hours exposed upon the shore to the air and sun along with Coryne pusUla, Laomedeajlexuosa,
and Sertularia pumila, which, like it, love the higher regions of the tide range. The hydranths
are then much contracted, and lie close to one another, so that the whole cluster, looking like
a round, fleshy mass clinging to the seaweed, is enabled to retain sufficient moisture until the
returning tide brings the sea again within its reach ; and then the hydranths once more stretch
themselves out to their full length, and the entire colony again expands itself in all its beauty
beneath the flowing water.

In every specimen which I have examined in its fully extended state the hydranth, at the
l)oint where it springs from the little tubes which form the rudimental hydrocaulus, is much
attenuated; soon afterwards it increases in diameter, and then again slightly thins away before
it ultimately enlarges into its club-shaped head. Throughout the whole there extends a
continuous gastric cavity, whose width varies from distance to distance with the varying diameter
of the hydranth, and which, during the contraction of the hydranth, is thrown into close zigzag
folds still visible through the surrounding tissues.

Clava squamata is in perfection dm'ing the whole of the summer months, and then the
heavy clusters of gonophores, grouped round the base of the club-shaped head of the hydranth,
add greatly to its singularity and beauty. It is strictly moncecious, each tuft-like colony giving
rise to gonophores, which are all exclusively male or all exclusively female.

The adult hydranth in its extended state sometimes shows a slight enlargement of the
extreme points of its tentacles, as if these manifested a tendency to terminate in the capitula
so characteristic of Coryne and certain other hydroid genera. The terminal enlargement of
the tentacles, however, in Clava is entirely transitory and dependent on a particular state of

CLAVA SQUAMATA. 245

contraction. Tt must not, therefore, be confounded with the capitula of Corijne, which arc not
only independent of the state of contraction, Ijut possess a structure different from that of the
rest of the tentacle. Heterocordyle Conyhearei and some other hydroids with filiform tentacles
show, like Clava squamata, a tendency to the terminal enlargement of the tentacles in certain
states of contraction.

One of the most distinctive features of Chiva squamata is to be found in the peciUiar cceno-
sarcal expansion which forms a hydrorhiza, by means of which the whole cluster becomes united
into a compound colony, and rooted to the surface which supports it (fig. 1).

Instead of being formed, as in the majority of IIvdroida, by a system of separate fililbrm
tubes, tlie component tubes of the hydrorhiza are here adnate along their sides, adhering to one
another by their very thin chitinous pcrisarc, so as to constitute a continuous expansion, which
is formed by sinuous anastomosing tubes, from whose free surface the rudimental stems, with the
little cup-like investment of perisarc, are given off at such short intervals that the hydranths
which they carry are closely crowded upon one another at their base.

The chitinous tubes of the hydrorhiza and hydrocaulus have been described and figured by
Dr. Strethill Wright, in what he regards as two distinct species, and names Clava memhranacea
and Clava cornea} but which are both identical with the H^dra sq/iainata of Midler, the Clava
squamata of the present Monograph. The first, however, to draw attention to the fact that Clava
is not properly a naked hydroid, as had been previously asserted in descriptions of this genus,
was Leidy, who has shown" that an American species, which he names Clava multicornis, but
which Agassiz refers to his Clava leptostyla, is invested at its base by a distinct perisarcal tube.
A similar character is presented by all the other species of Clava.

The development of the ovum in Clava squamata may be followed without much difficulty.
The female gonophores usually carry each a single ovum (figs. 3, G). I have occasionally met
with two ova in the gonophore, but never with more. The ciliated planula (figs. 7, 8) is of an
elongated club-shape and of a deep red colour. It is very contractile, and may sometimes be seen
stretched out in a straight line, sometimes more or less coiled on itself, occasionally to such an
extent that the two ends are brought into contact. After it has lost its cilia and become fixed
the first tentacles are produced in a verticil of four near the distal extremity (fig. 10). Imme-
diately below these, and alternating with them, another verticil of four tentacles is then deve-
loped, and then successive tentacles are thrown out behind those previously formed; but all trace
of any distinct verticillate arrangement soon becomes lost.

In the mean time, short fleshy processes (fig. 11) — which afterwards multiply, ramifj^, and
coalesce, so as to form the hydrorhiza — extend themselves from the base of the young hydroid, and
from these new hydranths are budded forth, and thus give rise to the clustered colonies of the adult.
At a very early period — shortly after the appearance of the first tentacles — a very delicate struc-
tureless pellicle may be traced over the body of the hydranth, from its base almost to the roots
of the tentacles (fig. 11). During the life of the hydroid this continues upon its body as a
scarcely perceptible film, but at the base it becomes thickened by new layers, so as to con-
stitute the tubular perisarc of the hydrorhiza and of the rudimental hydrocaulus.

1 Wright, in ' Proc. Roy. Phys. Soc. Ediub.,' 1857, p. 228, pi. xi, figs. 3, 4.

" " ^Marine Invertebrate Fauna of Rhode Island and New Jersey," iu ' Journal of Acad. Nat. Sc.
of Philadelphia,' vol. iii, second series, 1855, pi. ii, figs. 33, 34.

246 CLAVA MULTICORNIS.

Clava squaiiiata loves tlie more sheltered bays and estuaries of our coast ; and I have never
found it in situations exposed to the surf of the open sea. It is not uncommon in the more shel-
tered parts of Dublin Bay, and is abundant in most of the fiord-like inlets of the sea which
characterise the western shores of Scotland. It occurs also in abundance on the shores and rocky
islands of the upper parts of the Firth of Forth, where I have met with it in great perfection about
forty miles from the open sea, at Craigflower, the seat of Sir J. Colvile, a point where the fresh
waters of the river have not yet ceased to exert their influence on the life of the estuary.

It is a hardy species, and may be kept alive for many weeks in our aquaria.

2. Clava multicoenis, Forshal.
Plate II, figs. 1, 2.

Hydra multicornis, — Forskal, Descript. Animal., 1775, p. 131, and Icones, tab. .xxvi, fig. B b.

CoRYNE MULTICORNIS, — Lumarck, An. s. Vert., 1816, vol. ii, p. 62.

CoRYNE SQUAMATA, — Van Benecleu, Mem. siir les Tubulaires, p. 60, pi. v.

Clava multicornis, — Johnston [in part], Brit. Zooph., 1847 (pi. i, figs. 1, 2)? Leidy,
Marine Invert. Fauna of Rhode Island and jS^evv Jersey, p. 3, pi. xi,
figs. 33, 34.1 Hhicks, Brit. Hydr. Zoopli., p. 2, pi. i, fig. 1.

Clava repens, — Wright, Proc. Roy. Phys. See. Edinb., 1857.

Clava discreta, — Alhnan, Ann. Nat. Hist., 1859.

TROPHOSOME. — Htduocaulus consisting of minute tubes, about l-20th of au
inch in heiglit, distributed at distinct intervals over the hydrorhiza. Htdrorhiza
consisting of a ramified, tubular, creeping filament, wliose branches are not adherent
to one another. Htdraxths about a quarter of an inch in height, not clustered ;
tentacles in well-grown specimens, twenty or more.

GONOSOME. — ^GoNOPHORES forming very shortly pedunculated, almost sessile
clusters, wliich are aggregated on the body of the liydranth, immediately behind the
proximal tentacles.

Colour of hydranth and gonophores varying from very pale brown to light red.

Development of Gonosome. — April to September.

Habitat. — Creeping over stones, sea-weeds, &c., near low-water mark.

Bathymetrical distribution. — Laminarian zone and lower limit of Litoral zone.

Localities. — Coast of Denmark, Forskrd ; New Jersey, Leidy ; Firth of Forth, Dr. Wright ;
Firth of Forth, Firth of Clyde, Orkney and Shetland Islands, and various other parts of the
British Isles, G. J. A.

' Leidy's Clava multicornis is regarded by Agassiz as identical with his own Clava leptostyla. I
can see nothing, however, in Leidy's figure or description to distinguish it from the Clava multicornis
of Forskal.

CLAVA DIFFUSA. 247

The species here described appears to be that which Forskiil had in view when describing
his Hydra mnllicortm, which is phiinly a different species from the Hydra squamata of Midler
and nmst be regarded as a scattered rather than a cUistered form of the genus Clava.

The two forms, liowever, continued to be confounded by most subsequent writers. Joiuiston
describes them both under the name of Coryne squamata in tlie first edition of his ' Britisii
Zoophytes,' 1S38, and imder that of Clava mulficornis in the second edition, 1847, while his
figures, when they refer to a Clava at all, represent a scattered rather than a clustered form.

The first, however, who described with sufficient fulness a scattered as distinguished from
a clustered species of Clava was Wright, who, under the name of Clava rcpms, described a
scattered species from the Firth of Forth. Not being aware of Wright's paper, the same species
was shortly afterwards described by myself under the name of Clava discreta. I now believe
that this species is identical with the Hydra multicornis of Forskal, and that it must therefore be
recorded under the specific name assigned to his hydroid by the Danish naturalist.

Forskal's description is undoubtedly obscure, but I entirely agree with Mr. llincks's criticism
which leads him to refer the present species to the Hydra multicornis of Forskal.

Clava midticornis is a much smaller and less conspicuous species than Clava squamata, from
which it is at once distinguished by its scattered habit, being never clustered like Clava squamata,
and by its filiform hydrorhiza, which never forms as in the last-named species a continuous
expansion by the lateral adhesion of its branches to one another.

It attaches itself to stones much more frequently than is the case with Clava squamata,
which is almost always found on fuci or on timber piles between tide marks, and its bathynietrical
area is lower than that of this species. Its favorite resort is the under surface of detached
stones near low-water mark in situations but little exposed to the surf of the open sea. It would
thus seem to be a light-shunning animal — a habit in which it contrasts strongly with Clava
squamata.

3. Clava diffusa, Allman.
Plate II, figs. 3, 4.

Clava diffusa, — Allman, Ann. Nat. Hist, for January, 1863.

TROPHOSOME. — Hydkocaulus rising to a height of ahout half a line from a
creeping filiform HrDBORiiiZA. Htduanths slender, from one quarter to half an inch
in height ; tentacles about twenty.

GONOSOME. — GoNOPUOEES scattered singly and in small clusters upon the body
of the hydranth, along which they extend for some distance behind the proximal
tentacles.

Colour of hydranth and gonophorcs, light rose-colour.
Development of Gonosome observed in July.

248 CLAVA LEPTOSTYLA.

IlahUat. — Growing over the bottoms of exposed rock-pools.
Ballnj metrical distribution. — Laminarian zone.
Locality. — Shetland Islands, G. J. A.

This pretty little Clava comes very near to Clava multicornis — so near, indeed, that doubts
might at first be entertained as to the justice of regarding it as a distinct species. The scattered
condition of its gonophores, however, oflers a character which cannot be set aside. It is true
that in at least one other species of Clava, in which the gonophores are habitually clustered
immediately behind the proximal tentacles, there may occasionally be met with individuals in which
the clusters of gonophores are more or less separated from one another, and carried back for some
distance from the tentacles. This is the case in the Clava leptosti/la, Agas., of which a woodcut
representing this condition is given by Agassiz,' but I know of no case in which the segregation
of the gonophores is carried to such an extent as in the present form, in which the clusters are
not only widely separated from one another, but in which many may be seen completely broken
up into their component gonophores, which are then scattered singly upon the body of the
hydranth. It must also be noticed that this peculiarity was not confined to one or two
individuals, but was presented by every gonophore-bearing hydranth of the colony. If to this
character we add a greater slenderness of the hydranths and their delicate rose-colour, and
further take into consideration the difference in habitat, the hydroid just described being met
with at the lowest spring tides, when it is found upon the bottom of shallow rock-pools exposed
to the roll of the open sea, we shall find reason to justify us in regarding it as specifically
distinct from all other described Clavas.

It was upon the most exposed shores of the Out Skerries, a small rocky cluster at the
extreme east of the Shetland group, that our little clava was obtained. It occurred in shallow
rock-pools, overgrown with Lamitiaria. Occasionally some of the hydranths were found with
atrophied tentacles, as if showing a tendency to become converted into blastostyles.

4. Clava leptosttla, Agassi:.

Clava leptostyla, — Ayassiz, Contr. Nat. Hist. U.S., vol. iv, p. 218, pi. xx, figs. 11 — 16, and
pi. xxi. Also fig. 32, on p. 222. Alex. Agassi-, Illustr.
Catal., p. 170, fig. 274. Hinclcs, Brit. Hydr. Zooph., p. 6.

TROPHOSOME. — IlTDKORniZA consisting of a basal expansion formed by closely
approximated and coherent tubes, which become distinct only at the extreme
margin of the colony. Hydranths much attenuated for some distance from their
proximal end, where they spring from the rudimental hydrocaulus, and then, suddenly
increasing in thickness, continue cylindrical or slightly tapering to the base of the
clavate head ; tentacles from twenty to thirty-five in number.

^ ' Contr. Nat. Hist. U. S.,' vol. iv, p. 222.

CLAVA NODOSA. 249

GONOSOME.— GoNOPiiORES in shortly pedunculated clusters aggregated imme-
diately beliind the posterior tentacles, or occasionally with the clusters detached from
one anotlier and extending for some distance backwards.

General Colour — Red.

Development of Gonosome observed in June and July.
Habitat. — On Fucus vesiculosus.
Batli^metrical distribution — Literal zone.

Locality. — Massachusetts Bay, Professor Agassiz ; jMorecambe Bay, Lancashire, Mr. West
(teste T. Hincks).

The Clava leplostyla of Agassiz is so closely allied to the European Clava squamata that its
recognition as a separate species is by no means obvious. Its distinguishing characters would
seem to lie in its greater number of tentacles, and especially in the more cylindrical form of its
hydranths, which thus contrast with the more fusiform hydranths of Clava squamata. Mr.
Hincks believes that he has identified with it specimens of a Clava obtained from Morecambe Bay,
and on the strength of this identification I have retained it among British species. I have never
seen it, and the diagnosis given above is compiled from characters selected from the description
published by Agassiz.

5. Cl.4.va nodosa, Slrelhill Wright.

Clava nodosa, — Wriyht. In Proc. Eoy. Phys. Soc. Ediiib., for 18G2.

TROPHOSOME. — Hydrouhiza in the form of delicate threads, which "at intervals
twine themselves into a convoluted knot of membranous tubes, from which a single
polyp arises." — Wright.

GONOSOME.— ?

Colour. — •" Aurora red."

Habitat. — On the fronds of DeJesseria sanyuinea.
Locality.— Y\x\\\ of Forth, Dr. Wright.
Bathymetrical distribution. — Laminarian zone.

I have never met with this hydroid, and all I know of it is from Dr. Wright's short
description. The knotted condition of the hydrorhizal tubes is certainly very peculiar. A fuller
description of the entire hydroid, however, is much to be desired, more especially as no figure
accompanies the published account of it.

250 RHIZOGETON FUSIFORMIS.

EHIZOGETON. Agassiz.

Name. — From 'pi^a, a root, and •ya'ra.i', a neighbour ; so called from the situation of the
sporosacs on the hydrorhiza.

TROPHOSOME. — Hyduocaulus evanescent. Hydhouhiza a creeping tubular
stolon, invested by a delicate perisarc. Hydranths sessile on the hydrorbiza,
elongated, subcylindrical, carrying towards their distal extremity scattered filiform
tentacula.

GONOSOME. — Sporosacs springing from the hydrorbiza.

Rliizoyetoii is one of the new genera described by Agassiz in his ' Contributions.' It has
certain affinities witii Clava, from which it differs chiefly in its gonosome, for the gonophores arise
separately from the hydrorbiza, while in Clava they spring from the body of the hydranth.

Only one species has been as yet described.

*^* Ehizogeton fttsiformis, Agassiz.

RaizoGETON i'usiFOR5iis, — Ac/assiz, Contr. Nat. Ilist., U.S. vol. iv, p. 224, pi. xxii, figs. 17 —

23. Clark, Mind in Nature, p. 73, note, figs. 38, 39.

TROPHOSOME. — Hydranths attaining a height of from one eighth to one quarter
of an inch, invested by a delicate filmy pellicle, tapering distally to a blunt point ;
tentacles about twelve in number, borne by nearly the whole of the distal half of the
hydranth.

GONOSOME. — Gonophores oval, elevated on peduncles, which arise singly from
the hydrorbiza ; both peduncles and gonophores invested by a filmy perisarc.

General Colour. — Orange.

Development of Gonomme observed in INIay.

Habitat. — In rock-pools between tide marks.

Bathytiictrical dhtrihiition. — Litoral zone.

Locality. — Massachusetts Bay, Professor Agassiz and Professor 11. J. Clarke.

CORDYLOPHORA. 251

The description and figure of Ithixogelon fimforiim given by Agassiz arc those of male speci-
mens only, but Prof. 11. J. Clarke, who had an opportunity of observing both sexes, has described
and figured female colonies, lie has noticed that the ova, after the earlier stages of their
development have been passed in the interior of the gonophore, break through its walls, and
then remain for some time confined between the outer ectodermal layer of the gonophore and its
perisarcal investment.

Agassiz, from whose descriptions I have selected the characters out of which the diagnosis
just given has been constructed, informs us that the male gonophore becomes converted into a
hydranth after the discharge of its contents. This converting of a gonophore into a hydranth
has been already referred to in the former part of the present monograph,^ where it is I'cgardcd
as an abnormal phenomenon.

COEDYLOPHOEA, Allnian.

Name. — From /copSuXij, a club, and <popUo, I bear ; in allusion to the form of the trophosome.

Syncoryne, — Agassis.
(?) TuBULARiA, — Affard/i.

TROPHOSOME. — Htdrophyton consisting of a well-developed ramified htdeo-
CAULTJS, which springs from a creeping filiform HTDEOimiZA ; the whole invested by a
PEEiSARC. Htdranths fusiform.

GONOSOME. — Sporosacs borne on the hydrocaulus.

The genus Cordylophora was instituted by me many years ago " for a tubularian hydroid
singularly exceptional in its mode of life, for it was found in fresh water, and thus along with one
other genus, namely. Hydra, affords an exception to the otherwise universally marine habitat of
the Hydroida,

Agassiz,' believing that the hydroid on which I founded the genus Cordylophora is congeneric
with a hydroid discovered long ago by Cavolini in the Bay of Naples, and described by him under
the name of Sertdaria parasitica, would refer them both to the genus Syncoryne of Ehrenberg.

From this determination, however, I must altogether dissent. The name of Syncoryne was
introduced by Ehrenberg in 1833* in order that it might replace that of Stijmla, Sars' name for
a genus of hydroids exactly equivalent with the Coryne of Gartner. To this genus Syncoryne

^ See above, p. 204..

* ' Reports of the Meeting of the British Association,' held in Cork, 1843, and ' Ann. Nat. Hist.,'
xiii, p. 330.

' ' Contr. Nat. Hist. U.S./ p. 239. * ' Corallenthiere des llothes Meeres.'

33

252 CORDYLOPHORA LACUSTRIS.

Ehrenberg erroneously referred Cavoliiii's liydroid, and lie has in this respect been followed by
Agassiz, who, moreover, refers to it the species of the genus Cordi/Iojj/iora, which, no less than the
Sertularia parasitica of Cavolini, are absolutely excluded from it.

But in order that the genus Syncoryne may admit these forms, Agassiz finds it necessary
to modify it by removing from it all the si)ecies included in it by Ehrenberg except the Serfularia
parasitica of Cavolini, which, as has just been said, was erroneously placed there by Ehrenberg.

Now this reconstruction of Syncoryne in a sense in which the genus was not understood by
its founder is inadmissible, but even supposing it allowed, I must still differ from the distin-
guished North American zoologist in his generic association of Cordyhphora with the Serfularia
parasitica.

The Sertularia parasitica already taken by Van Beneden as the type of a new genus,
Corydendrium, Van Ben.,^ is certainly a very remarkable hydroid, and though much is still needed
for a satisfactory knowledge of it, we can gather from the figures and description left by Cavolini
that it has a curious complex hydrophyton, that it has a singularly extensile and dilatable pro-
boscis, and that its gonophores are phanerocodonic. In all these points Cavolini's hydroid stands
widely separated from the species of Cordylophora, so widely that it cannot be associated with
them in a common generic group.

1. COBDYLOPHORA LACUSTRIS, Alhiaii.
Plate III.

TuBtiLAKiA CORNEA ? — AijarHi, Kongl. Vetensk. Ak. Forhandl. Stockholm, 1816.

Cordylophora lacustris, — Allman, Brit. Assoc. Reports, for 1843; Phil. Trans., 1853, p.

367, pi. XXV and xxvi. Hincks, in Ann. Nat. Hist.,
for March, 1853; and Brit. Hydr. Zooph., p. 16, pi. iii,
fig. 2. Van Beneden, in Bui. Ac. Roy. de Belg., 2me
ser., vol. xxiii. No. 5, 1867. F. E. Schulze, Ueber
den Bau und die Enwicklung von Cordylophora
lacustris, mit sechs Kupfertafeln, Leipzig, 1871.

SvNCORVNA lacustris, — Agussi:, Coutr. Nat. Hist. U.S., vol. iv, p. 339.

TEOPHOSOME. — HrDROCAULUS alternately branched, attaining a height of from
two to three inches ; perisaec gradually losing itself on the neck of the hyclranth, and
with shallow annulations at the origins of the ultimate ramuli. Hydranths with
about sixteen tentacles.

GONOSOME. — Gonophores obovate, invested by an extension of the perisarc ; on
short stalks, springing from the sides of the ultimate ramuli on which they are alter-
nately disposed.

Colour. — Hydranth nearly colourless, but with a pale reddish-brown tint ; gonophores with a
reddish-brown spadix ; perisarc brown in the older parts, pale straw colour in the younger.

' 'Bull. Acad. Brux.,' 1844.

CORDYLOPIIORA LACUSTRIS. 253

Development of Gonosome. — June to August.

HahUut. — Attaclied to the under surface of floating timber, and to various submerged bodies
in fresh and slightly braekish water.

Localities. — Grand Canal Docks, Dublin, G. J. A. ; Commercial and West Indian Docks,
London, Dr. Bowerbank and G. ,I.A. ; a freshwater cistern near London, Busk; agricultural
drains near Lynn Regis, Dr. Lowe ; canal near Ostend, Van Beneden ; Schleswig, Semper ;
neighbourhood of Stockholm, Retzius ; near Rostock, F. E. Schulzc.

I first met with Cordi/lopliora lacmtris on tJie l)ottom of an old canal boat which had
been lying for a long time in the docks of the Grand Canal, near Dublin. Tliese docks form
the termination of a canal which receives its supply of water from lakes in the interior of the
country. The docks are many feet above the river Litfey, from which, however, vessels can be
admitted into them by means of a lock ; and though the part of the river with which they can be
thus made to communicate is uiuler the influence of the tide, the quantity of salt water which
may get access to the docks at the time when vessels enter them from the river is altogether
inappreciable.

Cordi/Iophora tacustris occurs also very abundantly and in great perfection on floating timber
in the Commercial and West Lidian Docks on the river Thames, where it is accompanied by
Spongitta Jlumatitis, and by Hydras, freshwater Polijzoa, and other freshwater mollusca, as well
as by Potomogetons, Lemnas and various other freshwater plants. Here also any sea water which
during the entrance of vessels may get admittance to the docks is quite inappreciable to the taste.

But the Cordi/Iophora has been also found in water supplied to the inhabitants of London for
drinking and other domestic purposes, for Busk has identified with it a hydroid met with in a
dark cistern filled with water obtained for these purposes from the river Thames. Quite recently
it has been found by Professor Van Beneden on shells of Dreissena poljjmorplia, and on timber
from a canal near Ostend. Specimens of Cordylopliora tacustris have been kept alive and in
health by myself for many weeks in the ordinary drinking water of the City of Dublin. This
water is brought from freshwater lakes which lie far away in the interior of the country, so that
there can be no question as to the present species living and thriving in absolutely freshwater.

I must here mention, however, that specimens obtained in the West Lidian Docks were,
after having been brought to Edinburgh, placed in some of the water supplied to the inhabitants
for domestic purposes, where they continued to live for some weeks, but with an evident loss of
vigour ; many of the hydranths had fallen off the branches, and fresh buds were but feebly
produced. Li this state of things I added a very small quantity of sea water to the fresh water
in my jars — less than one hundredth part of the whole— when my specimens, after a few days,
had quite recovered their original vigour, and had begun to throw out al)undance of healthy
hydranth buds.

From this it would seem that the specimens from the West Lidian Docks were unable to
dispense, at least suddenly, with the very minute quantity of sea water which may occasionally
gain access to the Docks, a quantity so small as to be entirely inappreciable to the sense of taste,
and which in no way interferes with the healthy growth of truly freshwater animals and plants
which, along with the Corydytophora, are found abundantly in the Docks.

It is highly probable that Cordytopltora lacmtris is an introduced species, possibly imported
into this country, as Breissena potymorpha has been supposed to be, on foreign timber, and like

254 CORDYLOPHORA ALBICOLA.

this mollusc is now making its way into our canals and rivers. It is a light-shunning animal
delighting in such obscure places as the under side of a large log of floating timber, and always
avoiding exposure to the direct rays of the sun. The specimens already mentioned as flourishing
in water supplied for domestic purposes to the inhabitants of London were found in a dark
cistern constructed in the interior of a house, and totally excluded from all access of daylight.

The hydranths of Cordijlophora lacustris are very mutable in shape, aud they may be seen,
according to their state of contraction, varying from a short fusiform or clavate figure to a nearly
cylindrical one. The tentacles when contracted are short and thick, but may be extended to
long thin thread-like appendages, which float listlessly in the water, as they yield to every
impulse of the passing current.

The gonophores afford a good example of that form in which the spadix becomes branched,
enveloping in its ramifications the ova or the spermatozoa. This branched condition of the
spadix I mistook at first for a system of radiating canals,' an error which subsequent observations,
aided by experience in the investigation of other hydroids, have enabled me to correct.

Dr. F. E. Schidze has recently made this species the subject of an excellent memoir, in which
he discusses many important points bearing in the minute structure of the Hydroida. (See
above. Part I, p. 228.)

*^* 2. CoRDTLOPHORA ALBICOLA, Kirchenpauer.

CoRDYLOPHoiiA ALBICOLA, — Kifchenpaiier. la a letter to Mr. Busk in Journ. Mic. Sci., 1861,

p. 283, pi. ix, figs. 12—14.

TROPHOSOME. — Htdrocaultjs attaining about one inch in height ; branches
alternate, annulated ; perisarc terminating abruptly below the hydranth.

GONOSOME not known.

Hahiiaf. — On buoys in tidal rivers near the highest limit of the tide.
Locality. — The River Elbe, Senator Kirchenpauer.

I have selected the charactei's given in the above diagnosis from Senator Kirchenpauer's
description and figures communicated by Mr. Busk to the ' Journal of Microscopic Science.'
Kirchenpauer believes that in the form of the body and tentacles of the hydranth characters may
also be found by which Cordyloplwra albicola shows itself to be a distinct species from
Cordylophora lacustris ; in these, however, I can see nothing beyond indications of a particular
state of contraction. It is in the well-marked annulation of the branches, not merely at their
origin but throughout, and in the abrupt termination of the chitinous perisarcal tube as shown
in Kirchenpauer's figure, that the only valid grounds of specific distinction are to be found.

The species was discovered by Senator Kirchenpauer on those buoys in the estuary of the
Elbe, which lie farthest from the upper limit of the tide, and consequently most removed from

' ' Phil. Traus.,' loc. cit.

TUBICLAVA. 255

the influence of the sea water. On tliose buoys which lay further down the river no Cordi/Iophora
was found, but in its stead abundance of truly marine hydroids.'

The observations made by Kircheni)aucr on the change of life as seen in the forms which
successively show themselves on the buoys in proceeding from the sea at the mouth of the river
upwards into tiie fresh water, arc full of interest and significance.

TUBICLAVA, Allman.

Name. — From Tubus, a tube, and Clava, a genus of hydroids ; in allusion to the form of the
trophosome, which resembles a Clava invested by a tubular perisare.

TROPHOSOME. — IlTDKOPnTTON consisting of a conspicuous htduocaulus in the
form of simple or branched stems, and of a creeping filiform htdborhiza ; the
whole invested by a chitinous peuisarc. Htdranths claviform, with scattered filiform
tentacles.

GONOSOME.— [GoNOPHouES adelocodonic, borne by the hydranth at the proximal
side of the tentacles.] ?'

It was in one of the shallow rock pools left Ijy the retiring tide at Torquay that I obtained
the trophosome of a hydroid which resembled that of a minute Clava in all respects, except in
the fact that the hydranths were supported on conspicuous though simple stems, which were
invested by a distinct perisare.

The specimens were unfortunately destitute of gonophores, but the trophosome brought to
my mind a very similar hydroid which I had obtained several years previously on the shores of
Dublin Bay, and in which the gonosome was present.

The Dublin Bay specimens, like those of Torquay, had a developed hydrocaulus, which was
invested by a perisare. The gonosome was exactly that of Clava inulticonils, consisting of
similar clusters of gonophores grouped round the base of the hydranth-head. I regarded it,
indeed, at the time as a mere variety of Clava multicornis,\\\ which the hydrocaulus had taken on
an unusual development, and I thought little more about it until the occurrence of the Torquay
hydroid brought it to my memory.

While I consider the Torquay hydroid with its well-developed hydrocaulus as generically
distinct from Clava, I am now disposed to regard it as identical with the hydroid of Dublin Bay.
If so, the gonosome is known, and we have all the data necessary for a complete generic
diagnosis.

' Kirchenpauer, " Die Seetonnen der Elbmundung." ' Abhandl. uaturwiss. Vereine in Hamburg,'
iv Band, 1862.

- See the general remarks on the genus given below.

356 TUBICLAVA LUCERNA.

Some time afterwards I obtained at Tenby another hydroid resembling the Torquay one in
all its more important charactere, but with a still more conspicuous hydrocaulus, for this was
occasionally branched, while it attained at least twice the height of that of the hydroid from
Torquay. Like this, however, the Tenby hydroid was unfortunately without a trace of
gonosome.

From a comparison of the three hydroids now mentioned, I consider myself justified for the
present in regarding the Tenby and Torquay hydroids as generically identical, though specifically
distinct, while that from Dublin Bay is probably of the same species as the Torquay one.

It is to the new genus thus represented by two species that I have given the name of
Tubiclava; but since I do not at this distance of time feel absolutely certain as to the identity
of the Dublin Bay and Torquay hydroids, the gonosome of Tubiclava must still be regarded as in
some respects hypothetical, and the genus itself as defined above must be accepted with just
so much of a provisional element.

If the gonosome of Tubiclava be formed as here considered probable on the type of that of
Clava, then Tubiclava is a Clava with a developed hydrocaulus, and is especially interesting as
a transition form between Clava and Cordylophora, its gonosome being that of the former, and its
trophosome that of the latter.

1. Tubiclava luceena, Allman.
Plate IT, figs. 7, 8.
Tubiclava lucerna, — Allman. la Ann. Nat. Hist., for June, 1863, and for May, 1864.

TROPHOSOME. — Hydrocaulus about two lines in height, sinii)le, cyhndrical,
with a slight dilatation at the base of the hydranth ; perisarc with shallow corruga-
tions. Htdranth when extended about equal in height to the stem, and with about
twenty tentacles, which are confined to its distal third.

GONOSOME. — [GoNOPUOREs in dense clusters, immediately below the proximal
tentacles.] ?^

Colour. — Hydranths milk white, with pale ochreous tint on the stomach cavity.
Development of Gonosome. — (?) observed in June.
Habitat. — Creeping over stones in rock pools.
Batliymetrical distribution. — Litoral zone.
Localities. — Dublin Bay and Torquay, G. J. A.

Tubiclava lucerna is a minute hydroid, and might easily be overlooked or without close
examination be mistaken for a creeping form of Clam. The hydranth is very contractile ; when

^ See for tlie evidence as to the nature of the gonosome the general remarks on the gen\is given
above.

TUBICLAVA FRUTICOSA. 257

fully extended it is nearly cylindrical, but may assume when contracted a short thick pyriform
shape. In its extended state it brings to mind somewliat the form of a lighted torch, a resem-
l)lance which has suggested the s])ecific name which I have assigned to it.

2. TUBICLAVA FRUTICOSA, Allmaii.
Piute II, figs. 5, 6.

TROPHOSOME. — Htdrocaulus attaining a lieight of four or five lines, more or
less branched ; perisarc smooth. Hydranths nearly cylindrical in extension ; tenta-
cles, from fourteen to eighteen, occupying about two thirds of the hydranth.

GONOSOME unknown.

Colour. — Hydranth deep vermilion passing into pale red in the cocnosarc, where it is invested
by the translucent pale straw-coloured perisarc.

Habitat. — Attached to rocks at extreme low water spring tides.
Bathymetrical distrihidion. — Lower limits of Laminarian zone.
Locality. — Tenby, G. J. A.

Tubiclava fruticosa forms small vermilion tufts on rocks which are exposed only at very low
spring tides. The hydranths when extended are nearly cylindical, but become clavate in contrac-
tion. They may occasionally be seen bent on one side so as to assume a nutant attitude. The
periderm is light straw colour, and so translucent as to allow the pale red coenosarc to be seen
tiirough it.

The budding hydranth is entirely enveloped in a delicate extension of the perisarc of the stem
which carries it, and within the sort of capsule tluis formed the tentacles of the bud are developed.
When it has attained sufficient maturity the hydranth breaks through its chitinous envelope, and
thus comes into immediate contact with the surrounding water.

MEEONA, Norman.
Name. — An arbitrary name without any special significance.

Tubiclava, — Norman, in Ann. Nat. Hist., 1864.
Merona, — Norman, in Ann. Nat. Hist., 1865.

TROPHOSOME. — HTDEOPHrroN consisting of a developed nYBROCAULUS and a
creeping filiform htdrorhiza ; the whole invested by a chitinous perisarc.
Hydranths claviform.

258 MERONA CORNUCOPIA.

GONOSOME. — ^Sporosacs borne on blastosttles, which spring directly from
the creeping hydrorhiza.

This genus has been instituted by the Rev. A. Merle Xorman, for the reception of a hydroid
which he at first referred to the genus Tubiclava. The gonosome, however, belongs to a type
generically different from tliat which we have reason to believe is the gonosome of Tubiclava, for
while in the latter genus the gonophores would seem to be borne directly by the hydrautb, they
are in Merona carried upon blastostyles which spring from the hydrorhiza.

Merona coENUcoPLffl, Norwian.

Tubiclava coRNUcopiiE, — Norman, in Ann. Nat. Hist., for Jan., 1864, pi. ix, fig. 5. Hincks,

Brit. Hydr. Zooph., p. ii, pi. ii, fig. 2.
Merona cornucopia, — Norman, Ann. Nat. Hist., for April, 1865.

TROPHOSOME. — Htdrocaulus consisting of shghtly curved trumpet-shaped
tubes, which are given off at short intervals from the creeping hydrorhiza, and
gradually increasing in diameter from the base to the summit, attain a height of one
fifth of an inch, or a little more, the tubes marked by a few slightly elevated trans-
verse lines of periodic growth. Htdranths with greatly elongated club-shaped
heads,

GONOSOME. — Gonophores forming mulberry-like masses on the summit of
very short blastostyles, which are invested at their base by a tubular extension of the
hydrorhizal perisarc.

Habitat. — Growing on the shells of living mollusca from deep water.
Bathymetrical distribution. — Deep sea zone.

Localifi/. — About twenty miles north of Unst in Shetland, Rev. A. M. Norman. Other
parts of the Shetland seas, Mr. Peach.

Merona cormicopice was observed by j\Ir. Norman upon the shells of Astarte sulcata and
Dentalium entails which had been dredged from a depth of from SO to 100 fathoms. The very
considerable depth thus frequented by this hydroid constitutes an interesting fact in its economy.
Mr. Norman also informs us "that in every instance it was observed upon shells still occupied
by the living mollusca, and that it had invariably assumed a position at the posterior extremity
of the shell where it would receive the benefits of the aqueous currents caused by the mollusc."

TURRIS NEGLECTA. 259

TURRIDyE.

TROPHOSOME. — Htdrocaulus rudimental or developed. Htdranths with
scattered filiforin tentacles.

GONOSOME. — GoNOPHORES medusiform; planoblasts with simple radiatiii"-
canals, and simple marginal tentacles.

TURRIS, Lesson.

Name. — From Titrris, a tower, in allusion to tlic form of the umbrella.

TROPHOSOME. — HTDUOPnYTOX consisting of a creeping filiform htdroriiiza
and a rudimental htdkocaulus, the whole invested by a perisauc. Htdranths
claviform.

GONOSOME. — GoNOPHORES phanerocodonic. Umbrella of mature medusa sub-
cylindrical ; manubrium massive, with a four-lipped mouth ; radiating canals four ;
marginal tentacles numerous, each with a bulbous base having a distinct ocellus.

The gonosome of this genus has long been known, but it is only within a few years that its
trophosome has been detected. For the discovery of the trophosome of Turris, we are indebtcil
to Gosse, who found that the eggs discharged from the medusa to which Lesson assigned the
name of Turris ncr/lecfa, developed themselves into a CIava-\\kQ trophosome, an observation
which was afterwards confirmed and extended by Strethill Wright.

Turris ^'eglecta, Lesson.

Turris neglecta (planoblast), — Lesson, Prodromus (1837), No. 38 ; Hist. Nat. des Acad.

(1843), p. 28-1. Forbes, British Medusae, p. 23, pi. iii, fig. 2.

(Plauoblast and trophosome), Gosse, Devonshire Coast, p. 348,

pi. xiii, figs. 6 — 10.

Clavula Gossii (trophosome), Strelhill Vi'r'njht. In Ediub. New Pliil. Journ., for July, 1859,

pi. viii, fig. 1.

34

2G0 CAMPANICLAVA.

TROPHOSOME. — Hydeanths about a line in height, scattered on the uydko-
RnizA, and with twelve or more tentacles.

GONOSOME. — Umbrella of mature medusa mitrate ; marginal tentacles sixty or
more.

Colour. — Hydranths and ovarian masses in the walls of the manubriuui bright crimson.
Habitat oi gonosome pelagic, of trophosome unknown.

Locality (of gonosome), European seas.

Observations are still needed for an adequate knowledge of the completely developed
tro})hosorae of this beautiful little iiydroid. The cUiated planula is of a fine crimson, and Gosse
traced its development as far as the production of a creeping stolon, and the formation of a
liydranth which he followed to the emission of its first four tentacles, while Wright, carrying his
observations further, saw the tentacles become multiplied to twelve, and noticed the production
of a chitinous perisarc on the stolon and rudimental stem. Beyond this, however, no observations
have been carried, and we as yet know nothing of the part of the trophosome from which the
gonophores are developed, nor of tlie condition of the medusas at the time of their liberation.

Wright gave to the trophosome of Tunis nec/lecia the name of Clavula Gossii, a name,
however, which, in accordance with the principles which must regulate the nomenclature of the
Hydroida, has to yield to that long previously given to the medusa.

CAMPANICLAVA, Allman.

Name. — From Campaiia, a bell, and Clara, a genus of hydroids ; in allusion to the medusiforin
gonophores and the clavadike trophosome.

Syncoryne, — Gccjenhaur.

TROPHOSOME. — HTDEOPnTTON a creeping filiform hxdeouiiiza, invested by
a PERisAKc ; iiYDRocAULUs undeveloped. Hydranths claviform.

GONOSOME. — GoNOPnoRES phanerocadonic, borne by the hydrorhiza ; umbrella
at the time of liberation deep beU-shaped ; manubrium destitute of oral tentacles ;
radiating canals four ; marginal tentacles tw'O, continuous with two opposite radiating
canals, and having bulbous bases without ocelli, two intervening smaller bulbs corre-
sponding with the terminations of the other two radiating canals in the circular
canal.

The genus Cawpanidava has been constituted for a remarkable hydroid described by Gegen-
baur under the name of Syncoryne Clcodora, but which cannot be received into the genus Syn-

CAMPANICLAVA CLEODOR^E. 261

coryne as understood in the restricted sense reqnired by our present views of the limitation of
hydroid genera. It includes a single species.

*^* Campaniclava cleodoe/E, Gegciibaur, sp.

Syncoryne cleodob*, — Gegenbaur, Generationswechselj p. 11, pi. i, figs. 3, 4.
Campaniclava cleodor.e, — AUman, in 'Ann. Nat. Hist./ for May, 1864.

TROPHOSOME. — IlTDROKniZA branched, carrying the hydranths at distinct
intervals along its course. Hydeanths one half a line in height, with from five to
eight tentacles.

GONOSOME.- — -GoNOPHOEES sessile, scattered on the hydrorhiza. Umbrella of
PLAXOBLAST, with a superficial band of thread-cells lying over each radiating canal,
and extending from the summit of the umbrella to its margin ; manubrium extending
to about half the depth of the umbrella-cavity, its mouth provided with four short lobes
set with thread-cells.

Development of Gonosome. — November to March.

Habitat. — Growing on the shell of living Cleodora friciinjndala.

Bathijmetrical distribiitioii. — Surface zone.

Locality. — Coast of Sicily, Gegenbaur.

The two-tentacled medusa of Campanictava cleodora: is without doubt destined to undergo
changes before attaining maturity. It is almost certain that at least two additional tentacles
become developed, one from each of the intermediate marginal bulbs, as indeed the observations
of Gegenbaur, who obtained in the open sea free medusa; which he refers to this species, but
which were provided with four tentacles, go far to prove.

In its habitat Campaniclava cleodora presents, as Gegenbaur notices, a feature of no little
interest, for it grows on the shell of a pelagic mollusc, so that not only do the medusje which
constitute its gonosome enjoy a free locomotive life, but the trophosome also spends its life upon
the open sea, always wandering over its surface like the active pteropod with wdiich it has associated
itself.

Among forty specimens of Cleodora tricu.spidata observed by Gegenliaur no less than four
fifths of the whole carried a colony of Campaniclava on their shells.

202 CORYDENDRIUM PARASITICUM.

CORYDENDRIUM, Van Beneden.
Name. — From Kopmn\, a clul), and clvcpm; a tree ; in allusion to the form of the trophosome.
Sertularia, — Cavolini, Mem. Polyp.

TROPHOSOME. — Hyduophyton consisting of a well-developed hydeocaulus,
and a creeping filiform hyduoriiiza ; the whole invested by a distinct perisarc.
Hydranths fusiform, with scattered filiform tentacles.

GONOSOME. — Planoblasts borne on the hydrocaulus. Form of medusa
unknown.

The present genus was instituted by Van Beneden for the Sertularia parasitica of Cavolini,
a hydroid which Ehrenberg had incorrectly referred to his genus Syncoryne ; Agassiz, believing
that the species of Cordyloplwra are congeneric with Cavolini's hydroid, refers them both to the
genus Syncoryne, which, however, he iinds it necessary to reconstruct for their admission. I have
already shown that this view is untenable.

*^* CORYDENDRIUM PARASITICUM, CavoUni.

Sertularia parasitica, — Ciwolini, Mem. Polyp., Sprengel's translation, p. 83, tab. vi, figs.

8—13.
Syncoryne parasitica, — Ehrenberg, Corallenthicre, Abhandl. der Akad. Wisseus., Berlin, for

1832, p. 295. Agassiz, Coutr. Nat. Hist. U.S.,

vol. iv, p. 339.
CORYDENDRIUM PARAS1TICU3I, — Vtiii Beiicchn, BuI. Ac. Hoy. de Bruxellcs, 1844, p. 313.

TROPHOSOME. — Hydrocaulus attaining a height of about two inches; much
branched ; fascicled^ in the main stems, single in the smaller ramuli. Hydranths
with a very extensile and dilatable hypostome.

GONOSOME. — GoNOPiiORES on short peduncles, springing singly from a point
close to the junction of the ultimate ramuli with the branches from which they arise.

General colour. — Dark red.

^ T use the term " fascicled '' here and elsewhere to indicate a composition out of many
longitudinally colierent tubes, each invested by its own perisai'c.

CORYDENDRIUM PARASITICUM. 263

Development of Gonosome. — Summer.

Habitat. — Growing on the liydrorhizal stolons of Eudendrium racemosnm (Cavolini).

Batkymetrical distribution. — Litoral zone ?

LocaJitij. — Bay of Naples, Cavolini.

Since the time when Cavolini, towards the end of the last century, described his Sertularin
parasitica, no one has met with this remarkable hydroid ; and though Cavolini's description of it
is, like all his others, full and lucid, we must wait for farther observations made with the advan-
tages aftbrdcd by our presc^nt knowledge of the Hydroida, in order that some points of which
we are as yet ignorant in its structure and habits may be cleared up.

That it produces phancrocodonic gonophores there can be no doubt from Cavolini's descrip-
tion and figures, though the Neapolitan zoologist takes them for budding hydranths. It follows
too from his description that the main stems consist of a fascicle of tubes which breaks up into
single tubes before entering the branches ; and though it would seem that he believed the bundle
of tubes composing the stem to be enveloped in a common pcrisarc, it is more probable that each
tube has its own perisarcal investment as in other fascicled hydroids.

A remarkable feature is found in the great extent to which the hypostome can be elongated
and dilated ; indeed, Cavolini's figure of the hypostome when thus distended reminds one of the
exserted proboscis of certain annelides.

A curious fact on which Cavolini dwells at considerable length is the occurrence at certain
periods of egg-like bodies in the interior of the hydrocaulus where they replace the coenosarc, and
from which they are subsequently discharged through the extremities of the branches now
destitute of hydranths.

Cavolini takes these bodies for the proper eggs of the hydroid, and even describes their
development into young trophosonies ; but as it is rather by inference than by direct observation
that he traces his young hydroids from the supposed eggs, we must here withhold assent from his
conclusions. Indeed, it is far more probable that the egg-like bodies are parasitical organisms
than that they have any direct relation with the hydroid. ^

Cavolini always found the present species growing from the root-like basis of the Eudendrium
racemosum after the hydi-anths of the Eudendrium had perished and disappeared. He accordingly
regards it as a parasite of the Eudendrium at whose expense he imagined it to live, and it was
this supposed habit which suggested to him the specific name by which he has designated it.
It is almost certain, however, that there is here no case of true parasitism, and it is very likely
that the Corydendmm parasitieum enjoys other habitats as well as that Avhich was alone known
to the celebrated Neapolitan observer.

^ We may here, however, compare Cavolini's account of these egg-like bodies with an observation
made by F. E. Schulze on Cordyloptiora tacustris. Schulze informs us that he had occasionally met
witli bodies indistinguishable from young ova imbedded in the ectoderm of the branches of
Cordi/loptwra. (Schulze, ' Ueber Cordylophora Lacustris,' p. 37). A very similar observation had been
made on Hijdractinia ecJdnata by M. de Quatrefages. See above, p. 2.23.

264 CORYNIDiE.

CORYNIDjE.

TROPHOSOME. — Hydrocaulus developed. Htdeaxths with scattered or more
or less spirally disposed capitate tentacles.

GONOSOME. — GoxopiiOEES fixed Sporosacs.

COEYNE, Gartner.
Name — From Kopvvn, a club ; in allusion to the form of the hydrantlis.

TunuLARiAj — Gmelin.
Sy NCOKYN E, — Ehrenberg.
Hermia, — Johnston,

TROPHOSOME. — Hydeopiiyton consisting of a simple or branching hydeo-
CAULUS, rooted by a creeping filiform iiydeorhiza : the whole invested by a chitinous
PEEiSAEC. Hydranths clavifomi, with scattered capitate tentacles.

GONOSOME. — Spoeosacs developed from the body of the hydranth.

The genus Curi/ne was founded liy .Joseph G;a-tner, the author of the famous botanical work
'De Fnictibus et Scminibus Plantarum.' It is described in a letter written by Gartner to Pallas,
an extract from which, with a figure of the species on which Gartner founded his genus, was pub-
lished by Pallas hi 1774, in the tenth fasciculus of his ' Spicilegia Zoologica.'

The limits of Corijnc, however, as a well-defined generic group, were afterwards disturbed
by the introduction into it of certain forms not contemplated by its founder. In this way the
lli/i/ra isqua/jiala of 0. E. JMiillcr — though a separate genus under the name of Clava had been
constituted for it by Gmelin — came to be included in the genus Corijnr, and we find Bosc/
Lamarck," and Cuvicr" adopting this association.

^ ' Hist. Nat. (Ics Vers.,' torn, ii, p. 230.

" ' An. s. Vert.'

^ ' Rca^nc Anininl,' 1817.

CORYNE. 265

In 1S29, however, Sars i drew attention to the lieterogeneous nature of the group Cori/wj
as then accepted by systematists, and proposed to break it up into two genera, retaining tlie
name of Corijnc for the Hydra squamata of Miiller, and referring the true Corynes of Gartner to
a separate genus, which lie named Stipula, apparently unaware tliat the forms which he desired
to include under his new genus were those for which Gartner had long previously founded his
genus C'ori/iic.

'riiough Sars has tluis done good service in discriminating foi'ius falsely associated, there was
no need for his new name, Stipula ; for it is plain that he ought to have retained in the genus
Coryne the forms for which this genus had been founded by Gartner, while he should ha\e
assigned the Hydra squamata to the genus Clava, already constituted for it by Gmelin.

We next find Ehrenberg^ accepting the dismemberment of Coryne, as proposed by Sars,
l)ut instead of objecting to the name of Stipula on the ground that it was needless, he objects
to it on the ground that it was already in use among the botanists ; and then, instead of referring
the Hydra squamata to its legitimate genus, Clava, and restoring to its proper place Giu'tner's
name of Coryne, he introduces the generic name of Syiicoryne, to supplant the Stipula of Sars.
The Coryne of Gartner, Stipula of Sars, and Syncoryne of Ehrenberg, are thus merely synouymes
of the same group.

But though the group under these different names was accurately determined, it began once
more to lose its unity by the introduction into it of forms gcnerically distinct from the original
Coryne, and a new dismemberment was accordingly called for.

This dismemberment I adopted some years ago.^ The grounds on which it must be based
will be found, not in the trophosome, but in the gonosome, the gonophores of which, in some of
the species included by modern systematists under the genus Coryne or Syncoryne, are adelocodonic,
while in others they are phanerocodonic ; and again, among these phanerocodonic gonophores
forms will be found which differ from one another by characters of generic value.

If we can satisfy ourselves as to the form Gartner had in view when he constituted his genus
Coryne, there can be no doubt as to those to which this name must now be restricted. The whole
question, therefore, necessarily turns on the identification of the Coryne pusiUa of Gartner.

It must be admitted that we cannot always determine with certainty the exact species
which the earlier systematists had iiefore them in their descriptions, for it was not until a more
accurate knowledge of the structure and life-history of the Hydroiua pointed to the significance
of certain characters previously overlooked that the true value of these characters could be recog-
nised in systematic zoology. There can be thus no doubt that many very distinct species, and
even genera, of hydroids have been described under the same name ; and though our attempts
at the identification of many of the hydroids referred to in the writings of the last century and
the earlier parts of the present can have little more than approximate results, still I believe that
we shall be justified in regarding as Gartner's " pusilla'' a well detennined and abundant species
of our shores.

It is by no means easy to find among the true Corynes characters available for specific dis-
tinction, and most of the forms which have been described as distinct are far from being so strongly

' ' Bidrag til Soedj'rencs Natuiiiistorie,' p. 4.

- ' Corallcnthiere,' 1833.

^ 'Ann. Nat. Hist.,' for May, 18G1.

26G CORYNE PUSILLA.

marked as to enable them to he unhesitatingly received as so many well- defined species. Indeed,
though I have enumerated under the genus Coryne several species with the diagnoses proposed for
them by their describers, I am well disposed to believe in the absolute validity of only three
of these, the Cor^/iie piisilla, the C'ori/m- nurjinuta, and the Cori/)ic ciespes.

1. CoiiYNE PUSILLA, Gartner.
Plate IV, figs. 1—7.

Coryne pusilla, — Gartner, in Pallas. Spicil. Zool., fasc. x, p. 40, tab. iv, fig. 8. Jolinston,

Brit. Zoopli., 1847, p. 39, pi. ii.
TuBULAiiiA coRYNA, — Gmelui, Lbm., 3834.
Coryne glandulosa, — Lamarck, An. s. Vert., tome ii, p. 62.
Syncoryne PUSILLA, — Elivenberg, Corallenthiere, 294.
Hermia GLANDUi.osA, Tohnston, Brit. Zooph., 1838, p. Ill, woodcut 12.

TROPHOSOME. — HTDROCArLUS attaining a height of about an inch, much and
irregularly branched ; pekisakc distinctly annulated, not sending off a sheath upon the
base of the hydranth. HrDR.iNTii with about thirty tentacles.

GONOSOME. — Spoeosacs globular; developed among tlie tentacles from tlie
body of the hydranth.

Colour. — Hydranths and sporosacs flesh colour, perisarc reddish-brown.
Development of Gonosome. — May to Septenilier.
Habitat. — Growing on fuci and on rocks between tide marks.
Batliymeirical dwtrlhution. — Litoral zone.

LocaUticfi. — Coast of Cornwall,? Gartner; east and west shores of Scotland, and Orkney
Islands, G. J. A. ; Shetland Islands, Rev. A. M. Norman and G. .1. A. ; Dublin Bay, G. J. A.

The determination of the true Coryne piisiUa of Gartner is by no means an easy task. The
figure in the " Spicilegia Zoologica" is rude, and far from being sufficiently exact for undoubted
specific identification, and I know of no species which exactly answers to the description.

It is thus, perhaps, impossible to determine with certainty the species which Giirtncr had
Ijefore him, and which, it would appear, he obtained on the coast of Cornwall, when he furnished
Pallas with the first description and figure on record of a Coryne. A species, however, which is
widely distributed round our coasts, and is in some localities very abundant, would seem to agree
as nearly as any other with Gartner's hydroid. It is this species which Johnston, in the first
edition of his " British Zoophytes," figures and describes under the name of Hermia (/landdosa,
adopting the specific name from Lamarck, who used it for the Coryne pusilla of Gartner. In his
second edition Johnston describes and figures the same hydroid, but now restores to it Gartner's
name of Coryne pusilla.

This determination I shall follow. It is the nearest which the data at our command entitles

CORYNE VERMICULARIS. 267

us to adopt, and I agree to it all the more willingly, as by doing so we retain in tliu nomenclature
of Zoology a name which has no slight historical claims on our acceptance.

Coryne pitsilla is a litoral species occurring between tide marks, where it may be found
attached to \'arious sea-weeds, more especially Fiicus nodosiis and Fucus scrralus. When well
developed it forms dense tufts, about three quarters of an inch or even an inch in height, and
with the stems attaining a thickness of about one quarter of a line, the whole conspicuous by
the fine deep flesh colour of the hydranths. Shorter and more slender varieties however, which
scarcely attain more than half these dimensions, are not uncommon. It is not unfrequently
associated with Clava squamata, though its zone of greatest perfection is a little below that of
the Clava. It may often be seen after the retreat of the tide, exposed for a considerable time
to the air and sun while waiting for the return of the sea, and protected from desiccation only by
the moisture of the surrounding sea-weeds and the water it still contains entangled among its
crowded branches.

It is a fine typical Cortjne, and while living foriiis imder a low power of the microscope
a most beautiful object for the zoophyte trough.

2. CORTNE A^ERMICULARIS, HillcJcS.

Coryne vekmicdlaris, — Hincks, in Ann. Nat. Hist, for Oct., 1806. Brit. Hydr.
Zoopb., p. 1, 2, pi. viii, fig. 2.

TROPHOSOME.' — HTnEOCAULUs branclied dicliotomously, and forming dense
slu'ubby tufts, whicli attain a height of about three quarters of an inch ; perisauc
annulated, especially towards the base; often smooth or slightly wrinkled on the
branches and upper portions of the stem. Hydranths " of great length (about one
sixth of an inch when mature), stout, almost cylindrical for half their length wlien
extended, and then tapering off very gradually towards the oral extremity ; tentacles
in irregular and very distant whorls, rather stout, with large capitula, about twenty-
five in number."

GONOSOME. — Sporosacs " borne at the base of the tentacles over the lower part
of the l3ody ; spherical ; shortly stalked."

Hatlnjmetrical dislriiiition.—Deei^-sea. zone.
Locality. — Shetland Isles; Rev. T. Hincks.

I have never met with any species which exactly agrees with that just described. Mr.
Hincks, from whose account of it I have given the above diagnosis, informs us that it is " dis-
tinguished by the great size and worm-like appearance of the polypites, and the sparing distribu-
tion of the tentacles over the body." It is, however, by no means easy to distinguish it from
Coryne pusilla.

35

16S CORYNE VAGINATA.

3. COUTNE VAGINATA, H/'uch.
Plate IV, figs. 8, 9.

CoRVNE PDsiLLA, var. MuscoiDEs, — JoluisloH, Biit. Zoopli., 1847, p. 12, p]. vi, figs. 4, 5.
CoKYNE VAGINATA, — Hiticks, in Anu. Nat. Hist, foi- 1866. Brit. Hydr. Zooph., p. 41, pi.
viii, fig. 1.

TROPHOSOME. — Htdrocaulus rising to a heiglit of from two to tliree inches,
alternately branched, each of the main branches sending off numerous short alternate
simple ramuli ; peeisaec distinctly and regularly annulated. Hydeanths borne upon
a narrow neck, which is surrounded by a loose membranous sheath, derived from the
perisarc, and losing itself on the body of the hydranth ; tentacles about twenty.

GONOSOME. — Spohosacs globular, developed from the body of the hydranth,
each si)ringing by a short peduncle from the axil of a tentacle.

Colour. — liydranths and sporosacs pale pink, perisarc light brown.

Development of Gonosome. — Summer and autumn.

Habitat. — Attached to sea-weeds and to the sides of rock pools, near low-water mark.

Batliij metrical Distribittion. — Lower part of Litoral zone.

Localities. — Coasts of Devonshire, Rev. T. Hincks and G. J. A. ; Coast of Cornwall and
Southern and Eastern Coasts of Ireland, G. J. A. ; North-east Coast of Ireland, Mr. William
Thomson.

Coryne va(jinata forms branching hydrosomes, longer, looser, and more flexile than those of
Coryne imsilla, while its large size and the symmetrical distribution of its branches renders it a
conspicuous and beautiful object in the clear rock pools where it is rooted.

One of its most striking characters is the sheath-like extension of its perisarc, by w'hich the
neck of every hydranth is invested in a loose transparent pellicle, a feature to which attention was
first called by Mr. Hincks. This is plainly a continuation of the most external or oldest layers of
the perisarc, which in the young state of the hydranth constitute an external capsule, in which
the entire hydranth-bud is included, and within which this bud passes through the earlier stages
of its development.

It is a very abundant and characteristic species on the southern shores of Devonshire and
Cornwall, and on the southern shores of Ireland. It occurs also on the eastern shores of Ireland,
where I have found it in considerable abundance in Dublin Bay, while it has been obtained by
Mr. William Thomson on the same shores, as far north as Strangford Lough. I have never met
with it on any part of the Scottish coast or on the coasts of the adjacent islands, the genus being
there chiefly represented by Coryne pusilla.

CORYNE RAMOSA. 269

4. CORTNE FRUTICOSA, HiuclcS.

CoRYNE FRUTicosA, — Ifiiic/cs, iu Auu. Nat. Hist., scr. 3, vol. viii, p. 158, [pi. vi. Brit.
Ilydr. Zooph., p. 4-i, pi. vii, fig. 2.

TROPHOSOME. — •liYDUocADLTJs busliy, slender, slightly and irregularly annulatcd
throughout, much and irregularly branched, with the perisauc delicate, branches erect,
long, closely set. Hydranths swollen below, tapering above ; tentacles about
twenty, rather long, witli small capitula, a verticil of about five immediately below
the mouth, the rest scattered.

GONOSOME. — GoNOPHORES densely clustered, chiefly about the lower part of
the body ; sessile ; very large when mature.

IlahHat. —On Fuci.

Batlnjmetrical Bistrihution. — Literal zone.

Locality. — Exmoutli, Rev. T. Ilincks.

I have never seen this species, and have contented myself with simply giving Mr. Hincks'
description of it. He fm-ther tells us that it " forms dense clustered bushy masses on Fucus."
It is plainly very nearly allied to the Cori/ne pusilla — so nearly that, though I have here accepted
Mr. Hincks' determination of it, I feel much tempted to regard it as a mere variation of that
species, more slender and less distinctly annulatcd than the typical form.

*^* 5. CoRYNE RAMOSA, Sars.

Stipula RAMOSA, — Suvs, Biclrag til Soedj'reiies Naturhistoric, p. 4, tab. i, fig. 1.
Syxcoryne kamosa, — Ehrenber(j, Corallenthiere, Abliandl. Akad. Wisseus., Berlin, 1832,
p. 295.

TROPHOSOME. — Hydrgcatjlus attaining a height of fiom one to two inches,
slender, much ramified, with the branches subalternate, and very divergent from the
stem at their origin ; perisarc annulatcd. Hydraxths elongate, subcylindrical, with
from twenty to twenty-five tentacles.

GONOSOME. — GoNOPHOREs oval, shortly pedunculate, springing from among the
tentacles towards the proximal end of the hydrantli.

270 CORYNE C^SPES.

Colour. — Hydrocaulus light brownish yellow ; hydraiith Ijrownish red ; gonophores pale red.

Development of Gonosome. — July.

Habitat. — On sea-weeds.

Bathymetncal Distribution. — Literal zone.

Locality. — Coast of Norway (Sars).

It was upon the present species that Sars, in the year 1S29, founded his genus Stijinla, with
the intention of removing to it the true Corynes which had, at the time when Sars proposed his
reform of the genus Coryne, been mixed up with Clava. We have already giveu the reasons
which render necessary the rejection of the nomenclature thus introduced by the celebrated
Norwegian zoologist.^

I have never met with the Coryne ramosa. It is described by Sars as occurring in great
abundance in the neighbourhood of Bexen, where it grows close to the surface of the water, on
a sea-weed which he names Facus nutans. Its taller growth and more elongated hydranths would
seem to distinguish it from Coryne 2}usillu, though I cannot regard its distinctness from this
species so fully established as to remove all doubts as to its validity as a separate species, while it
is distinguished from Coryne vayinata, not only by the absence of a sheath-like extension of its
perisarc, but by its more irregular ramification.

*-...* 6. Coryne c.espes, Allinan.

TROPHOSOME. — Hyduocaulus scarcely exceeding a quarter of an inch in height,
unbranched, or occasionally with one or two simple branches ; irregularly annulated,
the stems all closely crowded upon a hydrorhiza, which is formed by a creeping
entangled mass of tortuous tubes. Hydrantii elongated with about twenty-five
tentacles.

GONOSOME.— Sporosacs globular, scattered on the lower portion of the hydranth,
where they spring by short peduncles from the axils of the tentacles.

Colour. — Hydranths and spadix pale flesh-colour, perisarc reddish-brown.
Development of Gonosome. — March.

Habitat. — Spreading over the surface of submerged rocks.

Bathjmetrieal distribution.— V^^gion of the Cystoscira. About two feet below the surface of
the sea at low water.

Locality. — Gulf of Spezia.

Coryne caspes forms a dense moss-like growth, which spreads like a close turf over the
surface of submerged rocks, for many square miles in extent. It is a well-marked species ; its
short and, for the most part, simple stems, and its dense, crowded, moss-like growth, easily dis-
tinguish it from all the described species.

' See above, p. 2G5.

PROVISIONAL SPECIES. 271

It occurs ill considerable abuiulauce on the west side of the Gulf of Spczia, on rocks which
are never uncovered by the sea.

PROVISIONAL AND INDETERMINABLE SPECIES.

Certain hydroids have been from time to time described whose probable, or at least possible,
place is in the genus Coryne ; but which from the fact of our as yet knowing only their tropho-
somes, cannot be assigned to this genus otherwise than provisionally ; while others, though
referred by their describers to Cori/ne, have been so very imperfectly described as to render their
determination impossible.

Among those which arc in all probability correctly referred to this genus is the following : —

Coryne nutans, Allman (provisionally).

TB.OPHOSOME. — Hydrocaulus attaining a height of about four lines, mucli brauched ; branches
subalternately disposed, deeply and distinctly anuulated, the annulation of the hydrocaulus becoming less
distinctly marked towards the base. Hydranths depressed on one side of the stalk, so as to assume a
nutant posture; ovate, with about fifteen tentacles.

GONOSOME unknown.

The little coryniforiu hydroid here described was obtained in Burraforth Caves, Shetland, by
the Rev. A. M. Norman, who placed in my hands for e.xamination some specimens which had been
preserved in spirits.

These specimens were entirely destitute of gonosome, and its determination cannot, there-
fore, be regarded as otherwise than provisional. The trophosome, however, was well preserved,
and one of its most characteristic features consisted in the drooping attitude assumed by the
hydranths, a feature which has suggested the name by which I have designated the species.

Among provisional species must also be included the Coryne sessilis of Gosse.^ This is a
minute coryniform hydroid with undeveloped hydrocaulus, and with nearly colourless, very elon-
gated hydranths, having the tentacles more than forty in number, and arranged in about six
verticels. As the trophosome is the only part of the hydroid known, the determination of the
genus must be held ov.er until the discovery of the gonosome shall afford the necessary data. It
will then be most probably seen that the gonophores are phanecodonic, and that the liydroid
must consequently be removed from Coryne. Mr. Gosse himself, indeed, assigns to his determi-
nation of it only a' provisional value. He obtained it attached to a rock on the coast of
Ilfracombe.

' Gossc, ' Devonshire Coast,' p. 208j pi. xiv, figs. 2, 3.

272 ACTINOGONIUM.

The Coryne raniosa of Charaisso and Eysenhardt,' recorded by Ehrenberg," under the
name of Si/ncori/ne Chamissotiis, is also indeterminable from want of a detailed deseription of
the gonosome. It is a branching species, attaining a height of about half an inch, and having
its gonophores clustered on the body of the hydranth just Ijelow the tentacles ; but of the
nature of those gonophores we have, as may be expected, no account. It was obtained on the
coast of La Manche.

Among indeterminable species must also be mentioned three which were discovered by Bosc
on floating fucus in the open sea, and figured and described by him under the names of Cori/ne
proUfica, Coryne amphora, and Coryne filif era} Of these his Coryne fdifcra is probably a Clava,
but so very imperfect are the figures and descriptions of these hydroids that no zoologist of the
present day would venture to assign to them a definite place in his system.

ACTINOGONIUM, Alhnan.

Name. — From a/cnc, a ray, and -ydvoc, offspring.

Syncoryne, Van Beneden.
Coryne, Hincks.

TROPHOSOME. — Htdrophyton consistiDg of a developed iixduocaulus springing
from a creeping filiform htdrorhiza, the whole invested by a conspicuous perisarc.
Htdranths claviform, with scattered capitate tentacles.

GONOSOME. — Sporosacs developed from the body of the hydranth, and giving
origin to actinuljd.

The genus Actinorjoiiiujii is founded on a hydroid, described by Van Beneden under the
name of Syncoryne ptmilla, i)ut which, judging from the account given by the Belgian naturalist,
difi"crs altogether from the genus Syncoryne by the earlier stages of its development. Its mature
trophosome is that of a Syncoryne or of a Coryne, but its development is from an actinula instead
of a planula, a character Avhich I I'cgard as of sufficient importance to entitle it to a distinct
generic rank.

^ ' Nov. Act.,' \, p. 370, tab. xxiii, fig. 3 a, b.

' ' Corallenthiere,' p. 295.

^ Bosc, ' Hist. Nat. des Vers.,' tome ii, pp. 339-tO, pi. x.\ii, figs. C — 8.

ACTINOGONIUM PUSILLUM. 273

*^* AcTi\OGONiUM PUSILLUM, Vail Bcneilen.

Syncoryne rusiLLA, — Van Beneden, Mem. sur rEmbryog. des Tubulaiics, p. 53, pi. iii,

figs. 1—12.
CoRYNE Van Benede.vii, — Uincks, Brit. Ilydr. Zoopli., p. 15, pi. ix, fig. 1.

TROPHOSOME.- — Hydrocaulus attaining a height of about one eighth of an
inch, capillary irregularly ramified; pekisarc thin, transparent, slightly annulated.
Htdbauths with about twelve tentacles.

GONOSOME. — GoNOPHOBEs few in number, globular, on short peduncles borne by
the hydrauth at the base of the inferior tentacles. Actinulj^ mth four tentacles.

Colour. — Perisarc pale-yellowish.

Habitat. — Attached to the carapace of crabs and to various other bodies.

Locality. — Coast of Belgium, M. Van Beneden.'

The Syncorijne pusilla of Van Beneden is certainly not the Coryne pusilla of Gartner, though
the latter is cited by the Belgian zoologist under the synonymes. There is, moreover, some con-
fusion between the Syncoryne 2msilla of the ' Memoir on the Embryogeny of the Tubidarians' and
a liydroid to which Van Beneden refers this species in his later memoir (' Faune lit. dc Belgique').
The species of the later memoir bears medusiform planoblasts, and woidd seem to be a true
Syncoryne, though no description of it is given sulficient for diagnosis. M. Van Beneden
declares that it would be difficult to say whether it differs really from the Syncoryne decipiens of
Dujardin.

From the description and figures given by Van Beneden of the Actinoyonium jjiisillum it is
impossible to regard the body which has liberated itself from the sporosac as other than an
actinula. Van Beneden describes it as resembling a minute cuttle-fish with four arms (" poulpe a
quatre bras "), and he further compares it to the free planoblast described by M. de Quatrefages
under the name of Elentlieria.

Through the kindness of IM. Van Beneden I have received specimens of his Syncoryne
pusilla preserved in spirits, but with the gonosome, unfortunately, in so imperfect a state as to
render it impossible to demonstrate in it the important points described by the distinguished
Belgian zoologist. Mr. Hincks, however, has been more fortvraate, for in specimens supplied
some time ago by jM. Van Beneden to j\lr. Alder, he has succeeded in confir-tning the pre-
sence of actinulffi. He has also detected in the walls of the sporosac four canals extending
from the base to the summit.

The Actinoyonium pusillKm is thus one of the very few hydroids in which the development

' Johnston (' Biit. Zoopli.,' pi. iv, figs. 1, 2) figures among British species a eoryuiform liydioid,
which he refers, though not without doubt, to the Syncoryne pusilla of Van Beneden. We have not
yet, however, evidence which can be accepted of the discovery of \a.\\ Beneden's hydroid on the
British shores.

274 SYNCORYNIDiE.

has been shown to take place through actinula^, tlie onl}' other genera which have as yet
afforded evidence of this mode of development being Tahiilaria, Myriothela, and Hydra}

SYNCORYNIDv^.

TROPHOSOME. — Hydrocaulus developed or not developed, hyduanths with
scattered, or partly scattered and partly verticillate capitate tentacles.

GONOSOME.' — Gonopliores in the form of raeduslform planoblasts with four
radiating- canals and simple (rarely undeveloped) marginal tentacles.

SYNCOEYNE, Ehrenberg (in part).

Name. — From aw, indicating union, and Koovvn, a club, in allusion to the numerous club-shaped
hydranths united into a common colony.

TROPHOSOME. — Hydrophyton consisting of a simple or branching uyuro-
CAULUs, with a creeping filiform hydkokhiza; the whole invested by a chitinous
PERiSARC. Hydranths claviform, with scattered capitate tentacles.

GONOSOME. — Planoblasts, developed upon the body of the hydranth. Medusa,
at the time of liberation, with a deep bell-shaped umbrella ; manubrium moderately
large, destitute of oral tentacles ; marginal tentacles four, with bulbous bases usually
furnished with an ocellus ; or the medusse may never become free, the marginal
tentacles remaining at the same time in an imperfectly developed state.

When Sars saw the necessity of reducing to some kind of zoological consistency the
confused state into which the genus Coryne had fallen, he founded, as has been already said,*^ his
genus Stipula for the reception of all the known claviform hydroids with scattered capitate
tentacles. We have seen, too, that Ehrenberg's name of Syncoryne afterwards took the place of
Stijjula, and since then it has been by many authors adopted in its stead.

As the species thus included under Syncoryne increased in number by fresh discoveries, it
became evident that very different plans of gonosome were to be found among them ; we now
know that some carry only sporosacs, while others give origin to medusee, and still further these

' Sec Pmt I, p. 94. ■ Vide supra, p. 265.

SYNCORYNE SARSII. 275

medusa-bearing forms present such differences in their gonosomes that it is impossible to
comprise them all under a single genus.

I believe that our descriptive zoology of the IIydroida will be best served by retaining the
name of S/jncoryne, though in a more restricted and definite sense than that in which it has been
generally employed. We shall thus avoid a further complication of synonymy, and preserve
a sufficiently expressive and convenient name.

I shall thus employ Ehrcnberg's uame of Si/nconjne for those hydroids which with a
coryniform trophosome possess a gonosome which is referable to the form described above under
the character of the genus.'

Among some species of the genus Syncorijne we find the very exceptional condition of the
planoblast never detaching itself from the trophosome, although its well-developed umbrella com-
pletely fits it for free locomotion. Under these circumstances the manubrium becomes loaded with
generative elements, sometimes to such an extent as completely to fill the cavity of the umbrella,
Avhile the marginal tentacles remain imperfectly developed. This condition of the planoblast is
precisely that which is presented by the gonopliores of certain Siphonopuora.

*^* 1. Syncoutne Sarsii, Lovcn.

Syncoryne Sarsii, — Loven, in Transactions of the Kojal Swedish Academy of Sciences, 1833,
translated in Wiegmann's ArcLiv, 1837, p. 323, tab. vi, fig.
25. Sars, Fauna lit. Norv., p. 3, tab. i, figs. 1 — 6.

TROPHOSOME. — Hydrocatjltjs rising to the height of half an inch, slender,
sparingly branched, nearly smooth. Hydkanths with from twelve to sixteen
tentacles.

GONOSOME. — GoNOPHORES developed among the tentacles on the body of the
hydranth. Manubrium with the margin of the mouth slightly crenulated ; marginal
tentacles set with round wart-like clusters of thread-cells, and having a brownish-red
ocellus in their bulbous bases.

Colour. — Hydranths pale rose colour, manubrium of medusa clear red.

Development of Gonosome. — May and June.

Habitat. — On Laminaria and other sea-plants.

Batki/metrical distribution. — Laminarinn zone.

Localities. — Island of lloroc. Coast of Norway (Sars) ; The Cattegat (Loven).

I have never met with this species myself, and have drawn up the above diagnosis from Sars's
description of a Norwegian hydroid which he refers to the Syncoryne Sarsii, figured and so

' I have already defined the genus Sijncoryne in the sense here explained, " On the Constructiou
and Limitation of Genera among the Hydroida," 'Ann. Nat. Hist.' for May, 18G1.

3G

276 SYNCORYNE LOVENI.

iiiimed l)y Loven, but not described by liiui with sufficient detail for satisfactory diagnosis. In
this determination I ha\c followed Sars, though it must be admitted that, if Lovcn's observation as
to the occurrence in his species of certain gonophores which occupy the position of a hydranth on
tlie extremity of a branch be free from error, it is not easy to accept unhesitatingly the specific
identity of the two hydroids. It is possible, however, to explain the peculiar position of the
medusa in Lovcn's specimens by referring it to the atrophy of a hydrantli which may at one
time have terminated the branch, and by its subsequent disappearance have allowed the medusa,
properly lateral, to assume an apparently terminal position.

The Si/ucori/ne Sardl is one of the hydroids which form the subject of Loven's famous
?»lemoir, in which he shows the production of ova, by medusiform buds, and is thus one of the
first in which the production of medusa buds was noticed and described by a competent oliserver.

*:j^* 2. Syncokyne Lovexi, Sars.

SyNCORYNE iiACE.MosAj — Lovcu, in Trans. Hoy. Acad, of Sciences of Swedeu, 1835, trans-
lated in Wiegmau's Archiv, 1837, Band i, p. 321, tab. vi,
figs. 19, 20.

Syncoryne Loveni, — Sars, Fauna Ht. Norv., p. 2, note.

TROPHOSOME. — Hyduocaulus much branched, formhig intricate, flexuous, and
shrubby growtlis, which rise to a height of an inch and a half. Hydranths with
sixteen tentacles.

GONOSOME. — Planoblasts permanently attached, borne on short peduncles, whicli
spring from the body of the hydranth, just behind the posterior tentacles. Umbrella
elongate, bell-shaped, set with scattered thread-cells ; marginal tentacles rudimental ;
manubrium, with its investing mass of ova, nearly filling the cavity of the umbrella,
and witli its mouth minute, and surrounded by a circle of rudimental tentacles.

llahitaf. — On a rocky bottom among shells and sea-weeds.
Bathi/iiietrical distribution. — Laminarian zone ?
Locality. — Coast of Norway, Loven and Sars.

I have never met with this apparently well-marked species, and the diagnosis here given is
framed from Loven's memoir.

Spicoryne Loveni possesses much historical interest as having, along with Sijncoryne Sarsii and
Gonotliyraa Loveni {Campamilaria yenicidata of Loven), formed the subject of Loven's memoir, m
which Ehrenberg's views of the zooidal and sexual nature of the egg-bearing bodies of the
Hydkoida were for the first time confirmed.

Loven names the species Syncoi-ync rainosa, believing it to be identical with the Stipida

SYNCORYNE GRAVATA. 277

ramosa of Sars.i Sars, however, himself in a subsequent work- contends against this identifi-
cation, and proposes for Loven's species the name of Syncorync Lovmi.

Syncoryne Loveni is one of the few hydroids in which the medusa, though provided A\ith a
well-developed umbrella, remains permanently attached to the trophosome, developing and dis-
charging its generative elements at a comparatively early period of its existence, and then wither-
ing away without ever !)ecoming free.

3. Syncoktne GRAVATA, StrctMU Wright.

CoRYNE GRAVATA, — StretMll Wright, in Trans. Roy. Physical Soc. Edin., 1857, p. 338, pi. xI.k,

fig- 5.

TROPHOSOME. — Hydeocaulus simple (?) rising to the lieight of about one quarter
of an inch from a creeping- stolon ; perisarc smooth. Hydranths with from ten to
twelve short tentacles.

GONOSOME. — MEDUsji permanently attached, long cylindrical, springing from
among the tentacles on the body of the hydranth, umbrella with the marginal tentacles
undeveloped, each being represented by' a minute bulb ; manubrium piriform, and
Avith its mass of generative elements nearly filling the cavity of the umbrella.

Colour. — Hydranths and umbrella colourless, marginal bulbs of medusa 1)rown, manubriiuu
with its burden of spermatozoa milk white.
Development of Gono-iome. — Spring.
Habitat. — On the sides of a rock-pool.
BatJiymetrical distributio7i. — Literal zone ?
Locality.— ISorih Berwick, Firth of Forth (Dr. T. S. Wright).

I have never met with this species. The above diagnosis, which has been drawn up from Dr.
hjtrethill Wright's description, must be regarded as applying strictly only to the male, female
specimens not having yet been observed. The growth of the large gonophore, with its voluminous
mass of generative elements, occasionally causes the atrophy and almost entire disappearance of
the hydranth from which it springs.

Syncorync yravata is nearly allied to Syncoryne Loveni. It resembles it especially in its
gonophores, which, though phanerocodonic, develop and discharge their generative dements with-
out ever becoming detached from the trophosome.

' Sars, ' Eidrag til Sciedyrenes Naturliistorie.'
" ' Fauna lit. Norv.'

278 SYNCORYNE MIRABILIS.

*^* 4. Syncoryne MiEABiLis, Agassiz.

CoRYNE MIRABILIS, — Ayussiz, Contr. Nat. Hist. U.S., vol. iv, p. 185, pis. xsvii — xx.
Clark, Mind in Nature, p. 77, figs. 40 — -12.

TROPHOSOME. — Hydkocattlus consisting of branching stems, wliicb, without
any distinctly differentiated iiyduohuiza, adhere to some fixed object for a greater or
less extent of their length, and then become free, forming irregularly ramified tufts,
which usually attain a height of about half an inch ; peeisauc smooth. Hydhanths
with about sixteen tentacles.

GONOSOME. — Medusa borne on the body of the hydranth, either from among
the tentacles or from a point just behind the most posterior ones. They are of two
forms, free and fixed; the free gonophores (produced in the earlier months of the year)
have a wide umbrella, with well-developed marginal tentacles, set with irregular
clusters of thread-cells, and provided with an ocellus in the basal bulb ; manubrium of
moderate size. The fixed gonophores (produced at a later period) have a deep thimble-
shaped umbrella, imperfectly developed marginal tentacles without ocelli at the base,
and a very large manubrium, which, Avith the generative elements developed in its
walls, nearly fills the cavity of the umbrella, and even projects beyond it.

Development of Gonosome. — January to May.

Habitat. — Attached to rocks, sea-weeds, and floating timlier iu the sea and in the brackish
water of the estuaries of rivers.

Localili/. — Atlantic shores of North America, Agassiz and Ckrk.

I have compiled the above diagnosis from characters selected from tlie elaborate description
of the species by Agassiz, whom, along with Professor Clark, I have followed in referring the two
forms of gonophores described iu the diagnosis to one and the same species. I cannot, however,
but think that for so remarkable a fact, unsupported, as it is, by the analogy of other hydroids,
stronger evidence is needed than any that has been adduced by the American zoologists. Until
specimens shall be found in which the two kinds of gonophores exist simultaneously, or until
these can be shown to be consecutively de\'eloped from one and the same colony confined in our
aquaria, or otherwise subjected to continuous observation, we can scarcely accept the specific
identity of the two forms as proved.^

' Mr. Ilincks, liowever ('Brit. Hydr. Zoopli.,' p. 54), not only adopts the views of Agassiz and
Clark upon tliis point, but refers the Sijncorijne mirahilis, A gas., to the Sijncorync gravatu, "Wright,
regardins; the latter as the arrested condition of the former.

SYNCORYNE PULCHELLA. 279

Agassiz, however, is not only convinced of this identity, but lie maintains that the two
hydroids described by Loven imdcr the names of Syncoryne ramosu and ^yncoryne Sarvii, the
one with fixed and the other with free phanerocodonic gonophores, arc on]y different states of a
single species,' a view which, in my opinion, is quite untenable.

Indeed, I am well inclined to keep all the forms with coryne-likc trophosome and fixed
phanerocodonic gonophores in a distinct group, generically separated from the true Syncorynes.'
Tlie chief argument against thus treating them is derived from the observations of Agassiz and
Clark as to the production of fixed and free medusae by the Syncoryne mirabilis. These observa-
tions, if verified, would be fatal to the proposed separation, but, as already said, I am by no
means convinced of their being free from error.

The free medusa of Syncoryne mirabilis possesses at the time of its liberation a manubrium,
which is but little extensile, and scarcely exceeds in length half the vertical diameter of the
umbrella. Agassiz and Clark, however, believe that they have traced this medusa in the open
sea through every stage of growth up to the condition of a true Sarsia, with its elongated and
amazingly extensile manubrium loaded with generative elements, the very species which Agassiz
had already described and figured under the name of Sarsia mirahUis?

There can be no doul)t that the changes which a phanerocodonic gonophore may undergo
in the period between its detachment from the trophosome and its acquisition of a sexually
mature state are frequently very great, and there is no reason why the medusa of Syncoryne
should be exempt from them. The evidence, however, adduced by the American zoologists as to
the ultimate form of this medusa cannot be regarded as conclusive. There is necessarily a want
of direct and continuous observation, and the forms figured by Agassiz as progressive stages of
development do not, in my mind, tend to remove the doubt. The proof has certainly not been
yet given that the Sarsia form is that which the free medusa of Syncoryne must ultimately attain.

5. SrNCOErNE ptjlchella, Alhnan.

PI. YI, figs. 1—3.

Syncoryne pulciiella, — Allman, in Ann. Nat. Hist, for Juue, 1865.

TROPHOSOME. — Htdrocaulus consisting of sim])le stems rising at sliort intervals
from a creeping reticulated stolon, and attaining a height of ahout half an inch ;
PEUiSARC thin, and destitt^te of annulation, but showing some indistinct transverse
rugEe. Htdeanths with from fifteen to twenty tentacles.

' Agassiz, ' Contr. Nat. Hist. U.S.,' vol. iv, p. 190, note.

" To the genus thus constituted the name of Statocodium (crraToe, k(ocwv) might be couTcnieutly
given.

■' Agassi?;, " Coutributions to the Natural History of the Accaleplue of North America," ' ^lem.
Amer. Acad. Nat. Sc.,' 1850, y. 22J, pis. iv, v.

280 SYNCORYNE DECIPIENS.

GONOSOME. — Planoislasts borne on short peduncles in a dense cluster imme-
diately below the tentacles. Umbrella with its transverse and vertical diameters nearly
equal, its outer surface set with scattered thread-cells, and having two well-marked meri-
dional furrows extending along the middle lines of two opposite interradial spaces from
tlic summit of the umbrella to its margin, while two nearly obsolete furrows may occa-
sionally be seen ocupying a similar position in the remaining two interradial spaces ;
marginal tentacles very extensile, nodulated with clusters of thread-cells, which give
them a moniliform character when extended, one larger cluster terminating the
tentacle ; tentacular bulbs with a distinct ocellus.

Colour. — Body of hydrantli deep orange, becoming paler wlieii it passes into the pale orange
stem; maiuibriuni and tentaculai- bulbs of medusa deep orange.
Development of Gonosome. — April.

Habitat. — Rooted to the bottom of rock-pools near low-water mark.
Bat Ji //metrical dislrihution. — Lower part of Literal zone.
Locality. — Skelmorlie and Rotlisie, Firth of Clyde (G. .J. A. ).

Si/ncoryne pulchella, though of humble habit, is yet conspicuous by the l)right orange colour
of the hydranths and medusa-buds. Paler varieties, however, may occur. It is nearly allied to
the Syncoryne decijxiens of Dujardin, with which it agrees in the form of the planoblast, and in
the fact that these are all borne at the proximal side of the tentacula. From Syncoryne decipicns,
however, it differs in its simple habit, and in the more ovate form and more numerous tentacles
of the hydranth.

With Syncoryne eximia also the present species closely agrees in the form of the planoblast,
though the position of these on the trophosomes of the two species are very different. The
planoblasts, indeed, are scarcely distinguishalde from one another, except in the fact that the
longitudinal furrows of the umbrella do not exist in Syncoryne eximia.

*;}^* C. SY^xouYNE DECIPIENS, Bujardin.

Syncoryne decipiens (trophosome), sthenyo (planoblast), — Dujardin, iu Ann. Scr. Nat.j 3ieme

ser., 1845, tome iv, p.
275, pi. xiv, fig. B.

TROPHOSOME. — Hydkocaulus branched, attached by a creeping stolon, and
invested by a smooth perisakc. Hydranths nearly cylindrical, with eight or nine
tentacles.

GONOSOME. — Planoblasts produced immediately below the tentacles of the
hydranth ; transverse and vertical diameters of umbrella nearly equal, a longitudinal
furroAv in the middle of each interval of the radiating canals ; marginal tentacles
moniliform, with clusters of thread-cells, and having an ocellus in the bulbous base.

SYNCORYNE FRUTESCENS. 281

Locality/. — Loriciit, Dujardiii.

The Si/ncoryiw decipiens possess an liistoric interest in l)eing a species which had afibrdcd
material tor some of tlie earliest observations on the production of free medus;e hy the fixed
trophosonie, for it uas studied by Dujardin, who lias described it in his well-known memoir
as presenting an cxami)le of this phenomenon.

It is not easy, however, to collect from Dujardin's description characters suilicient for a
satisfactory diagnosis of the species ; those given above may be gathered from his memoir, and
are probably sufficient to secure this species from being confoiuidcd with any other.

Dujardin describes its medusa, which he names "Sthenyo" as undergoing some time after
its escape from the tropliosome, and while still in the confiuement of his vases, certain remarkable
changes, which consist chiefly in a complete retroversion of the umbrella. The changes noticed
by Dujardin would seem to be very similar to those which I have myself observed in the nearly
allied Sjncortjnc puhhella. (See above p. 203).

7. Syncoryxe rRUTESCExs, Allmaii.
Plate VI, figs. 4—6.

TROPHOSOME. — Hyduocaulus much branched, rising from a creeping stolon,
and attaining a bciglit of from one to tv^•o inches ; perisauc sraooth. Hydranths
oval, witb about fourteen tentacles.

GONOSOME. — Plaxobeasts springing by short peduncles from the body of the
hydranth at the proximal side of the tentacles ; transverse and vertical diameters nearly
equal ; umbrella studded with scattered tln-ead-celis ; marginal tentacles with slightly
elevated clusters of thread-cells, which render it verucose, but not properly moniliform ;
ocellus distinct ; manubrium with the mouth surrounded by a ring of thread-cells.

Colour . — Hydranth light [)ink ; manubrium of medusa, and basal bulbs of mai'ginal tentacles
light brownish red.

Bevdopment of Gonosome. — October.
Habitat. — On floating timber.
Locality/. — Kingstown, Dublin.

1 have onlv once met with this species, which approaches in uuuiy respects to the S^ncoiyitc
decipiens of Dujardin. From this species it is chiefly distinguished by its more oval Iiydranths
and their more numerous tentacles, as well as by the less moniliform disposition of the groups of
thread-cells in the marginal tentacles of the planoblast.

It was found attached to floating logs in a reservoir exposed to the tide, and constantly
supplied Ijy sea-water from Dublin Bay.

382 SYNCORYNE EXIMIA.

8. SyNCOUXXE EXIMIA, AUmOM.
Plate V.

CoiiYNE LiSTERii, — Aldcf, Catal., p. 12.

CoRYNE EXIMIA, — Allmdii, ill Aim. Nat. Hist, for August, 1859. Alder, Suppl. Catal., p. 3.

Hincks, Brit. Hydr. Zooph., p. 50, pi. ix, fig. 2.
Syncoryne ExniiA, — AUman, in Aim. Nat. Hist, for May, 1861.

TROPHOSOME. — Hydrocaulus much and irregularly branched, but with a
tendency to a unilateral disposition of the ultimate ramuli, springing from an adherent
stolon, and attaining a height of from two to three inches ; peuisaec for the most part
annulated at the origin of the branches, the annulation gradually losing itself in
shallow rugfc on the short ultimate ramuli. Htdbaisiths much elongated, with twenty
to thirty tentacles, the four distal tentacles forming a verticil.

GONOSOME. — Planoblasts scattered upon the body of the hydranth, where they
are borne on rather long peduncles, which spring each from the axil of a tentacle ;
umbrella with the transverse and vertical diameters nearly equal, its outer surface
studded with scattered thread-cells ; marginal tentacles very extensile, when extended
moniliform with spherical clusters of thread-cells, the terminal cluster the largest ;
basal bulbs with a very distinct ocellus ; manubrium scarcely passing beyond the
middle of the umbrella cavity, and having the mouth surrounded by thread-cells.

Colour. — Hydranths l)rip;lit pink, cocuosarc pale pink, visible through the clear straw-coloured
perisarc. Manubrium pale pink with bright rose-coloured base ; basal bulbs of marginal tentacles
rose-colour; ocellus deep carmine.

Development of Gonosome. — -Summer.

Habitat. — Attached to rocks, the stems of Laminaria, &c., at extreme low water springtides.

Batliymetrical Distribution. — Laminarian zone.

Localities. — Pirth of Forth and Shetland Seas, G. J. A. ; coasts of Northumberland and
Durham, Mr. Alder and Mr. Hodge ; south coast of Devonshire, Mr. Ilincks.

Syncorijne eximia is perhaps the finest species hitherto met with of the beautiful genus to
which it belongs. Its large size, the bright colour of its hydranths, often thickly set with budding
medusse, the graceful motions of those medusa when, after freeing themselves from the tropho-
some, they sweep through the water in paths which the eye is never tired of following, their
wonderfully extensile tentacles, like long strings of pearls streaming from the bell-margin in ever
varying curves, — all contribute to render Syncoryne eximia one of the most singularly beautiful
liydroids of our coasts. It is a lover of the laminarian zone, where the hunter at low water will
easily recognise it by its large intricate straw-coloured tufts tinged with pink, and its terminal
rose-coloured hydranths.

It is nn abundant, hydroid on the northern shores of Great Britain.

SYNCORYNE RETICULATA. 283

9. Stncortne ferox, Slrethill JVrifjht.

CoRYYNE FEROX, — StrethUl U'riffht, in Proc. Roy. Pln'sical Soc. Ediii. for Feb., 18G3.
Hincks, Brit. Hydr. Zoopli., p. 319.

TROPHOSOME.- — Htdrocaulus " single, smooth." Htdranths with the " ten-
tacles thick, short, having the capitate cluster of thread-cells scarcely larger than the
width of the tentacle."

GONOSOME. — Medus.e " developed heneath the tentacles, similar to those of
Corijne decqjiens."

I know nothing of this species except from Dr. Wright's description, which is chiefly occu-
pied by a comparison of it with the Sj/ncori/iie deci^nens of Dujardin, a hydroid which he informs
us *S'y/?co/y/«'/e?-o.r "resembles in the shape and mode of development of the medusoids ;" but
from which it differs in its " more robust and clumsy habit." He notices also the existence on
the hydrantli itself of a delicate perisarcal investment (" coletoderm" of Wright), which "passes
over the whole body and tentacles," causing the liydranth " to assume a dirty appearance, as it
often seems to support a downy growth of very minute algae."

It is probably a Firth of Forth species, though Dr. Wright makes no mention of either its
habitat or locality.

*^* 10. Syncortne reticulata, Alex. Agassiz.

Syndictign reticulatum, — Alex. Agassiz, in Agas. Contr. Nat. Hist. U.S., vol. iv, p. 34,

and in Illustr. Catal., North American Acad.,
p. 177, figs. 290 — 297.

TROPHOSOME.— HiTDROCAiiLiTS simple, or very sparingly branched, slender, not
more than one tenth of an inch in height. Htdranths large ; tentacles short, eight
or ten in number.

GONOSOME. — Planoblasts developed from among the tentacles on the lower
part of the body of the hydranth ; umbrella of nearly uniform thickness throughout,
with the transverse and vertical diameters nearly equal, and having its surface covered
with large scattered thread-cells, between which long, narrow, superficial cells are so
disposed as to give it a reticulated appearance ; velum well developed ; length of
manubrium somewhat exceeding half the height of the umbrella cavity ; marginal

37

284 PROVISIONAL SPECIES.

tentacles with lar^e, spirally arranged, closely set clusters of very elongated thread-
cells, the clusters increasing in size towards the extremity, where the tentacle terminates
in a club-shaped cluster larger tlian the others ; basal bulbs large, ovoid, each with an
ocellus.

Colour of Medusa. — A very light metallic blue tinge; basal bulbs of tentacles light brown.
Habitat. — Growing on Biphasia rosacea.
Locality. — Massachusetts Bay, Alex. Agassiz.

For the hydroid just described Mr. Alex. Agassiz has instituted a new genus, under the
name of Si/ndiciyon. I can find, however, nothing in his description of the hydroid to justify its
separation from Syiicoryne. The pecuHar network of superficial cells in the umbrella walls, which
seems to have suggested both generic and specific names, is an interesting histological feature,-
but not in itself a character which can be employed in generic diagnosis.

Certain Sarsia-\'\\& medusae found by A. Agassiz swimming freely in the open sea are
regarded by him as the adult form of the medusa of the present species. They may be so, l)ut
the evidence is not sufficient to justify us in regarding the identity as proved.

PROVISIONAL AND DOUBTFUL SPECIES.
" SvNCORYNE LiSTERii," Fan Beneden.

The following is Van Beneden's diagnosis of this species : —

" Tige cornea, annelee assez regulierement dans presque toute sa longueur, ramifiee. Ten-
tacules au nombre de seize environ." ' Recherches sur rEmlnyogenie des Tubulaires,' p. 54,
pL III, figs. 11, 12.

He subsequently informs us that the medusae " sont de forme spherique et apparaissent a la
hauteur du verticille inferieur ;" and he adds, " ils portent pensons-nous, quatres cirrhes." 'Faune
Lit. de Belg.,' p. 120, pi. V. fig. 5.

It is impossible to determine with precision the hydroid which Van Beneden had in view
under the designation of Syncoryne Listerii, or, indeed, to construct from the characters wdiich he
assigns to it, a satisfactory diagnosis ; and the difficulty is in no way diminished by tlie appended
.synonymes which certainly apply to a very different hydroid from that described and figured as
Syncoryne Listerii in either of the memoirs cited above.

M. Van Beneden believes his Syncoryne Listerii to be identical with tlie hydroid which under
the generic name of Coryne, but without a specific designation, has been described by Lister in
the Philosophical Transactions for 1834. This is undoubtedly a wrong determination. The
gonophore which M. Van Beneden assigns to his Syncoryne Listerii is phanerocodonic and
apparently that of a true Syncoryne, though his figure shows it in too imperfectly developed a
stage for accurate determination ; while Lister's hydroid possesses simple sporosacs and is a true
Cori/ne, almost certainly the Coryne vayinata of Hincks, and of the present monograjdi.

PROVISIONAL SPECIES. 2S5

" Syncokvne .Tohnstoni," Van BeiicJeii.

The hydroid to which M. Van Beiieden gives the above name is described by iiiiu as follows : —
" Les tentacules sont au nombre de douze, places sur troisrangs, s'etendant pendant le repos de
maniere a dcpasscr la longneur du corp ; le corp est legerement brunatre.

" La colonic est tres-irregulierement ramifiee et rampante : on voit souvent des tiges
droitcs et longues s'elever a peu de distance en dessous du coip des polipulcs.

" Le polypier meme est transparent, d'un jaune dory plus on nioins tortucux." ' Recherches
sur la Faunelit. de Belg.,' p. 120, pi. V, figs. 1—3.

Though neither figure nor description is given of the gonophore of this hydroid, we can
gather from an allusion made to it by M. Van Beneden that it is medusiform. Without, how-
ever, a fuller description of the gonosome, the species cannot be regarded as otherwise thm
provisional.

" Si'NCORYNE LovENii," Van Benedcn.

The hydranth in this species is described by M. Van Beneden as being very much elongated,
and having its tentacles short, and arranged in four or five alternating verticils with four tentacles
in each verticil; the colony ramified, with, the branches appearing at rather regular intervals and
never tortuous ; the perisarc thin and of a golden-yellow colour. See ' Recherches sur la Faune
lit. deBelg.,' p. 121.

If the verticillate arrangement of the tentacles were complete, this character would point to
a relation with Stauridium, and would necessitate the removal of the hydroid from the genus
Syncoryiie. We often, however, meet with a tendency in the scattered tentacles of this genus to
assume a verticillate disposition without forming true verticils, which is, perhaps, all that M. Van
Beneden desires to assert by the phrase " tentacules en etages."

No mention is made of the size attained by this hydroid, nor is there any description of the
gonosome. I have little doubt, however, that the species is distinct from any hitherto descril)ed,
though we must wait for the discovery of the gonosome before we can refer it with certainty to
its proper genus. It must, therefore, for the present be placed on the list of provisional species.

Under no circumstances, however, can it retain the name of Syncori/ne Lovenii, this name
having been already proposed by Sars for another species, the Syncoryne ramosa of Loven, who
thus designated his hydroid under the belief that it was identical with the Slijmla ramosa of
Sars, an identification which Sars has since shown to be incorrect. See above, p. 270.

" CoRYNE R0S.4RIA," Ale.T. Ayassiz.

Under the name of " Coryne rosaria," Mr. Alex. Agassiz describes a large /SV/rsm-like
medusa found swimming freely in the open sea on the Pacific coast of North America, and which
he believes himself justified in referring to acorynoid trophosome, which he describes as profusely
branched, attaining a height of from three to three and a half inches, and with its hvdranths
" slender and supported on remarkably long and attenuated stems.'' ' Illnstr. Catal.,' p. 177.

His description of the adult medusa, which he tells us " attains an enormous size, measuriuL"-

286 CORYNITIS.

more than an inch in polar diameter," is very full, and is accompanied by a figure. He o-ives
no figure, however, of the trophosome, and his description of it is hardly detailed enough for a
sufficiently distinctive diagnosis, while he adduces no evidence to connect with it the free-
swimniinK medusa.

COEYNITIS, M'Cradf.

Name. — From Kopmn], a club, in allusion to the club-shaped hydranths.
Halocharis, Agassiz.

TROPHOSOME. — Hydrocaulus absent, so lliat the hydrautlis are sessile on tlie
iixDRORHizA.^ Hydranths cylindroid.

GONOSOME. — Planoblasts springing from the hydranth. Medusa (wlien fully deve-
loped) having a deep umbrella ; manubrium massive, and destitute of oral tentacles ;
marginal tentacles four, each with a clavate extremity and with an ocellus on the
bulbous base ; roof of umbrella cavity rising in four overarched spaces, between the
roots of the radiating canals.

The trophosome of the present genus was described and figured by Agassiz xmder the name
of Halocharis." He was unacquainted, however, with the gonosome, and he shortly afterwards
changed this name to that of Corynitis, for he found that his Halocharis was identical with
Corynitis, a genus which had been previously instituted by M'Crady' for a liydroid, whose
gonosome, as well as trophosome, had been examined by this naturalist.*

The genus Corynitis comes very near to Syncoryne, from which it is separated chiefly by
the undeveloped or rudimentary hydrocaulus, and to Gymnocoryne, so far, at least, as its
trophosome is concerned, for the gonosome of Gymnocoryne has not yet been discovered. From
the last-named genus Corynitis is distinguised mainly by the fact that its distal tentacles do not
form a distinctly defined verticil.

' This character is adopted from Agassiz's description ; it is highly probable, however, that a
rudimental hydrocaulus will be found to e.'cist as in Clava.

- 'Contrib. Nat. Hist. U.S.,' vol. iv, p. 239, pi. xx, f. 10.

^ McCrad}', ' Gymnoph.,' p. 131.

* It is true that the name of CorijnUis had been already used by the botanist, for Sprengel had
employed it to designate a genus of Leguminosce. The Coryyiiiis of Sprengel, however, is synonymous
with Coronella, De Cand., and as this name has precedence in point of publication, Corynitis must
give way to it, and becomes thus freed for appropriation by the zoologist.

CORYNITIS AGASSIZII. 2S7

*^* CouTNiTis Agassizii, M' Crudij.

CoRYNiTis Agassizii, — McCrady, G3'irinoplitlialmata of Charleston IIarl)oiir, p. 133, pi. ix,

figs. 3 — -8. Agassi::, Coutr. Nat. Hist. U.S., vol. iv,
p. 340.

Halocharis spiralis, — Aytissiz, Contr. Nat. Hist. U.S., vol. iv, p. 239, pi. xx, figs. 10 — 10 c.

TROPHOSOME. — Hydhanth, when extended, liaving the form of a slender
cylinder, with the spirally arranged tentacles successively increasing in size from below
upwards.

GONOSOME. — GoNOPHORES borne among the lower tentacles, or just below them.
Medusa, when mature, conico-campanulate about three tenths of an inch in height ;
umbrella wall set witli scattered clusters of thread-cells, and with its summit very
thick ; marginal tentacles thick, and nodulated with large pads of thread-cells,
generative elements occupying about the proximal two thirds of the walls of the
manubrium, which is thence continued as a more contracted tube towards the obscurely
lobed mouth.

Colour of mature medusa. — IManuhriuiu deep red, with the generative masses in its walls of
a rich orange; basal bulbs and claviform terminations of marginal tentacles red.

Habitat. — " Growing on sponges a little above dead low water mark." — McCrady.
Bathi/mt'trical distribution. — Laminarian zone.
Localities. — Charleston, South Carolina ; McCrady.

The scattered tentacles of the hydranth assume in this species a more decided approacli to a
spiral arrangement than is generally met with in the members of its family.

The trophosome has been described by both McCrady and Agassiz. McCrady has further
shown that the medusa, while still young, has only two marginal tentacles, and that at this period
its umbrella walls are thin, and the overarched spaces in the roof of the umbrella cavity are absent.
These overarched spaces are very striking in the fully developed medusa, and the manulu-ium has
then the appearance of being suspended from them in such a way as to have led McCrady to
compare it to " one of the massive pendants hanging from a gothic ceiling."

While there is every reason to believe that McCrady is right in referring to his corynoid
trophosome the adult medusa whose description he has given, we must not lose sight of the fact
that he has not traced it by continuous observation to the bitentacular form which the medusa
presents whUe still attached to the trophosome ; the evidence of identity being afforded by what
he seems with justice to regard as intermediate stages of development, but which were seen onlv
in specimens captured in a state of freedom in the open sea.

288 GYMNOCORYNE CORONATA.

GYMNOCOEYNE, Hinch.

Name. — From yvfxvoq, naked, and curyne, a genus of hydroids, in allusion to the absence of a
chitine-clothed hydrocaidus.

TROPHOSOME. — ^Hydkocaulus undeveloped. HrDEORnizA a filiform stolon
invested with a chitinous perisauc. Htduanths sessile on the hydrorhiza, with
numerous tentacles, the distal tentacles disposed in a vertical round a conical hypo-
stome, the others scattered over the body.

GONOSOME not known.

The genus Gi/mnocorijne was established by Hincks for a little hydroid which he olitained
on the southern coast of Devonshire. Its most striking character by which it is distinguished
from all other coryniform trophosomes except that of Corynitis consists in the non-development
of a hydrocaulus.

It comes very near to the genus Corynitis, from which, indeed, in the absence of all know-
ledge of the gonosome, it is chiefly distinguished by the verticillate disposition of its distal ten-
tacles, and though a tendency to this disposition is seen, according to Agassiz, in Corynitis, it
does not seem to be there much more marked than it is in many species of Coryne and
Syncoryne, while according to j\Ir. Hincks the oral verticil of tentacles in Gymnocoryne is so well
marked and constant as to recall the characteristic verticil of capitate tentacles which surround
the mouth in Clavatella.

GXMNOCORYNE CORONATA, Ulncl'S.
Gymnocoryne coronata, — Hincks, in Ann. Nat. Hist., for Aug., 1871, pi. v, fig. 1.

TROPHOSOME. — Hydranths " very minute, slender, enlarging slightly upwards,"
with forty or more tentacles, eight or nine of which, thicker than the otliers, and
with larger capitula, form a verticil, encircling the oral extremity ; the others scattered
over more than three fourths of the body.

GONOSOME not known.

Colour. — Central })art of body reddish, hypostome opacpic white.
Locality. — Salcombe Bay, Devonshire, Mr. Hincks.

GEMMARIA. 289

This elegant little hydroid was found by Mr. Hincks inhabiting a deserted bivalve shell in
Salconibc Bay, on the soutli coast of Devonshire. He informs us that the hydranths do not
appear to measure more than one sixth of an incli in height. Tlie verticillate tentacles are stout,
and with their capitula large ; the seattered tentacles are very slender, with small capitula ;
and decrease very markedly in size towards the base of the hydranth. The hydranths when
fully extended become very slender. Hincks could not satisfy himself that a hydrocaulus was
here present even in the rudimental condition of this part in Clava. It is probable, however, that
its very imperfect development caused it to be overlooked.

GEMMARIA, WCradij.

Name. — Yxom. gemma, a gem ; so named from the sacs of gem-like thread-cells in the
walls of the umbrella.

Zanclea, — M'Crady, Gyrauophth.
Gemmaria (provisionally), — M'Crady, id.

TROPHOSOME. — Htdrocaulus developed, invested by a perisarc, and rooted by
a creeping filiform utdroriiiza. Hydranths claviform, with scattered capitate
tentacles.

GONOSOME. — Planoblasts developed from the body of the hydranth. Um-
brella at time of liberation deep bell-shaped ; manubrium moderately developed,
destitute of oral tentacles or lobes ; marginal tentacles two, developed from the distal
extremities of two opposite radiating canals, the alternate canals having each a small
tentacular tubercle at the corresponding point ; the tentacles commence each wnth a
large bulbous dilatation destitute of distinct ocellus, and are for the remainder of their
course set with pedunculated sacs filled with thread-cells ; from the base of each tentacle
and intermediate tubercle, a ccecal claviform tube, fiUed with thread-cells, extends for
some distance in the walls of the umbrella, parallel to the corresponding radiating
canal.

M'Crady captured in the open sea at Charleston, South Carolina, a medusa which he
regarded as an immature specimen of a form referable to the genus Zanclea of Gegenbaur. Having
some doubts, however, as to the correctness of this determination, he suggested the name of
Gemmaria as a provisional designation to be used in the event of more conclusive observations
failing to justify the allocation of the medusa in the genus Zanclea}

^ M'Crady, ' Gymnoplithalmata of Charleston Harbour.'

290 GEMMARIA IMPLEXA.

Some time ago' I described the medusa of Alder's Coryne implexa, and afterwards" drew
attention to the fact that this medusa was in all essential points similar to that described by
M'Crady, whom I followed in referring the medusa to Gegenbaur's genus Zanclea. A more
careful comparison, however, has since convinced me that while the American and British forms
present no difference which could justify a generic separation from one another, they cannot be
referred to Zanclea, and must therefore receive the generic name of Gemmaria proposed pro-
visionally by M'Crady.

In thus substituting the name of Gemmaria for that of Zanclea I find myself in accordance
with Mr. A. Agassiz, who has already recognised the difference between M'Crady 's and Gegen-
baur's medusae.^ Mr. A. Agassiz has also found in ]\Iassachusetts Bay another medusa which
he regards as a second American species of Gemmaria. Neither of the Amei'ican medusas, how-
ever, has been traced to its trophosome.

Gemmaria implexa, Alder.
Plate VII.

TuBULARiA IMPLEXA, — Alder, Catalogue Zooph., p. 18, pi. vii, figs. 3 — 6.

Coryne pelagica, — Alder, Catal. Zooph., p. 13, pi. vii, figs. 1, 2.

Coryne implexa, — Strethill Wright, Edin. New Phil. Journal. Alder, Suppl. Catal.

Zooph., p. 3, pi. X, fig. 4.
Coryne Briareus, — Allman, in Ann. Nat. Hist, for July, 1859.
Zanclea implexa, — Allman, in Ann. Nat. Hist, for May, 1864, and June, 1864.

Hincks, Brit. Hydr. Zooph., p. 59, pi. ix, fig. 3.

TROPHOSOME. — Hyduocaulus consisting of simple or rarely branched offsetts
from a creeping net-like iitdrorhiza, springing at short intervals from the hydrorhiza,
and attaining a height of about half a hne ; pebisarc, for some distance from the
hydrorhiza, transversely corrugated and opaque, and then becoming abruptly trans-
parent and smooth, and forming a cup-like dilatation, which is separated by a
considerable interval from the contained coenosarc, and extends as far as the base of
the hydranth. Hydbanths, when extended, cylindrical, exceeding by many times the
hydrocaulus in length, when contracted thickly club-shaped ; tentacles numerous,
forty or fifty.

' Notes on the Hydroid Zoophytes, in ' Ann. Nut. Hist.' for July, 1859. The hydroid is there
named Coryne Briareus. Wright and Alder, however, maintain its identity with the C implexa, Alder,
and though neither Alder's nor Wright's description will apply in all points to my hydroid, I prefer to
follow these zoologists in regarding the two hydroids as specifically identical, rather than run the risk
of further confusing the descriptive zoology of the Hydroida hy the retention of a doubtful species.

°- • Ann. Nat. Hist.' for May, 1864.

^ A. Agassiz, 'Illustrated Catal.,' p. 184. Hincks, however, does not recognise a generic difl'er-
euce between the Gemmaria of M'Crady and the Zanlcea of Gegenbuur. ' Brit. Hydr. Zooph.,' p. 59.

GEMMARIA TMPLEXA. 291

GONOSOME. — Planoblasts shortly pedunculate, forming a more or less dense
cluster, which surrounds the body of the hydranth in a plane which is much nearer to
the proximal tlian to the distal limit of the tentacles ; manubrium of medusa sub-
cylindrical, extending through about one half the vertical diameter of the umbrella-
cavity, receptacles of thread-cells in umbrella very superficial, extending from the
circular canal over about one quarter of the entire height of the umbrella ; thread-
cell-bearing sacs of tentacles oval, carrying a pencil of long vibratile cilia on their
free ends, and borne on very extensile peduncles.

Colour. — Hydranths white, with the gastric cavity very pale pink, perisarc on proximal
])ortion of hydrocaulus brown, on distal portion colourless. Manubrium and tentacular bulbs of
medusa brownish red.

Bevelopmenf of Gonosome. — April to June.

Habitat. — On stones in tide-pools, and on old shells, &c., from deep water.

Btttltijmetrkal Distribution. — Litoral to deep water zone.

Localities.— Coast of Northumberland, Mr. Howse, Mr. Alder, and Mr. Ilodge ; Firth of
Forth, Dr. T. S. Wright and G. J. A. ; Coast of Forfarshire, G. J. A.

The short hydrocaulus and large many-tentacled hydranth, with its body long and cylin-
drical or short and thickly clubshaped, according to its state of extension or contraction, con-
stitute marked features in this interesting hydroid. In every instance I found that those
Hydranths which carry gonophores have the body shorter and the tentacles less developed than
in those from which no gonophores are developed, but the arrest is comparatively slight, and in
no case can it be regarded as reducing the hydranth to the condition of a blastostyle.

Alder describes the perisarc of the hydrocaulus as consisting of two distinct coats, separated
by an interval from one another, the inner one annulated, the outer smootii. Nothing of this
kind was seen in any of my specimens, in which the perisarc always became very thin and trans-
parent for some distance below the hydranth, where it formed a wide tube, including a consider-
able interval between its walls and the coenosarc which occupied its axis. This interval was
traversed by numerous regularly disposed transverse ileshy processes, which radiated from the
coenosarc and became attached to the inner surface of the perisarc.

It would seem from Mr. Alder's description that the hydrocaulus is sometimes much more
developed than in the specimen here figured. I have occasionally found specimens with a some-
what longer hydrocaulus, which might even send off a branch, but the difference was never so
great as to justify our assigning to it a value greater than that of an unimportant variation.

The planoblasts have been described at length in a former part of this Monograph (Part I,
p. 224), and it is these which will especially arrest the attention of the observer, whether he
regards the curious piriform sacs filled with thread-cells which lie one over each radiating canal, or
the pedunculated sacs, also loaded with thread-cells, which are borne in multitudes along the
entire length of the marginal tentacles. The peduncles of these sacs are singularly extensile,
and may sometimes be seen stretched out to a great length, carrying the sacs on their extremities,
and giving to the tentacles the appearance under the naked eye of being covered with a dense
growth of some parasitic mould. On being touched they immediately contract, and become

38

292 DICORYNIDiE.

closely aggregated along the tentacle. It is in this retracted state that the tentacular appendages
have been represented by Alder in his figure of this species.

DIC0RYNID.1J.

TROPHOSOME. — Hyduocaulus developed, invested by a perisarc. Htdranths
with verticillate tiliform tentacles.

GONOSOME. — GoNOPUORES in the form of natatory ciliated sporosacs, with two
simple, ciliated basal tentacles.

DICORYNE, Allman.

Name. — Sic, double, and /copu'i'ij, a club ; so called from its two club-shaped zooids — the hydranth

and the blastostyle.

TROPHOSOME. — Hydrocaulus consisting of branched or simple stems, which
arise at intervals from a creeping filiform hydrorhiza. Hydranths fusiform, with
a single circlet of filiform tentacles surrounding the base of a conical hypostome.

GONOSOME. — Sporosacs developed upon blastostyles, detaching themselves as
natatory planoblasts, ciliated over their entire surface, and having two filiform tenta-
cles diverging from the proximal end.

The very remarkable sporosacs of Bicoryne separate it by a wide interval from every other
genus. These sporosacs when mature become free, and lead a life as natatory as that of a
medusiform planoblast. Natation, however, is effected not by the systole and diastole of an
luubrella, but by the vibration of the cilia with which the entire surface of the sporosac is clothed.
So important are these characters, that Bicoryne must become the type of a distinct family of
gymuoblastic hydroids.

DICORYNE CONFERTA. 293

DicoRTNE CONFERTA, Alder.

Plate VIII.

EuDENDRiuM ? coNFERTUM, — Alder, Catal., p. 13, pi. i, figs. 5 — 8.

DicoRYNE STRICTA, — Allman, in Ann. Nat. Hist, for Nov., 1859.

DicoRYNE CONFERTA, — Allman, in Ann. Nat. Hist, for August, 1801. Althr, Suppl.

Catal., p. 4, pi. viii, figs. 1, 2. Hincks,Zv\\,. llydr. Zooph.,

p. 105, pi. xviii, fig. 1.

TROPHOSOME. — Htdrocaulus composed of much-branched stems, which attain
a height of about half an incli, with usually some short simple ones, all crowded on a
creeping, closely reticulated hydeorhiza ; branches subalternate, ascending at a very
acute angle from the stem ; perisauc continued as a cup-like expansion for a very
short distance over the base of the hydranth, marked by transverse though rather
indistinct corrugations towards the hydrorhizal extremity of the stem. Htdrantii
with about sixteen tentacles alternately elevated and depressed in extension.

GONOSOME. — Blastostiles claviform, borne on the extremities of cylindrical
stalks, which are covered with a perisarc, and spring, some from the main stem and its
branches, others directly from the hydrorhiza. Gonophores oviform on short
peduncles, densely clustered round the blastostyle ; planoblast, with its two diverging
tentacles, nearly equalling in length the rest of the zooid.

Colour. — Perisarc opaque earthy brown; hydranths ash-bvovvn.
Development of Gonosome. — Summer and autumn.
Habitat. — Investing the shells of Buccinmn and of other Gasteropoda.
Bathpiietrical Bisfribtition. — Larainarian and Coralline zones.

Localities. — Orkney Islands and Firth of Forth, G. J. A. Shetland Islands, Rev. A. M.
Norman. Coast of Northumberland, Mr. Alder.

Some years ago I dredged from a depth of about three fathoms in the Orkney Seas an old
Buccinum shell which was invested with a moss-like covering due to a hydroid which proved on
examination to possess characters which excluded it from every previously defined genus. I
accordingly assumed it as the type of a new genus, and named it Dicoryne sfricta.

The same hydroid, however, had been previously described by ]\lr. Alder as a Eudendriiun
under the name of Eudendriuw confertum, though with some doubt as to the correctness oi'
referring it to this genus ; and as his description was drawn up from imperfect specimens, no
gonosome having been developed in them, I had overlooked the identity of his hydroid with
mine. There can, however, as Mr. Alder afterwai'ds ascertained, be no doubt of this identity,

294 BIMERID^.

and though the generic name of Dicori/ne must be retained, the specific name of stricta must
yield to that of covferfa.

The trophosome of Dkoryne conferta resembles that of Heterocordi/le Conyhearei so closely
that it is difficult to find between them differences to which even a specific value can be attributed.
It is, however, quite another thing with the gonosome. That of Dkoryne is widely different
from that of Heterocordyle ; for while the gonophores of Heterocordyle are constructed upon the
common type of the sporosac, those of Bicoryne present a structure which is absolutely unique,
having nothing like it in any other hydroid with which we are acquainted.

Dk-oryne conferta resembles Hydractinia echinata, in the fact of its selecting for its
habitat the surface of univalve shells. While Hydractinia echinata, however, is almost always
found on dead shells, whose cavity has been taken possession of by a hermit crab, Bicoryne
usually occurs on shells still occupied by their molluscan owners.

A more detailed account of Bicoryne conferta is given above (Part I, p. 226), where it has
been made a subject of special anatomical study.

BIMEUIBJE.

TROPHOSOME. — Htdeocaulus developed and Invested with a perisaec, or
rudimental. Htdkanths with the hypostome not abruptly differentiated ; tentacles
filiform in a verticil round the base of the hypostome.

GONOSOME. — GoNOPHOEES in the form of fixed sporosacs.

GAEVEIA, StretUll Wright.

CoRYTHAMNioN (provisionally), — Allman.

Name. — From Inch Garvie, one of the rocky islets in the Firth of Forth.

TROPHOSOME. — Hydrophtton invested with a conspicuous perisarc, and con-
sisting of a branching htdrocaulus, rooted by a filiform nxDRORHizA. Htdranths
fusiform, with filiform tentacles, which are disposed in a single verticil round the base
of a conical hypostome.

GONOSOME. — Sporosacs developed from the hydrophyton, where they are borne
each on the summit of a short ramulus.

It was at low water spring tides on some of the small rocky islets of the Firth of Forth,

GARVEIA NUTANS. 295

that I first obtained the hydroid on which the genus Garveia has been founded. Not finding any
pubUshed account of it I described it ^ as a new Edimdriinii-Y)ke, hydroid, but though I beUcved
it entitled to rank as a distinct genus, I abstained from publishing it as such without further
comparison, considering it safer to wait for the result of a critical examination of the various
Eudendnim-\\ke hydroids with which I was then engaged. I therefore recorded it simj)ly as a new
species, under the name of Eudendrium baccatum, suggesting for it provisionally the name of
Coryihamnion^ to be used in the event of its generic distinctness being established.

Simultaneously with the publication of my description of the new hydroid there appeared a
paper by Dr. Strethill Wright,' in which the same animal was described under the new generic
name of Garveia. As Dr. Wright's paper, however, purports to be a report of a communication
made by him in the previous January to the Royal Physical Society of Edinburgh, I willingly
yield to his claim of priority, and allow the name of Garveia to take precedence over that of
Corythamnion.

One of the most remarkable characters of this curious genus is the position of the
gonophores, each being borne on the summit of a true branchlet, where it takes the place of the
hydranth on an ordinary branchlet.

The gonophore-bearing branchlets are much smaller than those which bear the hydranths, and
at first sight might be mistaken for the mere peduncles of the gonophores. A little examination,
however, will show that they are proper branchlets, not only invested by their perisarc like the
rest of the hydrophyton, but quite distinct from the peduncle of the gonophore, which they
support upon their summit. In the only representative of the genus as yet known, they are
dilated at the summit exactly like the hydranth-bearing branchlets, and from this funnel-like
dilatation the true peduncle of the gonophore springs.

Gaeveia nutans, Strethill Wright.
Plate XII, figs. 4—11.

Garveia nut.'Ins, — Strethill Wright, in £din. New Phil. Journal for July, 1859,
pi. viii, fig. 5. Hincks, Brit. Hydr. Zoopli., p. 102, pi. xiv,
fig. 4.

Eudendrium bacciferum [Corythamnion bacciferum provisionally], — All/nan, in
Ann. Nat. Hist, for July, 1859.

TROPHOSOME. — Htdeocaulus attaining a height of about half an inch, much
branched, with the ultimate branchlets for the most part regularly alternate ; main
stem fascicled ; peeisaec slightly corrugated on the branches, but without proper

' 'Ann. Nat. Hist.' for July, 185'.).

icnpmn, a club, and Oa/^viov a liltie shiulj.
* ' E(liiil). New Phil. Journal' for July, 18r)9.

29G GARVEIA NUTANS.

annulation, and continued over the body of the hydranth in the form of a funnel-
shaped cup ; ultimate branchlets frequently bent abruptly to one side just below the
hydranth. Hydkanths with about ten tentacles.

GONOSOME. — GoxopnoREs large, oval, borne on long peduncles, which rise out
of the summit of short, simple, funnel-shaped ramuli, which spring from the sides of
the hydrocaulus.

Colour. — Hydrantlis and terminal portion of branchlets orange red ; main stems and
branchlets reddish-brown ; gouophores deep orange.

Development of Gonosome. — May to July.

Hahitat. — Attached to stones or growing on other hydroids or seaweeds, near the low-water
limit of spring tides.

Bathytiietrical Didribiition . — Laminarian zone.

Localities. — Pirth of Forth, Dr. Strethill Wright and G.J. A.; Morecambe Bay, Lanca-
shire, G. J. A.

Garveia nutans, though a small hydroid, is eminently conspicuous among its more sombre
associates by the bright orange red of its hydranths and of the younger portions of the hydro-
caulus, and by its very large berry-like deep orange sporosacs ; while the way in which the
hydranth-bearing branchlets are frequently bent with an abrupt inflexion on themselves near their
termination serves further to impress upon it a peculiar and characteristic physiognomy.

The tentacles of the hydranth are uniformly directed upwards, showing no tendency to the
alternately elevated and depressed attitude so common in other genera.

The main stem is thickest near the root, and is here distinctly fascicled or composed of
aggregated tubes. It becomes gradually attenuated as it gives off its branches, and finally both
the main stem and the branches consist of a single tube.

The funnel-shaped extension of the perisarc which is continued over the body of the
hydranth, nearly as far as the bases of the tentacles, reminds us of the hydrotheca in the
calyptoblastic hydroids. It is, however, of a different morphological significance, and not being
free like a true hydrotheca from the body of the hydranth, is incapable of receiving the tentacles
and hypostome in contraction. In extreme retraction the body of the hydranth just below the
base of the tentacles is forced over the edge of the funnel-shaped cup in the form of a projecting
ring.

The gouophores are usually developed in great abundance, giving to the hydroid the appear-
ance of being covered with orange-coloured berries. The short ramuli from whose summits they
spring arise from the sides of the main stem and principal In-anches. From the edge of the
funnel-shaped cup in which these little ramuh terminate there usually hang the shreds of a
delicate chitinous membrane which had confined the gonophore in an early stage. Through this
membrane the gonophore subsecpiently burst its way, as it continued to be carried up out of the
cup on its elongating peduncle.

The gonophore, though not passing beyond the condition of a sporosac, possesses
rudiraental radiating canals, and is thus of much interest in its bearing on the general

BIMERIA. 297

morphology of the hych'oid gonosome. Its pcpuharitics Iiave been referred to in the former
part of the present monograph. Sec p. 44.

BIMERIA, StretUll Wright.

Name. — From Inch Bimer, one of the rocky islets in the Firth of Forth.
Manicella, Allman.

TROPHOSOME.— Htdeophtton invested with a conspicuous peeisabc, and
consisting of a branclied nTDUocAULUs rooted by a filiform nTDRORniZA. Htdkantiis
fusiform, witli a conical hypostome and a single circlet of filiform tentacles ; perisarc
continued over the body and hypostome of the hydranth to within a short distance of
the mouth, and forming a sheath on the proximal portion of each tentacle.

GONOSOME.— Sporosacs developed from the hydrophyton.

In the Annals of Natnral History for July, 18.59, I described as a new genus, under the
name of Manicella} a hydroid with the characters enumerated in the above generic diagnosis.

In the same month's number of the ' Edinburgh New Philosophical .Journal,' Dr. Strethill
Wright instituted under the name of Bimeria a new genus for a hydroid, which, notwith-
standing some important points of discrepancy between Dr. Wright's description and my own
(especially Dr. Wright's denial of the existence of a hypostome in his hydroid), is almost certainly
the same species for which I had founded the genus Manicella. There is thus no actual priority of
publication on either side ; but Dr. Wright's paper is given as a report of a communication made
some months previously to the Royal Physical Society of Edinburgh, and is the same as that in
which, as already said (see above, p. 295), he characterises his genus Garvcia. I therefore
willingly concede to an esteemed coUaborateur the right of naming the new genus, and accept
Bimeria as its legitimate designation.

Bimeria vestita, Wright.
Plate XII, figs. 1—3.

Bimeria vestita, — WriijM, in Edinburgh New Phil. Journ. for July, 1859, pi. viii,

fig. 4. Bincks, Brit. Hydr. Zooph., p. 103, pi. xv, fig. 2.
Manicella fusca, — Allman, iu Ann. Nat. Hist, for July, 18.59.

TROPHOSOME. — Hydrocaulus attaining a height of about half an inch, much
branched, and having its ultimate ramuli disposed with a regularly pinnate arrange-

' A diminutive noun, from manica, the long Roman sleeve which performed tlie office of a glove ;
tlie allusion is to the extension of the perisarc over the body and tentacles of the hydranths.

298 WRICHTIA.

ment; perisauc spirally corrugated at the origin of the ultimate ramuli, and rendered
opaque throughout by the accumulation in it of earthy particles and siliceous
sand. Hyduanths with about sixteen tentacles, which are usually held with
the alternate ones elevated and depressed in extension ; the perisarc continued over
each tentacle for somewhat more than a third of its length, as well as over the
greater part of the hypostome, on which it terminates abruptly at a short distance
from the mouth.

GONOSOME. — Sporosacs (male)^ oval, with a ramified spadix, pedunculated,
arising alternately along the length of the ultimate ramuli ; perisarc extended over
the entire peduncle and sporosac, spirally corrugated on the peduncle.

Colour. — The entire liydroid is of a dull brown colour.
Development of Gonosome. — June.

//«(5(V«i^.— Attached to other hydroids, seaweeds, &c., near low-water spring tides.
Bathjmetrical distribution. — Larainarian zone.

Localities. — Firth of Forth; Dr. Strethill Wright and G. J. A. Morecambe Bay, Lanca-
shire; G. J. A. Coasts of Yorkshire and of South Devon ; Mr. Hincks.

Bimeria vestifa is rendered very remarkable by its faculty of excreting chitine from its surface
to an extent unknown in any other hydroid. The chitinous sheaths which invest the bases of the
tentacles confer upon the hydranths a very singular aspect, suggesting the idea of a half-gloved
hand. The gonosome is completely covered, so that the only naked parts of the hydrosome are
the terminal portions of the tentacles, and the hypostome for a short distance just below the
mouth.

WRIGHTIA, Allman.

iVftwe.— After Dr. T. StielMl Wriijhl.

ATitACTYLis, — Wright.

TROPHOSOME. — Hydrophtton consisting of a hydrocaulus in the form of
simple funnel-shaped stems, developed at intervals from a creeping filiform hydro-
RHiZA, the whole invested with a conspicuous chitinous perisarc. Hydranths
retractile, fusiform, with filiform tentacles in a single circlet surrounding a conical
hypostome.

1 1 have not vet had an opportunity of examining the female gonopliores.

WRIGHTIA. 299

GONOSOME. — Spokosacs springing directly from the sides of tlic liydro-
caulus.

An important modification of Ehrenberg's genus Eudendrhnn, or rather of Van Bcneden's
amended form of this genus, was proposed some years ago by Dr. Strcthiil Wright, who drew
attention to the existence in the genus as then generally accepted of two forms of hydranth,
which he rightly considered as sufficiently distinct to justify the disuiembernient of the genus.

In accordance with this view he instituted his genus Atradylis for such forms of the older
genus as are characterised by a fusiform shape of the hydranth and a conical hypostome ; the
greater number of these species, moreover, presenting a more or less complete retractihty within the
suuunit of the hydrocaulus, though nothing like a projjcr hydrotheca is ever developed.

The forms, however, which Dr. Wright included under his genus Atrarli/Us presented two
generic types. One of these had already been characterised under the name of Periyonimus by
Sars, who described both the trophosome and gonosome ; ' while another had, under the name of
Bougainvillia, been long ago described by Lesson, who, however, was acquainted only with the
medusa.^ That the Bougainvillia of Lesson is the medusa of a hydroid whose type is found in the
Eudendrium ramosum of Van Beneden ( = Alradylis ramosa of Wright), has been shown both
by Wright and Agassiz. To this hydroid and its allied species the name of Bougainvillia must
accordingly be restored, and indeed we find Agassiz already arranging such forms partly under
Lesson's name and partly under ]S[argelis, Steeustrup's name for a medusa which can scarcely
be regarded as distinct from Bougainvillia.

Some time afterwards a third form, distinct from both Perigonimiis and Bougainvillia, obtained
admission into Wright's genus Atractylis. This was the Afracfglis arenosa of Alder, a form
which, unlike Perigonimus and Bougainvillia, was now for the first time made known.

To Alder's Alradylis arenosa, which has been accepted by Wright as a genuine Atradylis,
it would therefore be necessary to restrict the name Atradylis, if this name could under any
circumstances be retained by the zoologist. Unfortunately, however, Atradylis cannot be
admitted as the name of a zoological genus, for it had been long ago appropriated by the botanist,
being an old Linnsean name for a genus of composite plants.

There is thus no alternative but to assign to Atradylis arenosa some other generic name,
and that of Wrightia naturally suggests itself, as commemorating the labours of an original and
able observer of the Hydroida.'

' Sars, ' Faun. lit. Norv.'

"• Lesson, in 'Ann. des Sci. Nat./ v, 1836.

' The name of Wrightia had been already employed by Agassiz as a generic designation for
certain campanulariau hydroids. The forms, however, which Agassiz would include under his genus
llric/tttia had been previously assigned by Van Beneden and Hincks to two new generic groups, under
the names of Campanulina, Van Ben., and Calycella, Hincks. The IVrightia oi Agassiz must therefore
be suppressed in favour of the previously instituted genera of Van Beneden and Hincks. Mr. Alex.
Agassiz has indeed already rejected the name of Wrightia, on the grounds that the forms to which
Prof. Agassiz had applied it had been previously named. (See ' Hlustr. Catal. N. A. Acal.,' p. 71.)

39

300 WRIGHTIA ARENOSA.

Wkightia akekosa, Alder.

Atractylis ARENOSA, — Aider, Suppl. Catal., p. 7, pi. x, figs. 5 — 7. Wriyht, Quart.
Journ. Mic. Sci., vol. iii, New Series, pi. iv, figs. 7^ — 10.
Hincks, Brit. Zoopli., p. 88, pi. xvi, fig. 1.

TROPHOSOME. — Hydrocaulus minute, simple, irregular in form, and with
dilated trumpet-shaped summit, attaining a height of about one line, invested with
particles of sand, mud, or other foreign matter. Htdranths with about twelve long,
slender, muricated tentacles, alternately elevated and depressed in extension ; the entire
hydranth retractile within the dilated summit of the trumpet-shaped hydrocaulus.

GONOSOME. — Sporosacs (female') oval, very large in proportion to the size of the
hydroid.

Colour of Hj/dranfhs.—M\W white ; Wright.
Development of Gonosome. — September ; Wright.
Habitat. — On the under side of stones and the roots of Laminaria.

Localities. — Tynemoutli and Cullercoats, Coast of Northumberland, Mr. Alder; Largo,
Firth of Forth, Dr. Strethill Wright ; Coast of Yorkshire, Mr. Hincks.

This remarkable little hydroid was first descril)ed by Mr. Alder, who, however, was
unacquainted with its gonosome. He referred it to the genus Atradj/lis of Wright. It was
subsequently found in its fully developed state by Dr. Strethill Wright, who was thus enabled to
study the structure and development of the gonosome, and who, as has been mentioned in the
former part of the present Monograph (see p. 49), has shown that the ova, after escaping from
the sporosac, became enveloped in a gelatinous mass, which continues adherent to the ruptured
sac. In this mass they undergo further development, and are ultimately liberated from it as
planulae.

Mr. Hincks infoi-ms us that it is very common under stones and on Laminaria roots at
Filey Brig, Yorkshire.

^ The male gonophores have not been observed.

HYDRANTHEA MARGARICA. 301

HYDEANTHEA, Einchs.

Name. — From Hydra, a genus of hydroids, and m'floc, a flower.

TROPHOSOME. — Htdrophtton consisting of a rudiraental htdrocaulus and a
creeping filiform nYDROUuizA, the whole invested by a chitinous perisarc. Hy-
DRANTns fusiform, mouth surrounded by a circlet of filiform tentacles.

GONOSOME. — Sporosacs crowned by a ribbed cap, and developed from the
hydrorhiza.

The genus Ili/draufkea has been instituted by Mr. Ilincks for a little hydroid which he first
described under the name of AtracAylis mart/arica, but which he afterwards in a letter to myself
referred to a new genus, which under the designation of Ilydranlliea he found it necessary to
construct for it. There can be no doubt of the propriety of making this remarkable little hydroid
the representative of a new genus. From the only form with which there is any chance of
generically associating it — the Wrightia {Atractylis) arenosa, Alder — it is obviously separated by
its rudimental hydrocaulus and by the peculiarity of its gonophores.

It is much to be desired that we had a more detailed account of the curious little hydroid
on which the geiuis Hi/dranihen has been founded, and more especially of the cap-like appendage
described as crowning the gonophores. Mr. Hincks further infoi'ms us that from the base of the
gonosac there " proceed four much-branched vessels, terminating near the top of it in blind
extremities, and immediately enclosing the ova, which fill with a dense mass the interior of the
cavity." He further states that he could detect no trace of a spadix. There can be no doubt,
however, that the branched csecal tubes which embrace the ova represent a ramified spadix.

HtdRANTHEA MARGARICA, Eilichs.

Atr.\ctylis MARGARic.i, — Hiucks, in Ann. Nat. Hist, for Jan., 1863, pi. ix, figs. 3, 4.
Hydranthea maragarica, — Hincks, Brit. Hydr. Zooph., p. 100, pi. xix, fig. 2.

TROPHOSOME. — Rudimental htdrocaulus in the form of minute cup-like pro-
cesses, which spring- from the htdrorhiza, and support the hydranths and gonophores.
Htdranths elongate, tentacles about twenty-four in number, alternately elevated and
depressed in extension, every alternate tentacle having at its base a projecting cluster
of bean-shaped thread-cells.

GONOSOME. — Sporosacs large, sub-globular, pedunculate, springing from cup-
like processes of the hydrorhiza, which are situated singly or in pairs close to the base
of a hydranth.

302 STYLACTIS.

Colour ofIIi/drmil/t.—yf\\\iQ.
Habitat. — On Flustra fuliacea.
Bathymdrical didntmtion . — Coralline zone.
Locality. — Ilfracombe, Rev. T. Hincks.

Mr. Hincks informs us that the most striking character in the present species " is the series
of projecting bosses round the base of the tentacular ring. Wlien examined with the microscope
these are seen to consist of a number of elongate, bean-shaped thread-cells, which are piled
together so as to form silvery-white prominences on the lower side of the tentacles." He believes
that " they occur only on the alternate arms," where they " form a unique garniture.''

The species was dredged from a depth of ten fathoms, where it occurrred exclusively on
Flustra folia cea .

STYLACTIS, AUniau.

Name. — From aruAoc, a column, and uKric, a ray.
PouocoRYNE, — Sars.

TROPHOSOME. — IIydiiocaulus not developed. Htdkouiiiza formed by a net-
work of anastomosing- stolons invested with a chitinous peeisarc, but without a super-
ficial layer of naked ccenosabc. Htdbanths sub-claviform or cylindrical, with the
tentacles filiform, and arranged in a single circlet round the base of a conical hypostome.

GONOSOME. — Sporosacs, borne on the hydranths at the proximal side of the
tentacles.

I have fouiul it necessary to constitute the genus St^lactis for the reception of two hydroids
described by Sars, both of which he referred to the genus Fodocoryne, while one of them was
at first regarded by him as a mere form of Fodocoryne carnea, though he afterwards expressed
his belief in its specific distinctness fi'om that hydroid.

They are both, however, distinct from Fodocoryne, for, independently of other differences,
the fact of their gonophores being in the form of simple sporosacs instead of medusae renders it
impossible to admit them into the medusa-bearing genus Fodocoryne.

For an opportunity of examining living specimens of Stylactis I am indebted to Dr. Da
Plessis, of Nice, and in these I was enabled to satisfy myself that not only is the gonosome
different from that of Fodocoryne, but that the hydrophyton is constructed on a plan entirely
different from that of the hydrophyton of this genus, for it consists of a network of anastomosing
tubes, invested with a chitinous perisarc, and without any superficial covering of coenosarc, such
as we meet with in Fodocoryne and Ilydractinia. (See woodcut, fig. 79, p. 306.)

STYLACTIS SARSII. 303

In the Sti/laclis (PodocQry,)ie) fudcola of Sars the hydrorhizal network sends off nuinerous
cliitinous spines, resembling those of the true Podocori/nes and of Ili/dractinia. They appear to
be simple processes of the chitinons perisarc'

*^* 1. Stylactis Sarsii, AUman.

PoDocoRYNE cARNEA, — Surs, Faiin. lit. Norv., 1846, p. 7, tab. ii, figs. 5 — 11.
Stylactis Saksii, — AUman, in Ann. Nat. Hist, for ]May, 1861.

TROPHOSOME. — Hydeantiis with from twelve to tliirty tentacles, those carrying
gonophores quite similar in size and otherwise resembling those on which no gono-
phores have been developed.

GONOSOME. — Spobosacs globular, arranged in a verticil at about the middle
of the body of the hydrantli.

Colour. — Hydranths pale red, gonophores with yellowish-red spadix, ova crimson.
Development of Gonosome observed in March.
Habitat. — On empty univalve shells in the sea.
Bathi/melrical distribution. — Coralline zone ?
Locality. — Coast of Norway, Sars.

Notwithstanding the marked difference in the gonosome between this hydroid and the
Podocorync carnea, Sars, to whom alone our knowledge of it is due, believed it to be only a
particular state of Podocoryne carnea. In this, however, he was certainly mistaken, as he him-
self afterwards admitted." There is nothing to justify the belief that the same colony produces

' Hindis would apply Pliilippi's name of Dysmorptiosa to the forms which, under tlie name of
Stylactis, I separated geuerically from the true Podocorynes with which they were associated by Sars.
Now, there is uo doubt that Dysmorplwsa has precedence, not only over Stylactis, but over Podocorijne.
It was not, however, to the Stylactis form that Philippi intended his name Bysmor/j/tosa to apply, for
though his description is far from being as full as could be desired, it is sufficiently so to leave no
doubt that he bad before him either a true Podocorync or a Hydractinia, and, it will apply just as well
to one of those genera as to the other.

If we prefer to regard Podocoryne as the genus Philippi had iu view, it is tliis name, and not
Stylactis, that Dysmorptiosa must supplant, and then the name of Podocoryne may be applied e.x-
clusively to the Stylactis forms included by Sars along with the others under the general name of
Podocoryne. As the name of Podocoryne, however, has been already employed, for the medusa-
bearing forms alone, it will be better not to introduce further confusion into the existing nomenclature
by transferring this name to the sporosac-bearing Stylactis forms,

" ' Bidrag til Kunds. ora Middelbavet's Littoral Fauna,' in which Sars expresses his belief that
the hydroid here described is specifically distinct from Podocoryne carnea, thougii generically associated
with it. I must here acknowledge my obligations to my friend Dr. Charles Wilson, of Edinburgh, to
whose extensive acquaintance with the languages of Scandinavia I am indebted for translations
of such portions of its zoological literature as are referred to in this mouograph.

304 STYLACTIS FUCICOLA.

medusa at one time and sporosacs at another, a phenomenon of which no sufficient evidence is
afforded by any hydroid, and the present form must accordingly be separated, not only specifi-
cally, but genericaily, from Podocoryne carnea)

*i^ 2. Sttlactis rucicoLA, Sars.

Podocoryne fucicola, — Sars, jNIiddelhavets Litteral-Fauna, p. 40, tab. ii, figs. 6 — 1.3.
Stylactis fucicola, — Allnan, in Ann. Nat. Hist, for May, 1864.

TROPHOSOME. — Hydkorhiza emitting at intervals from its free surface conical
cliitinous spines, about half a line in height. Hxdranths about one line and a half
in height; those destitute of gonophores having eight to twelve tentacles, those
carrying gonophores (proliferous hydranths) much more slender than the former, and
with only from four to eight tentacles.

GONOSOME. — Spokosacs globular, shortly pedunculated, three to seven on
each hydranth, scattered or forming an imperfect verticillate group near the middle of
the hydranth.

Colour. — Hydranths pale red, spadix of gonophore pale red.
HahUaf. — Attached to sea-weeds a little below the surface of the sea.

Baih 11 metrical disfribiifioii. — Extending downwards to about a foot from the surface of the
sea in the region of Ct/stoseira."
Locality. — Messina, Sars.

' In that most pleasant and instructive book, the ' Sea-side Studies' of Mr. G. H. Lewes, it
is stated that in Aglaophenia myriophyllum, a calyptoblastic hydroid, some of the corbulse were found
"filled with eggs, and some with medusfe" (' Sea-side Studies,' p. )IQ&). So far as I can gather from
the context, it would seem that it is the true phanerocodonic medusa which is here meant, and that
the eggs did not proceed from these, but had their origin in sporosacs. There is notliing impossible
in this, and coming, as the statement does, from so accurate and conscientious an observer, and so acute
a reasoner as Mr. Lewes, who, moreover, has had the rare opportunity of examining living specimens
of Aglaophenia myriophyllum, I should not feel justified in at once rejecting it. But for all this, I
believe that there is here some error of interpretation. The structure of the Hydroida, and the
nature and significance of their zooids, has since Mr. Lewes wrote become much better known ; and if
it were really as he supposes, we should have in this case the solitary instance which would invalidate
some of the best established generalizations in the whole life history of the Hydroida.

" The genus Cystoseira replaces round a great part of the Mediterranean shores the Laminaria
of our more northern latitudes, and I have named the Bathymetrical region, which is characterised by
its growth, the " Region of Cystoseira." Unlike the Laminarian zone of our own seas, it is not
necessarily exposed during any period of the limited tide-range of the Mediterranean.

STYLACTIS INERMIS. 305

The Styladh fucicoJa was discovered by Sars iu the haven of Messina, and referred by him
to tlie genus Podocori/ne, from which, iiowever, for reasons already mentioned, it has been found
necessary to separate it. The common basis, or hydrorhiza, spreads over the upper surface of
sea-weeds, and is described by Sars as composed of creeping, horny, branched, minute tubes, often
anastomosing with one another, and sending off here and there vertical, hornv, conical processes,
similar to those which occur in Fodocoryne carnea, but longer and more numerous. He makes
no allusion to the existence of a superficial naked layer of coenosarc on the hydrorhizal network,
as in Podocoryne and Hydractinia, and it is pretty certain that nothing of the kind exists.

Those hydranths which are destitute of sporosacs have usually, when adult, eight tentacles,
which in extension are held alternately elevated and depressed. The hydranths which bear
sporosacs are as long as the others, and, indeed, sometimes even longer, but they are always much
more slender, while their tentacles are shorter and fewer, usually from two to six.

The sporosacs are situated low down on the body, at about two fifths of its length from the
tentacular circlet.

Stylacfis fiicicola has never been found, like some allied forms, on the surface of empty
shells, but always " on sea-weeds, especially on a little Ijrownish-red, doubly pinnated species
(Fucus cartilayineus, Cav., non L.) which grows in abundance on stones in the haven of Messina,
close to or within a foot of the sea's surface." — Sars.

*^* 3. Stylactis ineemis, AUman.
Woodcut, fig. 79, page 306.

TROPHOSOME.— Htdroriiiza destitute of spines, and formed by a layer of
closely approximated, tortuous, anastomosing tubes. Sterile and proliferous hydranths,
differing in size ; sterile hydranths about two lines in height, and with about twenty
tentacles, arranged in two closely set, alternating series ; proliferous hydranths about
half the size of the sterile ones, and with about twelve tentacles, also disposed in two

GONOSOME. — GoNOPnoRES oval, shortly stalked, forming an irregular verticil at
a short distance below the proximal tentacles.

Colour. — Hydranths and gonophores varying from a very pale pink to brownish.

DevelojjiiH'ni of Gonosome. — April.

Habitaf. — Spreading over the surface of sea-weeds.

Bafhymetrical Distribution. — Region of Cystoseira.

Locality. — Nice, Dr. du Plessis.

306

HETEROCORDYLE

I am indebted to Dr. dii Plessis, of Nice, for the specimens which have enabled me to
define the present species, and from which the accompaiiyinp- woodcut has lieen made. Tiiey were

developed in his aquarium, where they spread them-
selves over the surface of sea-weeds, which were
obtained from the neighbouring shores at probably
a slight depth from the surface of the sea. They
afforded me the means of determining the real
nature of the hydrophyton in the genus Sfj/hcfix,
and to show that it is very different from that of
Podocoryne, in which the surface of the hydro-
rhizal expansion is covered by a stratum of abso-
lutely naked ccenosarc, a character which Fodocoryne
possesses in common with Hydractinia. The
hydrophyton of Sh/Iadis, on the other hand, while
it is almost entirely reduced, as in Hydractinia
and Podocoryne, to the condition of a hydrorhiza,
is formed, like most other hydrorhizse, on the
common type of a network of anastomosing
chitine-covered tubes.

The hydranths of Slylactis ineniiis are sur-
rounded at their origin by a narrow cup-like pro-
cess of the hydrorhiza. They are very mutable in
shape, and run through a range of forms depending
on different states of contraction, similar to what we
meet with in Podocoryne and Hydractinia. The
tentacles, when in a semi-contracted state, assume
a somewhat club-shaped form. The sporosacs, which are clustered round the body of the
hvdrantli, at a little distance l)elow the tentacles, varied from two to six in num])er. The more
mature sporosacs were filled with ova at the time of examination.

A, a colony of the natural size, spreading over a piece of

sea-weed.

B, portion of a colony magnified, a, hydranth, from which

no gono])hores are developed (sterile hydrantli), ex-
tended ; h, same contracted ; c, gonophore-hearing (or
proliferous) hydranth.

HETEROCORYDLE CONYBEAREI. 307

HETEROCORDYLE, Allman.

Name.— Yxom steooc, dissimilar, and /vopou'Ai,, a club; in allusion to tlic two lands of club-
shaped zoids, the hydranths and the blastostylcs.

TROPHOSOME. — Htdrophyton composed of a simple or branching hydrgcaulus
rooted by a creeping filiform hydroehiza, the whole invested by a chitinous pesisarc.
Hybbanths fusiform, with a single circlet of filiform tentacles round the base of a
conical liypostome.

GONOSOME. — Sporosacs borne upon blastostyles.

The genus Hefcrocordi/le has been founded for a hydroid discovered on the southern coast
of Ireland. Its trophosome is that of Dicorym, but from this genus it is entirely separated by
its very different gonosome.

Heterocordyle Conybearei, Allman.

Plate X, figs. 4—7.

Hkterocordyle Conybearei, — Allman, in Ann. Nat. Hist, for May, 1864, and July, 1861,

pi. ii. Hincks, Brit. Hydr. Zooph., p. 107, pi.
xviii, fig. 2.

TROPHOSOME. — Hydrocaulus consisting of numerous stems crowded upon the
creeping hydroriiiza, some much branched, and attaining a height of about four
lines, others shorter and simple ; perisarc transversely corrugated, extending for a
short distance over the base of the hydranth. Hydranths with about twelve
tentacles, usually held straight in extension, with the alternate ones elevated and
depressed.

GONOSOME. — Blastostyles club-shaped, borne on the summit of very short erect
processes, which spring from the UYDRORmzA, and are invested with a chitinous peri-
sarc ; distal extremity of blastostyle thickly set with thread-cells. Sporosacs numerous,
nearly sessile, densely crowded, commencing a little behind the distal extremity of the
blastostyle, and thence extending to within a short distance of its base.

40

308 HETEROCORYDLE CONYBEAREI.

Colour. — Ash brown.

Beveloj)Ment of Gonosome. — Observed in September.
Habitat. — On old univalve shells in the sea.
Bathymetrical distribution. — Laminarian and coralline zones.

Localities. — GlengarifF Harbour, County of Cork, G. J. A. ; Oban, Argyieshire, Rev. T.
llincks.

This very distinct hydroid, the only representative as yet discovered of its genus, was dredged
in considerable abundance from a rather muddy bottom in the Harbour of Glengarifi', on the
south coast of Ireland. It was always found investing old gasteropodous shells, whose cavity
was in every instance tenanted by a hermit crab. It so closely resembles Dieori/iie conferta in
form, colour, and general habit, that it was at first mistaken for this species, and it was only the
closer inspection of its gonosome which showed how very widely it differed from it.

Ileterocordyle Coni/bearei spreads itself in the form of a dense moss-like covering over the
shell of which it has taken possession, extending in all directions by means of a reticulated
stolon, which sends up at short intervals the stems of the trophosome and the blastostyles of the
gonosome. The stems are for the most part well developed, and much, but rather irregularly,
branched, while a few simple and much shorter stems, each carrying a hydranth on its summit,
spring here and there from the creeping base. The tentacles, when extended, are held alter-
nately elevated and depressed, indicating a composition of the tentacular circlet out of two closely
approximated verticils. When partially contracted the tentacles frequently present a slight
enlargement of their extremities, giving them a somewhat clavate form. The perisarc extends
for a very short distance over the base of the hydranth, and then terminates aliruptly by a
defined edge.

The blastostyles are liorne on the summits of very short stalks, which spring like the stems of
the trophosome from the creeping stolon, and which, like them, are also invested with a chitinous
perisarc. The blastostyles are clavate, but as they admit of considerable extension and con-
traction their form is a varying one. Their distal extremity is rounded, and thickly set with
thread-cells.

The gonophores spring from the sides of the blastostyle, to which they are attached by
peduncles so short as to render them almost sessile. They are crowded upon the blastostyle, a
short space at its distal extremity and another at its proximal being the only parts free from
them. They are oviform and constructed on the type of the simple fixed sporosac ; the
spadix is large, and in the female embraces a large solitary ovum, over which the endotheca
extends as a well-marked layer, while a very distinct cctotheca, in which no evident structure can
be detected, envelopes the whole.

The male gonosacs resemble the female in form, and in the mode in which they are
grouped upon the blastostyle. The spermatogenous mass is developed round the spadix, and
is confined by an exceedingly delicate investing membrane (endotheca), while the whole is
surrounded by a distinct but apparently structureless ectothecal sac.

It was in company with an accomplished microscopist, Henry Conybeare, Esq., during a
dredging excursion in the beautiful Bay of Glengariff that this remarkable hydroid was first
obtained, and I have yielded to the gratification of dedicating it to my friend, and thus connecting
it with the name of a family to which the progress of natural science is largely indebted.

CIOiNISTES RETICULATA. 309

CIONISTES, StreiMll JFright.

Name. — Froni kImv, a ])illar, probably in allusion to the presence of blastostyles.

TROPHOSOME. — Hydrophtton consisting of a retiform iiydeoehiza without
develo])e(l iivdrocaulus. Hydranths sessile on the liydrorhiza, minute, with a single
verticilfof hliform tentacles.

GONOSOME. — Spouosacs borne by blastostyles, which spring from the
liydrorhiza.

Cioiiistes is a genus constituted by Dr. Strethill Wright for a liydroid of which he met
with\i single specimen (male) in the Firth of Forth. It appears to be distinguished from
Heterocordtjle chiefly by the absence of a developed hydrocaulus.

CiONiSTES RETICULATA, StrefliUl Wright.

CioNisTEs RETrcuLATA, — Sirethill Wright, in Ann. Nat. Hist, for August, 1861. Hincks,

Brit. Hydr. Zooph., p. 135.

TROPHOSOME. — HYDrvOnuizA " consisting of a close network of flattened tubes,"
from which the hydranths and blastostyles " spring at distant intervals." Hyduanths
" minute, white, with a single row of short tentacles."

GONOSOME. — Blastostyles gradually enlarging from the base to the summit ;
SPOUOSACS numerous, borne on the sides of the blastostyle.

liahUat. — " Growing on an old shell."
Zoc«/«7y.— Granton, Firth of Forth, Dr. S. AVright.

I have never seen this hydroid, which, judging from Dr. Wright's short description of it —
scarcely fidl enough, however, for a complete diagnosis — must be a form of considerable interest.
The only illustration which accompanies the description is a diagramatic outline of one of the
blastostyles, with a male gonophorc attached to it.

310 BOUGAINVILLIA.

BOUGAlNFILLIDyE.

TROPHOSOME.' — Htdkocaulus more or less developed, invested by a con-
spicuous PEiiiSAiic. Htdranths witli the hypostome not abruptly differentiated ;
tentacles filiform, in a single circlet round the base of the hypostome.

GONOSOME. — GoNOPiiOKES phanerocodonic, with four simple radiating canals,
and with the marginal tentacles simple, and distributed either singly or in clusters.

The bydranths of all the species included in this family are distinguished from those of the
Eudendnda, which they otherwise closely resemble, by the hypostome not being abruptly differ-
entiated, as in the latter, but, on the contrary, passing continuously into the body of the
hydranth.

BOUGAINVILLIA, Lesson.

Name. — In commemoration of the distinguished circumnavigator. Admiral de Bougainville.

EuDENDRiUM, — Vuii Beneden.
Atractylis, — Strethill JVriyht.
HippocRENE (planoblast), — Brandt.
Margelis (planoblast),— Steens^z-iyj.

TROPHOSOME.— Hyduophyton consisting of a branching htdrocatous, wdiich
is rooted by a filiform nrDuonuizA. Htduanths fusiform, with a conical hypostome.

GONOSOME.— Planoblasts borne by the hydrocaulus, and having at the time of
liberation a deep bell-shaped umbrella, with the manubrium shorter than the height
of the bell-cavity, and with four simple labial tentacles, each of which carries at its
extremity a little capitulum of thread-cells ; radiating canals, each terminating at its
junction with the circular canal in a bulb, from which two tentacles are developed,
each tentacle having an ocellus at its base.

Before attaining maturity, the labial tentacles become dichotoraously branched, and the
marginal bulbs carry each a fascicle of munerous tentacles, every one of which has an ocellus at

its base.

The genus Bou^ainvdlia was founded by Lesson for a hydroid whose medusa, however, was

the only part with which he was acquainted.

BOUGAINVILLIA RAMOSA. 311

Forbes^ bad a suspicion tliat the medusa whicli he had liimself described under the name
of Buiu/ainvillia liritamnca was related to the FAidendrimn ramosmi of Van Beneden, while
]\rCrady- was also led to refer an allied American medusa to a Ei(dendrmiii-YikG trophosome.
Strctliill Wright'^ subsequently believed that he could trace the planoblast of Van Beneden 's
Etidendrium ramosum through intermediate stages into a medusa which closely resembled the
Bougainvillia Britannica of Forbes, and the relation thus maintained received confirmation from
a similar observation afterwards made by Alex. Agassiz * on the medusa of Bougainvillia super-
ciliarix, an American species. Though the evidence from which Forbes's Bowjainvilliu Britannica
might be referred to the Eiidendriim ramosum of Van Beneden as its proper trophosome is
hardly complete, the observations of Wright and Agassiz leave no doubt as to the planoblast
of this trophosome being a true Bougainvillia medusa.

A\"right, having shown that Van Beneden's Eiidendrium ramosum cannot be retained in the
genus Eudendrium, constructed for it that of Atracfi/lisJ" It must now be referred to Lesson's
genus Bougainvillia, of which, indeed, it may be assumed as the type.^

1. Bougainvillia uajiosa, Van Beneden.

Plate IX, figs. 5—7.

HippocREN'E Britannica (planoblast), — Forbes, in Ann. Nat. Hist., 1841, vol. vii, p. 8)-, pi. i,

fig. 2.
Eudendrium ramosum, — Van Beneden, Embryogenie dcs Tubulaires, p. 3(5, pi. iv, exclusive

of syiion3'mcs. Fauue lit. de 13elg., p. 112, pi.
vi, vii.
Bougainvillia Britannica (planoblast), — Lesson, Acalepbes, p. 291. Forbes, Xakcd-eyed

Medusse, p, 62, pi. xii, fig. 1.
TuBULARiA RAMOSA, — Dcthjell, Rare and Remarkable Animals, p. 64, pi. xi.
Atractylis RAMOSA, — Stretkill IVriyht, Proc. Roy. Phys. Soc. Edinb., vol. i, p. 469.
Margelis RAMOSA, — Affussiz, Contr. Nat. Hist. U. S., vol. iv, p. 314.

Bougainvillia ramosa, — Allman, in Ann. Nat. Hist, for !May, 1864. Hincks, Brit. Ilydr.

Zoopb., p. 109, pi. xi.x, fig. 2.

' ' British Naked-eyed Medusa;,' p. 68.

- " Gymnophthalmata of Charleston Harbour," 'Proc. Elliott. Soc.,' April, 185".

' ' Proc. Roy. Phys. Soc. Edinb.,' April, 1858.

^ Agas., 'Contr. Nat. Hist. U. S.,' vol. iv, p. 291.

= See above, p. 299.

° The generic name of Marytlis was given by Steenstrup (' Vidensk. Medel.,' 1849, p. 43) to the
medusa ^vhich Forbes has described under the name of Bougainvillia Britannica, and Agassiz has
accepted the genus Margelis, believing that Forbes's Bour/ainviUia Britannica is generically distinct
from the original Hippocrene Bougainvillii of Brandt, for which Lesson founded the genus Bougain-
villia. I cannot, however, see in the differences between the two medusa; — dift'erenees which are to
he found chiefly in the wider and more rounded form of the manubrium and in the more profuse
bifurcation of the oral tentacles of Brandt's species — characters of sufficient importance for generic
separation.

31:2 BOUGAINVILLIA RAMOS A.

TROPHOSOME. — IIydhocatjlus attaining a height of from one to three inches,
much hranched, tlie ultimate branches for the most part alternate ; main stem and
principal branches fascicled towards the base ; perisarc with shallow annulations
at tlie origin of the branches. Htdranths with about twelve tentacles, which in
extension are usually carried nearly straight, with every alternate one elevated or
depressed ; perisarc continued over the body of the hydranth as far as the roots of
the tentacles, in the form of a cup-like extension, into which the hydranth, in extreme
contraction, may be almost completely withdrawn.

GONOSOME. — GoNOPHORES on moderately long peduncles, which spring from tlie
ultimate ramuli, on which they occur either singly or in sub-vcrticillate groups of two,
three, or more.

Colour. — Body of hydrantli with manubrium and marginal bulbs of medusa pale pink,
ocelli carmine, perisarc of liydrocaulus straw-colour.

Develojjment of Gonosome. — Autumn.

Habitat. — Attached to rocks near low-water mark, and to other liydroids and old shells from
the coralline and deep-sea zones.

Baflii/metrical disirihution. — Laminarian to deep-sea zones.

Localities. — Coast of Belgium, Van Beneden ; Firth of Forth, Dr. S. Wright ; Oban, Argyle-
shirc, Rev. T. Hincks ; south coast of Devonshire, Rev. T. Hincks and G. J. A. ; coast of
Northumberland, Mr. Alder.

Bougainvillia ramosa, in its complete form, presenting both trophosome and gonosome, was
first described by Van Beneden, who, believing it to be identical with the Eudendriiim ramosum of
Ehrenberg, assigned to it this name. Ehrenberg's Eudendrium ramosum, however, is an entirely
different hydroid, being identical with the Tuhularia ramosa of Linnaeus, the " small, ramified,
tubular coralline " of Ellis. We now know, as has been just said, that the adult medusa of
Vail Beneden's Eudendrium ramosum is either the Bouffaijivillia Britannica of Forbes or a closely
allied form. Were we sure of its identity with Forbes's medusa, the law of priority would
compel us to assign to Van Beneden's hydroid, not only the generic name of BoiigainviUia, Imt
the specific name of Britannica. In the absence, however, of absolute proof of this specific
identity it will be wiser to retain Van Beneden's specific name of ramosa, wdio, moreover, was the
first to give a good drawing and description of the species.

BoiigainviUia ramosa forms profusely ramified tree-like growths on the various bodies to
which it attaches itself, and in the confinement of our arpiaria will continue for many days to
throw ofi" successive crops of medusa;. The medusa, immediately on liberation, darts away through
the surrounding water, the marginal tentacles being at the same time contracted and thrown
back upon the umbrella in the form of short, thick, curved, somewhat clul>shaped appendages.
When it comes to rest the tentacles are extended, and when the medusa is floating passively on
the water they are stretched out for some distance nearly horizontally, and for the rest of their
length hang down vertically in the water.

I have not succeeded in getting the young medusae to undergo in confinement the changes

BOUGAINVILLTA RAMOSA. 313

by ^vllich tlicy l)econic converted into the mature Bougaiwcillia form. -Indging, however, from
the observations of Wriglit and Agassiz, and from the condition of the Jiouf/ainvillia Brilamiicu
medusa, which is not improbably the adult planoblast of this species, the change consists mainly
in three successive bifurcations of the labial tentacles, in the multiplication of the marginal
tentacles, by which each basal bulb, instead of carrjdng only two tentacles, as at first, carries a
fasciculus composed of many, and in the greatly increased thickness of the umbrella, whose cavity
now occupies but a small portion of the entire volume. In this state the medusa has assumed a
nearly gol)uIar form.

The labial tentacles of the medusa (Hate IX, fig. 8) to which the name of Hoti(/uinvUlia Brifan-
iiica has been assigned are not placed on a level with the month, but spring from a point some
distance above it. Every ultimate branch of these tentacles terminates in a little cluster of
thread-cells, which is not imbedded in the substance of the tentacle, but is elevated on the
summit of a very short peduncle. The entire tentacle is solid and is composed of an external
thin ectodermal layer, enclosing a distinctly cellular endodermal pith.

Each of the four fasciculi of marginal tentacles springs from a crescentic cushion-like projec-
tion of the umbrella margin. Every tentacle carries its ocellus on the iimer side of the thickened
base.

The generative elements are contained in four oval projections of the manubrium walls, which
extend symmetrically from the base of the manubrium to within a short distance of the labial
tentacles. These generative lobes, the labial tentacles, and the radiating canals with the marginal
tentaculiferous cushions, do not alternate with one another, but are situated, each set in one and
the same meridian plane.

When the medusa is floating on the water the marginal tentacles are usually extended and
thro^^•u back upon the umbrella, their slender extremities at the same time bending outwards in a
graceful curve. In this position the basal cushions become everted, and the ocelli, now directed
outwards, are fully exposed.

The manubrium, with its generative lobes, and the basal bulbs of the marginal tentacles, where
they spring from their common cushion-like support, are of a fine golden-yellow colour.

The retractility of the hydranth in BoiiyainviUla ramosa is very great. In complete retraction
the whole hydranth, to within a short distance of the tips of the tentacles, is withdrawn within the
perisfirc, which here forms a cup-ldce expansion. This cup, however, is nothing more than a thin
extension of the perisarc over the body of the hydranth, to which it is more or less closely
adherent, and during whose retraction it is thrown into transverse rugae. It is thus entirely
different from the hycbotheca of the campanularian and sertularian hydroids.

After the death and disappearance of the hydranth the ramulus which had supported
it continues to bear the gonophores, and these may now be often seen to form umbel-like groups,
terminating the branchlet.

In Van Beneden's original memoir on the present species a specimen _ is figured and
described in which some of the branches carry on their sides groups of piriform bodies, which
ajjjjcar to bud one from the other in such a way as to acquire the form of fan-shaped clusters.
I cannot offer any opinion as to the nature of these bodies. In a subsequent memoir,'
M. Van Beneden suggests the probabihty of their being male gonophores, the female gono-

' ' Ilcch. sur la Faune lit. do Belgique,' p. 1 13.

314 BOUGAINVILLIA FRUTICOSA.

phores showing themselves under the form of medasfe — a view, however, which, with the very
imperfect knowledge we as yet possess of the bodies in question, cannot be entertained without
further confirmation.

In the same figure a branch is represented which, instead of terminating in a hydranth, forms
a large fusiform dilatation. We are reminded liy it of a very similar condition occasionally
presented by certain Syncorynes, and which is there due to the presence of a parasitical Pichno-
gonidan (see p. 200). Indeed, Van Beneden describes the dilated branch in his species as
containing an oval body, which " avait bien I'air d'un parasite."

3. BoiIGAINVILLIA TRTJTICOSA, AUmUU.
Plate IX, figs. 1—4.

BouGAiNviLLiA rRUTicosA, — AUmaii, in Ann. Nat. Hist, for July, 1864. Hincks, Brit.

Hydr. Zooph., p. 110.

TROPHOSOME. — Htdrocaulus rising to the height of about two inches, much
branched, with the main stems composed of aggregated tubes, branches sub-alternate ;
PERisAuc without annulation, but marked on the smaller branches by shallow trans-
verse corrugations, w'hicli become obsolete on the larger branches and main stems.
Hydrantiis with about fourteen tentacles, which are more or less curved in extension,
with usually every alternate one elevated or depressed; body of hydranth nearly
cylindrical in extreme extension, and with the perisarc continued over it for a short
distance as a thin pellicle, which on extreme retraction appears as a membranous
corrugated cup, investing the hydranth for about a third of its height.

GONOSOME. — GoNOPnoRES on moderately long peduncles, borne upon the ulti-
mate ramuli, where they occur irregularly scattered along the length of the ramulus.^

Colour. — Hydrauths jialc red ; manubrium and marginal bulbs of medusa, up to the period
of liberation, pale red. Perisarc straw colour, becoming browaier and more opacpie on the older
parts of the hydrocaulus.

Develojnnent of Gonosome. — Autumn.

Ilahitat. — Attached to buoys and floating timber.

Localities. — Mouth of Kcnmare River, coast of Kerry, G. J. A. ; Firth of Forth, G. J. A.

' In my original description of BougainvilUa frulicosa I assumed as a distinctive character the
disposition of the gonopliores, -which I then believed were entirely confined to one side of the support-
ing ramulus. Subsequent examination has convinced me that this character is not the usual condition,
and cannot be retained in the diagnosis.

BOUGAINVILLIA SUPERCILIARIS. 315

Bouc/ainviUia fniiicosa is closely allied to BouguinvilUu ramosa. It differs from it, however,
ill the more cylindrical and more slender form of the extended hydranth, and in the fact that the
tentacles, when extended, are decidedly curved, while they are nearly straight in BougaimUlia ra-
mosa, as well as in the considerably less extent to which the hydranth is invested by the membranous
dilatation of the chitinous perisarc, and in the absence of distinct annulation at the origin of the
branchlcts. While in Boiif/aiiiviUla ramosa the contracted liydranths are almost entirely con-
cealed within the dilated perisarc, in TiougainviUia frulicosa the liydranths, in extreme retraction,
have the whole of the tentacles and at least half the body exposed. Another difference will,
perhaps, be found in the way in which the recently liberated medusa carries its marginal tentacles,
which in Bongainvillia fndicosa I have never seen extended in the plane of the codonostomc!
before bending vertically downwards, the habitual attitude of these appendages in BouyainviUia
ramosa.

Whether the mature medusa differs in any respect from that of BoiiyainvUlia Britannica,
we have not yet facts to enable us to decide.

I first met with Bou(/auwillia fntticosa loaded with gonophores and growing in abundance
on a large piece of floating timber in the estuary of the Kenmare River, Co. Kerry, and after-
Avards in less perfection, and without gonophores, on a buoy in the Firth of Forth. It is a
singularly beautiful hydroid, and when first transferred from the sea to the zoophyte trough of
the microscope, before its health and vigour have become impaii-ed by the confinement of our jars,
offers a spectacle unsurpassed in interest by any other species — every branchlet crowned by its
graceful hydranth, and budding with medusae in all stages of development, some still in the
condition of minute buds, in which no trace of the definite medusa-form can yet be detected ;
others, in which the outlines of the medusa can be distinctly traced within the transparent
ectotheque ; others, again, just casting off this thin outer pellicle, and others completely freed
from it, struggling with convulsive efforts to break loose from the colony, and finally launched
forth in the full enjoyment of their freedom into the surrounding water. I know of no form in
which so many of the characteristic features of a typical hydroid are more finely expressed than
in this beautiful species.

*^* 3. BOTJGAINVILLIA SUPEECILIAEIS, AgaSSlZ.

HippocRENE suPEiiciLiAiiis (mcdusa), — Agassiz, North American Acalephse, part i, pi. i — iii.

BouGAiNviLLiA SUPERCILIARIS, — Ayassiz, Coutr. Nat. Hist. U.S., vol. iv, p. 289, pi xxvi, figs.

1 — 7. Alex. 'Affassiz, Catal. North American
Acalephfe, p. 153, figs. 232—240.

TROPHOSOME. — Htdeocaulus attaining a height of about two inches, and
forming clusters in which the branches are given off " rather irregularly, though more
or less alternately or spirally," the principal branches and ramuli more or less dis-
tinctly annulated at their origin. Htdb.a:nths with from fifteen to twenty tentacles,
strongly marked by annular clusters of thread-cells ; hypostome " very short, forming

41

316 BOUGAINVILLIA CAROLINENSIS.

a mere conical papilla ;" pcrisarc continued as a thin film over the posterior part of
the hydranth.

GONOSOME. — GoxopnoRES scattered irregularly on the ultimate ramuli. ]\Iature
MEDUSA Avith a broad and round manubrium, and with the abial tentacles profusely
branched.

Habitat. — Attached to rocks, shells, and occasionally to seaweeds.
Locality. — Atlantic coast of North America.

The Boin/ainvillia siqjerciliari.s was first described by Agassiz, but only from the medusa,
with which alone he was at the time acquainted, and whose structure he has made the subject of
a laborious and exhaustive treatise.^ He afterwards discovered the trophosome, which he figures
in his ' Contributions to the Nat. Hist, of the United States,' while an effective woodcut of the
adult medusa, as well as of the entire hydroid, is given by Alexander Agassiz in his ' Illustrated
Catalogue of North American Acalephae,' where he has traced the medusa from the period of
its detachment from the trophosome to its fully developed state.

The mature medusa appears to belong to the form with a wide and rounded manubrium
and with profusely branched labial tentacles, to which Agassiz would restrict the name of
Boiigaitivillia, assigning to those with a more cylindrical manubrium and more sparingly
branched labial tentacles Steenstrup's name of Maryelis. I cannot, however, as already said,
accept the generic value of this difference. (See above, p. 311.)

I have never met with Boi'f/ainvillia siiperciliaris. It has not been discovered on the
European side of the Atlantic, and the diagnosis here given has been framed from Agassiz's
description.

*^* 4. BorGAiNviLiiA Caeolixensis, M'Crachj.

HiPPOCRENE Carolinexsis, — M'Crachj, Gyranopb. Charleston Harbour, p. 164, pi. x, figs.

8—10.
Margelis Carolinensis, — A(jassiz, Contr. Nat. Hist. U.S., vol. iv, p. SW. Alex. Agassiz,

Illustr. Catal, N.A. Acalepbre, p. 156, figs. 241—248.

TROPHOSOME. — Htdrocaulus attaining a height of from eight to twelve
inches ; " the main stem is stout, and tapers gradually ; the main branches begin
close to the root, and thus form clusters, from which branch off irregularly secondary
branches, which are quite slender and ramify but little." — Alexander Agassiz.

GONOSOME. — GoxoPHOREs "making their appearance all over the stem." —
A. Agassi:. Mature medusa with narrow manubrium, and with labial tentacles about
three times bifurcated.

' ' Contributious to the Natural History of the Acalephae of North America.'

BOUGAINVILLIA MUSCUS. 317

Colour. — JMain stem greyish green, liydrantlis delicate rose colour.

Habitat. — " Growing on Fiiscii veslculosus in great abundance." — A. Ar/assiz.

Bafhi/tiietrical distribiition. — Literal zone.'

Localities. — Charleston Harbour, M'Crady ; Naushon, Massachusetts, A. Agassiz.

A medusa obtained abundantly in Charleston Harbour was originally described by M'Crady
under the name of Ilippocreiie Carolinensis, and referred by him to a Eudendrinm-Yihi tropho-
some from, the same locality, but which he does not describe with sufficient detail for a satisfac-
tory diagnosis.

The same medusa was subsequently traced by Alexander Agassiz to a trophosome of which
lie has given the description just cpioted in the diagnosis of the species.

In the large size attained by the trophosome, Bougainvillia Carolinciisis far surpasses every
other species of Bougainvillia as yet discovered. It is, indeed, in this respect equalled by very
few hydroids.

The position of the gonophores on the main stem, as well as ou the branches, brings to
mind, as ]\lr. A. Agassiz remarks, the condition of the same parts in the Perigonimus muscoides
of Sars.

5. Bougainvillia muscus, Allman.

Plate X, iigs. 1—3.

Perigonimus muscus, — Allman, in Ann. Nat. Hist, for January, 1803.

Bougainvillia muscus, — Allman, in Ann. Nat. Hist, for May, ISGi. Hi7ic](s, Brit. Hydr.

Zoopb., p. 111.

TROPHOSOME. — Hydrocaulus consisting of numerous erect non-fascicled
stems, which spring at intervals from a creeping stolon and attain a height of about
half an inch ; main stems sending off short branches, which are themselves for the
most part without further ramification ; perisarc slightly corrugated. Htdeanths
with about sixteen tentacles, alternately elevated and depressed in extension ; body of
hydranth invested with a thin continuation of the perisarc, which extends nearly to
the base of the tentacles.

GONOSOME. — Gonophores borne on rather long peduncles, which spring singly
or rarely in greater number from the hydrocaulus, at a short distance below the
hydranths.

Colour. — Hydrocaulus light brown, hydranths light reddish brown, manubrium and marginal
bulbs of medusa vermilion.

' Such, at least, I assume to be the area of its distribution in depth ; the Fitcus vesictitosus, on
which it has been found by Mr. A. Agassiz, being eminently characteristic of the Literal zone.

31 S PROVISIONAL SPECIES.

Development of Gonosome. — September.
Habitat. — In tide pools near low-water mark.
Bathymetrical distribution — Laminarian zoue.
Locality. — Torquay, G. J. A.

The small size of the present species, its more simple ramification, and the fact that its
stems consist of a single tube instead of being fascicled or composed of numerous tubes coalesced
into a dense bundle, at once distinguish it from the other British species, while these features
present a combination of characters which also prevent its being confounded with any non-
British forms hitherto described.

Hincks ('Brit. Hydr. Zooph.,' p. 112) speaks of a form intermediate between Boiic/ainvillia
musciis and B. ramosa, having fascicled stems like the latter, but in other respects more closely
resembling B. muscus. It is probably a variety of B. ramosa.

The medusa, on liberation, diflPers in no respect from that of BougainvUtia ramosa at the
same period. We have as yet no evidence regarding its adult form, which probably resembles
that of Bouyainvillia ramosa.

PROVISIONAL SPECIES.
" BouGAiNviLLiA Mertensii," Alex. Agassiz.

Under the name of " Bovgainvillia Mertensii" Mr. Alexander Agassiz describes, but
without any figure, a trophosome which, he informs us, " grows cpiite luxuriantly, attaining
a height of nearly two and a half inches ; the stems are very stout, particularly in the main
branch, which, near the base, is exceedingly robust ; the branches are, at least, three times as
stout as those of the hydrarium (trophosome) of Bougainvillia superciliaris, which is slender
and always branches quite loosely."^

The trophosome thus described is referred by its discoverer to a medusa which he regards
as identical with that obtained by Mertens in Behring's Straits, and described by Brandt under
the name of Hipimcrene BougainviUii, but for which Professor Agassiz proposes the name of
" Bougainvillia 3'Iertensii."~

As, however, the medusa has not been traced to the trophosome, there is no direct proof of
the relation between the two, and the species cannot, therefore, be yet regarded as established.

The medusa appears to belong to the form with rounded manubrium and very nuich
branched labial tentacles. Both it and the trophosome were obtained upon the western coast of
Mexico, and they are thus Pacific representatives of the Bougainvillia form as seen in other species
on the east and west shores of the Atlantic.

* Alex. Agassiz, ' Illust. Catal.,' p. 152.

- Louis Agassiz, ' Cont. Nat. Hist. U.S.,' vol. iv, p. 314.

DIPLURA. 319

"EuDEiN'DUIUM rVLSILLUM," SrOS.

Among provisional species must be here included a hydroid described by Sars, in his
' Middelhavet's Littoral-Fauna,' p. 45, tab. i, figs. 14 — 10, under the name of " Eiulcndrmn.
pnsillimi." It is certainly, however, not a Eudendriicm, and is probably cither a Bouijainvillia or
a Perigonimus. The gonosome has not been observed, and though Sars gives a very full descrip-
tion of the trophosome, it is impossible, without a knowledge of the gonosome, to assign the
hydroid to its proper genus.

The following is Sars's diagnosis : —

" Eudetidrium pusilluiii. Sm-culis e tubulo repente filiformi ramoso lajvi erectis humilibus,
annulatis, hyalinis; ramis paucis (3 — 4), sparsis, brevibus, crassis, flexuosis seu tortis, annulatis,
simplicibus aut dichotomis, apice instar tubae aliquantum dilatato et a corpore seu capitulo
animalis distante ; capitulo aniuialium elongato, subcylindrico, non retractili, albo, tentaculis
circiter 20 uniserialibus."

The species was found in the neighbourhood of Messina, tolerably abundantly growing a
little below the surface of the sea on a brownish-red seaweed. It is quite a small form, attaining
a height of only an eighth of an inch. The hydranths manifestly belong to the BojyainviUia or
Pcrlffonimus type, and not .at all to that of Eudendriinn. Sars's description is accompanied by some
good figures of the trophosome.

DIPLUEA, Green, in Lit.

Name. — From cittAoc, double, and oupa, a tail ; in allusion to the two tentacles, which are
developed like a tail from one of the marginal bulbs of the planoblast.

CoRYNE, — Sleenstrup.
Steenstrupia, — Agassi: .

TROPHOSOME. — Htdkophtton solitary, rooted by a filiform hydeorhiza, and
surmounted by a claviform iitdeanth.'

GONOSOME. — Planoblasts springing from the body of the liydrantli at the
proximal side of the tentacles. Umbrella deep bell-shaped ; manubrium with simple
or quadrilobate mouth ; radiating canals terminating each in a marginal bulb, one of
which, in the mature medusa, carries a pair of filiform tentacles, the others being
destitute of tentacles.

^ For the description here given of the trophosome I have no data beyond the figure which
accompanies Steenstrup's memoir ; and as this figure is evidently a mere sketch, and plainly inexact
in its details, the above diagnosis founded on it is necessarily incomplete. Prof. Steenstrup's original
specimens have unfortunately been lost, and we do not possess data suflScient for the determination of
even the family to which Diplura ought to be referred. Its allocation to the BougainviUidce must,
therefore, be accepted with reservation.

320 DIPLURA FRITILLARIA.

One of the many observations which Steenstrup so liappily correlated in his famous treatise
on the ' Alternation of Generations' was made on a little tubularian hydroid which he obtained
off the coast of Iceland, and which, with its verticil of pendant medusa-bells, so closely resembled
a plant of Fritillaria imperialis, with its verticillate cluster of drooping flowers, that the Danish
zoologist gave to it the name of Coryne Fritillaria.

According to the views, however, by which we must now be guided in our classification and
nomenclature of the Hydroida, Steenstrup's hydroid is no Coryne, nor is it referable to any
other genus which had, up to the time when Steenstrup recorded it, been described.

Our knowledge of the trophosome is far from being as complete as could be desired. Indeed,
we are not informed whether the hydrophyton is naked or clothed with a chitinous perisarc, and
the disposition of the hydranth-tentacles can by no means be inferred with certainty from either
the figure or description given by Steenstrup.

The gonosome, on the contrary, is described and figured with considerable detail, so that no
difficulty is met with in obtaining such generic characters as may be aff'orded by this part of the
hydroid.

Some years ago Professor J. R. Green^ discovered in the Bay of Dublin a medusa which so
closely resembles that described by Steenstrup as developed from his Coryne Fritillaria that,
though nothing is known of the trophosome from which Green's medusa was produced, we are
justified in considering it for the present as congeneric with Steenstrup's hydroid.

Green described his medusa under the generic name of Biplonema ; but as this had been
already appropriated by the botanist, he subsequently substituted for it, in a letter to myself, the
name of Biplnra.

It was on these grounds that I had elsewhere' proposed that the name of Biplura Fritillaria
should take the place of Coryne Fritillaria.

'^^^ DiPLURA ruiTiLLAEiA, Steenstrup.

Coryne Fritillaria, — Steenstrup, Ueber den Generationswechsel, and Ray Society's Trans-
lation, p. 27, pi. i, figs. 41 — i6.
Steenstrupia Fritillaria, — Agassiz, Cont. Nat. Hist. U.S., vol. iv.
DiPLURA Fritillaria, — Allnan, in Ann. Nat. Hist, for May, 1864..

TROPHOSOME. — Htduocaulus attains a lieight of about half an incli, slender.
Hydranths with five or six tentacles.

GONOSOME. — GoNOPnouES pendulous, arranged in a verticil of four round the
base of the hydranth. Medusfe, at the time of their liberation, in tlie form of a
quadrangular bell, with a single rudimental, knob-like, marginal tentacle ; the two
tentacles given off from the marginal Inilb of the mature medusa very long and
moniliform.

> ' Ann. Nat. Hist.' for May, 1864.

PERIGONIMUS. 321

Development of Gonosome. — April and May.

Habitat. — Ou empty shells of Balani, &c., from the bottom of the sea.

Locality. — Off the town of Reikcwig, in Iceland, Professor Steenstrup.

I take for granted that Steenstrup is correct in regarding as the mature form of the plano-
blast of his Corj/ne FritUlaria a mcdus;i which he found swimming at large in the open sea, in
the neighbourhood of the spot which aflbrdcd him his specimens of the trophosome, and both
the generic and specific descriptions given above are based on this supposition. We cannot,
however, shut our eyes to the fact that in this identification we are dealing with a mere inference,
and that the free-swimming medusa has not been traced back by direct observation to the
tropliosome.

Steenstrup describes a lobcd body as existing at the common base of the two marginal
tentacles in his free medusa. He regards this as the generative organ of the medusa, a view
which, however it may have been justified at the time when he wrote, will not be shared in by
any one at present. There can, I think, be little doubt that the body in question is a cluster of
young medusa-buds originating in a situation in which this form of gemmation is by no means
rare among the Hydroiba.

PERIGONIMUS, Sars.

Name. — From -foJ, around, and yon/uoc, productive ; so named from the disposition of the
gonophores round the stem and Ijranches of the trophosome.

Atractylis (in part), — Stretliill IVriylit.

TROPHOSOME. — Htdeophyton consisting of a brandling or simple htbko-
CAULTJS rooted by a filiform nxmiOEmzA. Hydrantiis fusiform, with a conical
liypostome.

GONOSOME. — Planoblasts developed from the liydrophytou. Umbrella, at the
time of liberation, deep bell-shaped, with the oral extremity of the manubrium either
simple or more or less deeply lobed ; marginal tentacles either two or four, not in
clusters, and with bulbous bases, which are not furnislied with distinct ocelli.

The genus Perigonimus was founded by Sars' for a hydroid which he obtained on the
Norwegian coast. The position of its medusae, scattered over the hydrocaulus, appeared to him to
offer a feature so peculiar as to entitle this character to be assumed as one of the chief grounds
for the establishment of a new genus, while it further suggested to him the generic name.

1 'Fauna lit. Norv.,' Irste Liefer., p. 8.

323 PERIGONIMUS MUSCOIDES.

Tliis position of the gonophores, however, we now know to be equally a character of other
genera, while still other characters of a specific rather than a generic value were also included by
Savs in his diagnosis. Yet notwithstanding this, the hydroid on which the genus Perigonimus
was founded is undoubtedly entitled to generic rank, and, with some shght modifications of the
original diagnosis, Perigonimus must stand as an established genus.

The genus Atradi/lis of Wright, as already mentioned,^ includes forms which are referable
to the T criyonimus of Sars, and others which must l)e transferred to the BougainviUia of Lesson,
while, cjuite recently. Van Beneden" has instituted his genus Dinema for a form which cannot be
generically separated from Perigonimus.

The only character on which the claims of Binema to be separated from Perigonimus I'est
are found iu the fact that at the time of liberation the medusa of Dincma has only two marginal
tentacles, while that of the Perigonimus muscoides of Sars has four. This, however, is a very
unimportant difference, for there can be no doubt that the tentacles increase in number with the
age of the medusa, and that the two-tentacled Binema acquires four or more tentacles as it
advances towards maturity. I have accordingly already united^ under the genus Perigonimus,
not only forms in which the medusa, at the time of its liberation from the trophosome, has four
tentacles, but those in which the number of tentacles at that period does not exceed two. The
changes, indeed, which occur in the medusa as it approaches sexual matMity are to be chiefly
sought for in the increase of the number of marginal tentacles, each new one being intercalated
at the middle point between two older ones.

*^* 1. Peeigonbius muscoides, Sars.

Pekigonimus muscoides, — Sars, Fauna Lit. Norv., crste Lieferung, p. 8, tab. i, figs.

19—21.

TROPHOSOME. — Hybrocaulus attaining a height of from two to three inches,
much branched, the branches thinner than the main stem, and, as well as the main
stem, sending off numerous short, scattered, simple hydranth - bearing ramuli ;
PERiSAKC marked with longitudinal, somewhat undulating strias, not annulated.
Hydranths with a circlet of from eight to twelve tentacles in two closely approxi-
mated series.

GONOSOME. — GoNOPHOKES borne on short peduncles and scattered over the
main stem and principal branches. Medusa, on liberation, with four marginal
tentacles.

1 See p. 299.

' ' Recherches sur la Faune litt. de Belgique,' p. 127, pi. i.K and x.

' " Ou tlio Coustructiuti and Limitation of Genera," S:c., ' Ann. Nat. Hist.' for May, 1861.

PERIGONIMUS REPENS. 32a

Colour. — Pcrisarc grcyish-ycUow, hydranth vermilion, medusa with its manubrium reddish,
and the basal bulbs of its marginal tentacles reddish-ljrown.
Development of Gonosome. — August.

Habitat. — Attached to other hydroids, the tests of large ascidians, &c.
Batltymetrical distrihiition. — Coralline zone ?
Locality. — Coast of Xorway, Sars.

I have never met with this species. The diagnosis given above is selected from the
characters published by Sars. The species is historically interesting as affording the first-
described example of a genns which we now know to be one of the richest in specific forms
among the gymno1)lastic hydroids.

2. PERiGOxnius REPENS, SlretMll JFright.

Endendrium pusillum, — IVrif/ht, in Proc. Roy. Phys. Soc. Ediu., vol. i, p. 231, \A. xi,

figs. 8, 9.
Atr.-vctyles repexs, — IVriglit, iu Proc. Roy. Phys. Soc. Ediu., vol. i, p. 450, pi. xxii,

figs. 4, 5.
Perigonimus repex.s, — Hincks, Brit. Hydr. Zooph., p. 90, pi. xvi, fig. 2.

TROPHOSOME. — Htduocaultis simple, or slightly branched, attaining a height
of from one eighth to one quarter of an inch ; perisakc dilated at the summits of the
hydrocaulus. Htdrantiis partially retractile, with from four to twelve tentacles held
in an alternately elevated and depressed position.

GONOSOME. — GoNOPHORES nearly sessile on the hydrocaulus ; umbrella thimble-
shaped ; marginal tentacles four, two opposite tentacular bulbs carrying very long
ones, and the others very short ones.

Colour of hydranths white.

Habitat. — " Attached to scrtularians, and to the back and legs of the spider-crab," Wright.
Localities. — Firth of Forth, Dr. Wright ; Coast of Northumberland, Mr. Alder ; Ilfracombe,
IMr. Hincks.

I agree with Hincks in regarding the Atractylis repensoi Wright as the same hydroid which
he had previously described under the name of Endendrium pusillum. Indeed, it is impossible to
find any characters of specific value by which the one may be distinguished from the other.
Wright does not himself appear to recognise any difference between them, and while his change
of the generic name is founded on legitimate zoological grounds, his change of the specific name,

42

324 PERIGONIMUS MINUTUS.

is probably a mere ovcrsiglit. In order to avoid furtlicr confusion I liavc followed llincks in
adopting Wright's specific name of " repens.''

I have never met with this species, and the diagnosis here given is framed from Dr. Wright's
description.

3. Perigonimits minutus, Allman.

Plate XI, figs. 4— (5.

Perigonimus minutus, — Allman, in Anu. Nat. Hist, for January, 1863.

TROPHOSOME. — Htdeocaulus consisting of simple stems which rise at intervals
from a creeping stolon and attain a height of one eighth of an inch ; peuisakc smooth.
Htdeanths with seven or eight, rarely twelve, tentacles, which are held straight but
very irregularly in extension ; body of hydranth partially invested by a cup-like exten-
sion of the perisarc.

GONOSOME. — GoNOPHOiiES borne upon long peduncles, which spring at various
heights from the hydrocaulus. Medusa with the umbrella contracted towards the
summit, so as to give it a conical form ; two opposite marginal bulbs large, each
carrying a very extensile tentacle, two alternate bulbs much smaller and without
tentacles ; manubrium short, with the margin of the mouth fom--lobed.

Co/o?o-— Hydranth ash-brown, manubrium and marginal bulbs of medusa ash-brown,
perisarc yellowish-brown.

Development of Gonosome. — August.

Habitat. — Forming a fringe round the operculum of Tarriiella communis.

Bathymetrical distrihution. — Coralline zone.

Locality. — Busta Voe, Shetland.

The present species comes \'ery near to the Perii/onimus {Atracfylis) rejjcns of Wright, and,
indeed, is regarded by llincks as identical with it. It differs from it, however, in the irregular
disposition of the tentacles of the hydranth, and in the form of the medusa, which in Dr. Wright's
species shows no approach to the conical form of Periyonimus minutus, while it further differs
from it in the long peduncles which support the gonophores.

The tentacles of the medusa are very extensile ; when contracted they are, like many other
species of Perigonimus, rolled into a spiral.

Perigonimus minutus was abundant in the only locality where it has been as yet obtained,
and where it was entirely confined to the operculum of living Turritellas. Out of nearly thirty
specimens of Tmritella communis which I had dredged in Busta Voe and examined, not one was
free from this remarkable little hydroid.

PERIGONIMUS PALLIATUS. 325

4. Perigonimus sessilis, Strethill Wright.

Endexdru.m sessile, — Wrhjhl, in Proc. Roy. Phys. Sue. Ediii., vol. i, p. 237, pi. xii'

figs. IG, 17.
Atractylis sessilis, — llri/jhl, id., p. 450.
PERiGo.\nius SESSILIS, — Al/maii, in Ann. Nat. Hist, for May, 18(!1.. Hhicks, Brit. Hydr.

Zooph., p. 93, pi. xvii, fig. 1.

TROPHOSOME. — Hydrocaulus obsolete or rudimental. Hydranths sessile, on
a creeping reticulated stolon, or slightly raised above it on the rudimental hydrocaulus»
invested as far as the tentacles with a cup-like extension of the perisarc ; tentacles
attaining the number of eight, equal in length, straight in extension, and then carried
alternately elevated and depressed.

GONOSOME. — GoNOPiioRES sessile on the creeping stolon. Medusa with the
umbrella thimble-shaped, and having four marginal tentacles, alternately very long
and very short.

Ilahitat. — Oil shells from deep water, and on rocks.
Localifif. — Firth of l-'ortli, Dr. Wright.

I have never met with this apparently well-marked species. The diagnosis here given has
been extracted from Dr. Wright's description. Though the trophosome would seem to be very
distinct from that of Periyonimus piisillus, the medusse are described by Dr. Wright as differing
in no respect from those of that species, cither in size or form.

5. Perigonimus pailiatus, Strethill Wright.

Atractylis palliata, — Jl'riffht, in Ann. Nat. Hist, for August, 1861, pi. iv, figs. 6, 7.
Parigommus PALLIATUS, — AUiHan, in Ann. Nat. Hist, for May, \?>M. Hincks, Brit. Hydr.

Zooph., p. 93, pi. xvii, p. 91.

TROPHOSOME. — Hydeoc.vulus but slightly developed, and consisting of short,
closely set stems, which spring from a creeping reticulated stolon. Hydeanths very
minute, enveloped in a thick gelatinous coat as far as the border of the mouth,
tentacles eiu-ht, alternate.

32G PERIGONIMUS VESTITUS.

GONOSOME, — GoNOPiiORES developed from tlie hydrorliiza ; medusa thimble-
shaped, with two marginal bulbs which carry long tentacles, and alternate with two
smaller bulbs destitute of tentacles ; manubrium with the oral margin four-lobed.

Colour of hydranths white.

Habitat. — On a shell inhaljited by a hermit-crab.

Locality.— Y\x\\\ of Forth, Dr. Wright.

I have never met with this remarkable little hydroid. " When first observed," says Dr.
Wright, " its closely set and dense white polypes, surrounded Ijy their gelatinous envelopes, were
mistaken for a mass of minute ova."

G. Perigonimus VESTITUS, Allmaii.

Plate XI, figs. 1—3.

Perigonimus vestitus, — Allman, in Ann. Nat. Hist, for July, 18G4. Hindis, Biit. Hydr.

Zoopli., p. 94-.

TROPHOSOME. — Hydkoc.a.ulus consisting of numerous stems rising at short
intervals from a reticulated stolon, and attaining a height of from half a line to two
lines, becoming greatly dilated towards the summits where they pass uninterruptedly
into the body of the hydranths, mostly simple, but occasionally with one or two
short lateral branches ; peuisakc coarse, roughened by adherent particles of sand.
Hydranths with from six to ten tentacles, which are rendered hispid by minute
clusters of thread-cells, and are lield straight in extension with the alternate cues
elevated and depressed ; posterior part of the body invested by the rough perisarc,
which is thence continued as a delicate, smooth membrane over tlie remainder of the
body nearly as far as the mouth.

GONOSOME. — GoNOPHORES elevated on long peduncles, which spring from the
hydrocaulus, and occasionally also from the hydrorhiza, the peduncle for about its
proximal half covered by a continuation of the chitinous perisarc. Medusa, at the
time of liberation, oviform, the cavity of the umbrella being very dee]) and the
codonostome much contracted; umbrella-waUs very thin, and with numerous scattered
thread-cells immersed in them, two opposite marginal tentacles, and two intermediate
marginal bulbs destitute of tentacles, manubrium with four shallow lii:is.

PERIGONIMUS SERPENS. 327

Colour. — Hydrantlis pale j-ellow, luaimbriuiu and marginal Indbs of medusa pale rcddisli,
perisarc yellowish brown.

Development of </o)iosome. — .1 une.

Ilahitai. — On an old Buccerium shell found in a rock-pool near low-water mark.

liathjmetrical distribution. — Lammarian zone?

Zoca%.— Eirth of Eorth, G. J. A.

Perii/onimus vestitus was met with in the Eirth of Eorth in June, growing on an empty
Buccerium shell, where it was associated with IL/clractiuia ecldnaia. In the continuation of the
perisarc over the body of the hydranth, as well as in general habit, it comes very near to the
Periffoinmus j^cHiafus of Wright, from which, however, it difi'ers, judging from Dr. Wright's
description and figures, in its more developed liydrocaulus, in the thinner and more membranous
character of the perisarcal investment of the hydranths, in the position of the gonophores — which
are here borne almost exclusively on the liydrocaulus, only an occasional one being here and
there developed from the hydrorhiza, while in Veriyonimus palliatus they are described as being
confined to the hydrorhiza — and in the form of the medusa whose contracted codonostome gives
to the umbrella at the time of liberation an oviform shape, while in Peric/onimus palliatus the
umbrella does not become contracted towards the codonostome, and is accordingly nearly cylin-
drical in form.

In medusas, about the tenth day after their liberation, the form had undergone consideraljle
change, the umbrella having become nearly spherical. No increase, however, had taken place in
the number of marginal tentacles.

The trophosome is rendered particularly striking by the great dilatation of the stems, which
graduate into the body of the hydranths without any well-marked line of demarcation. In the
medusa the tentacles, which are very extensive, become rolled into an elegant spiral when
contracted.

7. Pemgonimus serpens, Allman.

Plate XI, figs. 7—9.

Perigo.nimus serpens, — Allman, in Ann. Nat. Hist, for January, 1803. Hincks, Brit. Hydr.

Zoopb., p. 93, pi. xvi, fig. 3.

TROPHOSOME. — Htduocaulus consisting of short, simple stems, rising at sliort
intervals from a creeping reticulated stolon, and attaining a height of about two lines,
clothed with a very delicate transparent perisauc, which is destitute of annulatiou, and
loses itself at the base of the hydranths without forming distinct cup-like dilatations.
Hydeanths, with twelve or fourteen slightly alternating tentacles.

328 PERIGONIMUS LINEARIS.

GONOSOME. — GoNOPiioRES elevated on long peduncles ■wliicli are borne by the
creeping stolon. Medusa dome-shaped, with the vertical and transverse diameters
nearly equal ; manubrium reaching to about half the depth of the bell ; marginal
tentacles two, opposite, very extensile, with large basal bulbs ; two very small inter-
mediate marginal bulbs.

fb/o/'/-.— Ilydranths and cojiiosarc, as well as the manubriuiu and marginal bulbs of the
medusa reddish orange.

Development of gonosome. — Autumn.
Habitat. — Growing on the stems of other hydroids .
Bathjmetrical Distribution. — Coralline zone.
LocaJltij. — Torliay, G. J. A.

This beautiful little hydroid was dredged from about twelve fathoms in Torbay, where it
invested the stems of Pliimularia sefacea. Though small, it is rendered very conspicuous by the
bright orange colour, not only of the hydrantlis, but of the ccenosarc, which is quite visible
through the delicate transparent perisarc. When about to detach itself from the trophosome the
medusa Ijursts through its uiembranous ectotheca, the remains of which may be usually seen
after its rupture still attached to the top of the peduncle. At the same tiuie, the two tentacles
which had been previously folded back into the cavity of the umbrella become liberated from their
confinement, and extend themselves in the surrounding water. The medusa, now divested of its
ectotheca, still remains for some little time attached to the summit of the peduncle, until at
length, after repeated convulsive efforts, it breaks loose from the trophosome, and henceforth
leads the life of a free zooid in the open sea.

When the planolilast is at rest, or is floating listlessly in the water, the tentacles are usually
extended to a great length, and hang in the form of threads of extreme tenuity from the umbrella
margin ; but, on the slightest irritation, each tentacle is instantly drawn up assuming a beautiful
spiral in the act of contraction. The tentacles are also rolled into a spiral when the medusa, by
the contractions of its umbrella, is propelling itself through the surrounding water.

8. Peeigonimus lineaeis, Alder.

Atkactylis linearis, — Aider, Suppl. C'ntal., p. 6, pi. x, figs. 1, 2, 3.

Perigonimus linearis, — Ailinan, in Ann. Nat. Hist, for May, 1864'. Hincks, Brit. Hydr.

Zooph., p. 96, pi, xvii, fig. 3.

TROPHOSOME.^ — Htdeocaulus consisting of unbranched (?) stems, which rise
at short intervals to the height of a quarter of an inch from a reticulated stolon ;
perisarc destitute of annulation, but slightly wrinkled near the base of the stems.

DOUBTFUL SPECIES. 329

Hydraxths slender, retractile, witli eight long muricated tentacles, held alternately,
elevated and depressed.

GONOSOME. — GoNOPnORES pear-shaped, or sub-globular, borne by the liydro-
caulus in groups of two or three together; medusa "globose, slightly truncated below,
with a contracted aperture ; four moderately sized sub-clavate tentacles arising from
four semicircular yellowish lobes at the margin of the umbrella ; sub-umbrella small,
with four radiating canals, the centre occupied by a mass of yellowish or orange
granules, apparently ova; ])eduncle (manubrium) inconspicuous, branched at the
base." — Alder.

Ilahitnl. — On TvrriteUa communis and other shells.
Bal/ij/metrical disirihufio/i. — Deep-water zone.
Loccdiiy. — Coast of Northumberland, Mr. Alder.

It is with great hesitation that I refer tins species to the genus Pcriyonhnus. I have never
seen it, and the characters given above are those assigned to it by Mr. Alder. So far as the
trophosomc is concerned, there is nothing to exclude it from Perifjonimus, but if the medusa bo
correctly described, there is some reason for regarding it as belonging to a distinct generic
group.

The characters of the medusa, howevei', are not given with sufficient detail and precision to
afford material for the definition of a new genus. Indeed, it seems pretty evident that the
medusae of the specimen which furnished the subject of Mr. Alder's description and drawing were
not in a perfect state, that they had undergone more or less change from exposure to unfavorable
conditions, and that their essential characters had been thereby obscured. We must wait,
therefore, until further observations shall have made us better acquainted with its gonosome,
before we can venture upon a precise characterization of this curious little hydroid. Until then
we may retain it in the genus Periyonimus, to which it appears to come nearer than to any other
genus hitherto defined.

DOUBTFUL SPECIES.

The " Atr.\ctylis bitentaoulata" of Wright (' Proc. Roy. Phys. See./ Edin. for Feb., 1S63,
p. 43, pi. i, fig. 5), and the " Atractylis quadritentaculata" of Wright [id. fig. C) are probably
immature forms of some species of Perigonimus. In the absence, however, of all knowledge of
the gonosomes it is impossible to assign them with certainty to any established genus.

They are very small forms. The " AtradijUs bitentaculatd' is described as consisting of
minute, nearly sessile hydranths, springing from a retiform hydrorhiza, and furnished each with
two erect tentacles. The hydrantlis " have a habit somewhat like that of Lar, of quickly

330 EUDENDRIDtE.

bending down tlic body until the mouth is brought close to the surface on which the zoophyte
grows." It was found on a pecten shell dredged in the Firth of Forth.

The other form, " AtradijUs quadritenfaculafa,'" is described as having its hydranths sessile,
columnar, short, provided with four alternate tentacles, two of which are long and depressed, two
short and nearly at right angles to the body of the hydranth. It " was found creeping along the
the side of a large vessel containing shells and zoophytes dredged from the Firth of Forth."

EUDENDRID^.

TROPHOSOME. — IItdrocaulus developed, invested with a peeisaec. Hydranths
with the hypostome abruptly differentiated from the body, and with a single set of
verticillate fiUform tentacles.

GONOSOME. — GONOPHORES fixed SPOEOSACS.

The hypostome, of all the species included in the family of the Eudendrida, is rendered very
remarkable by being abruptly differentiated from the body of the hydranth instead of passing
continuously into it as in other gymnoblastic hydroids. In tliis respect it resembles the hypo-
stome of the CampannhmdcE among the Calyptohlastea.

ETJDENDRIUM, Ehrenherg (in part).

Name. — From ju, Avell (beautiful) and Uv^pov, a tree, so named from the tree-like form of the
trophosome.

TuBULARiA, — Linnaus.
Sertularia, — Cavolini.

TROPHOSOME. — HTDEOPnYTON consisting of a branching iitdeocattlus rooted
by a creeping filiform HYDEOEnizA. Htdeanths flask-shaped, or oval, with the
hypostome expanded at its distal extremity so as to be more or less trumpet-shaped ;
tentacles forming a verticil just below the hypostome.

EUDENDRIUM. 331

GONOSOME. — Spoeosacs developed from the body of the liydrantli at the
proximal side of the tentacles, or from the hydrocaidus. Male sporosacs polytha-
lamic, female sporosacs monothalamic.

Ehrcnbcrg/ perceiving that the Tabnlaria of the eailier autliors inchuled more than one
generic form, proposed to distribute tlieni inider two genera. The dismemberment to which
the old genus Tubiilaria was thus suljjectcd must be regarded as an important step in the classi-
fication of the Hydkoida ; but Ehrenberg scarcely recognised the true grounds of the division,
for he based it on the simple or branched condition of the trophosome, leaving the simple forms
in the genus Tithularia, and bringing the ramified forms into a new genus, to which he assigned
the name of Eudendriiim.

The characters thus selected are of minor value, and the genus Eudendrmm, as defined Ijy
Ehrenberg, included not only forms which are properly separated from Ttibularia, but also true
Tulidnria, such as Tuhularia larijnx. The celebrated Prussian microscopist, however, did not
allow the more important characters, such as those derived from the disposition of the tentacles,
entirely to escape him, though he merely pointed to them as possibly affording to future observers
grounds for further generic distinction.

Van Beneden^ seems to have been the first who felt the necessity of restricting the genus
Uudendrium to such forms as had their tentacles arranged in a single verticil, and of restoring
to the genus Tuhularia those in which the hydranth is provided with two verticils of tentacles,
regardless of the simple or branched condition of the trophosome.

Still later, Sars^ founded the genus Ferigonimus for certain species which continued to l)e
included in the restricted genus Eudendrium, assuming as his chief grounds of separation the
position of the gonophores, which he believed to be confined to the hydranths in the true
Eudendriums, but distriliuted along the stems and branches in Perif/onlmus. This character,
however, is of comparatively little value, a much more valid one being found in the form of
the hydranth, as was first distinctly shown by Dr. Strethill Wright. (See general remarks on
the genus Wriyldia, given above.)

The gemis Eudendrium must now be understood as limited by the characters assigned to it
in the diagnosis just given. All the species hitherto discovered are ramified, and there are no
hydroids on which the plantdilce {)hysiognomy is more decidedly impressed than we find it to be
on the various species of this beautifid genus.

The peculiar form of trophosome described above in the generic diagnosis has as yet becu
found associated with only one form of gonosome. Until, therefore, some new observations sluall
come to invahdate our generalization, we shall be justified in referring any trophosome with the
characters above enumerated to the genus Eudendrium, whether the gonosome be kuoivn or not.

This is in strikhig contrast with certain other hydroid trophosomes, such as the coryuoid
trophosome, which may belong to at least three genera, and that of Periijoniriius, which, as we
have akeady seen, may also have such different forms of gonosome associated with it as to render
it impossible, without this last element, to decide on the genus of the hydroid.

1 " Coralleuthiere," ' ,Vbliaudl. der Berlin. Alcad.,' 18.33, p. 297.
" ' Recherches sur I'Erabrvogenie des Tuliulaires.'
^ ' Fauna lit. Norveg.,' Erste Liefer, p. 8.

43

332 EUDENDRIUIVI RAMOSUM.

1. EUDENDBIUM RAMOSUM, LlmmilS.
Plate XIII.

CORALLINA TUBULARIA GRACILIS ET RAMOSA, AXILLIS RAMULORUM CONTORTIS. " Small ramified

tubular Coralline." — Ellis, Corall., p. 31, pi. xvi, fig. a ; xvii, fig. a a.
TuBULARiA iiAMOsA, — Linn., Syst. Nat., edit. x.
TuBULARiA TRicHoiDES, — Pullas, Eleuclius, p. 84.
EuDENDRiuM RAMOSUM, — Ehreuberff, Corallentbiere, Abhandl. Berl. Akad., p. 29G. Johnston,

Brit. Zooph., 2nd edit., p. 4G, pi. vi, figs. 1 — 3.

Hincks, Brit. Hydr. Zooph., p. 82, pi. xiii.

TROPHOSOME. — Hydrocaulus much branched, fascicled at its base, and attain-
ing a height of four inches or more ; primary ramification irregular, after which the
branches become regularly alternate and mostly distichous in their arrangement,
giving oiF all along their length, from their upper or distal sides, short, usually simple
ramuli, which support the hydranths on their summits ; peuisakc firm, annulated at
the origin of the branches, or even along the entire length of the smaller branches.
Hybkanths usually with about twenty tentacles, frequently atrophied in the male.

GONOSOME. — Sporosacs, in the male, composed of two spherical receptacles,
and springing from the body of the hydranth in a verticil behind the tentacles, \yhich,
however, often become atrophied and disappear. Sporosacs in the female, piriform,
scattered, springing, some from the body of the hydranth at the proximal side of the
tentacles, and some from the hydrophyton just below the hydranth.

Colour. — Hydraiitlis vermilion, perisarc dark reddish brown, in the older parts of the
colony often nearly black ; male gonophores with vermilion spadix modified to rose colour as
seen through the pale imperfectly transparent contents of the receptacles ; female gonophores
orange red.

Develoj)me?it of (jotiosome. — April.

Habitat. — Attached to oyster-shells, stones, &c., in the sea.

Bathymetrical distribution. — Coralline zone.

Localities. — On the Kentish shore, Ellis ; JMorecamb Bay, Lancashire, G. .T. A. ; south
coast of Devonshire, Rev. T. Hincks.

The beautiful hydroid just defined may be conveniently assumed as the type of the genus
Eudendrium . There can be no doubt that it is identical with the " small ramified tubular coral-
line " of Ellis, and that it is, therefore, the original Tubularia raiiiosa, this being the binary

EUDENDRIUM RAMOSUM. 333

designation by wliicli the descriptive phrase of Ellis was superseded in the ' Systema Naturse.'
It is consequently a very diiferent hydroid from the Etidcndrium ramomm of Van Bencden.

It was first described and figured by Ellis from specimens obtained upon the coast of Kent,
but without any reference to the gonosome, an omission which may be explained by the fact tliat
Ellis's specimens had been gathered in August, and therefore after tlie gonophores had dis-
appeared.

The hydrauths, which are of a fine vermilion colour, are very conspicuous even to the naked
eye, while the beauty of the male colonics is greatly enhanced by the presence of the gonophores.
These are two-chambered and spring in a regular verticil from the body of the hydranth, about
half way between the origin of the tentacles and the summit of the supporting branch.

The hydranth in the male colonies is at first, while it continues to retain its full development,
embraced by its circlet of gonophores from whose midst the tentacles rise in a graceful cani-
panulate plume. Frequently, however, the hydranth, as if starved by the growing gonophores,
and their increasing demand for nutrition, becomes atrophied, the tentacles entirely disap[)ear,
and the whole becomes changed into a short column, which carries the gonosacs in an umbel on
its summit. With the atrophy of the tentacles the mouth seems also to disappear, and the
hydranth has thus assumed all the characters of a blastostyle. It is a blastostyle, however,
merely from atrophy, and must not be confounded with a true blastostyle, which, though strictly
homologous with a hydranth, belongs, from its origin, to a special and independent morphological
modification.

In the female colonies the gonophores are without the synunetrical arrangement so striking
in the male, and, instead of being disposed in a regular verticil, are scattered on the body of the
hydranth, and on the distal end of the branch. They are of a reddish-orange colour, and
increase in maturity as we trace them downwards on the branch. The tentacles of the hydranth
are here, also, as in the male, often dwindled ; but I have never met with that complete atrophy
which is of such frequent occurrence in the male.

The perisarc constitutes a firm, elastic, horn-like covering. The hydrocaulus in full-sized
specimens is about half a line in thickness towards its base, and here presents a fascicled structure
formed by the mutual adhesion of several simple capillary stems. It soon, however, loses this
condition, and then continues as a singled branched tube for the rest of its course. The compa-
ratively small diameter of the stem, even at that part where the fasciculated condition exists, and
the fact that the fasciculation is confined to those parts which are near to the attached end of the
hydrosoma, has caused this character to be overlooked in the description given by Ellis, while
Pallas ascribes to the species an absence of fasciculation as one of its leading diagnostic charac-
ters. This error has doubtless been strengthened by the fact that many of the specimens wdiich
have been found entangled in the lines of the fishermen, or otherwise detached from their place
of growth, have been so torn away as to leave most or the whole of the fascicled portion behind
them.

334 EUDENDRIUM RAMEUM.

2. Ettdendbium RAMEUM, Pullas.

TuBULARiA RAMEA, — PdUus, Eleuduis, p. 83. Ball/el/, Rare and Remarkable Animals, vol. i,

p. 5, pis. vi, vii, viii, ix.
EuDENDRiu.M RAMEUM, — Johnstoti, Brit. Zooph., 1847, p. 45, pi. v, figs. 1, 2. Hincks, Brit.

Hydr. Zoopli., p. 80, woodcut fig. 8.

TROPHOSOME. — Htdrocaulus profusely branched, attaining a height of from
three to six inches, fascicled in the main stem and principal branches ; main stem
attaining a thickness of more than a quarter of an inch, and as well as the principal
branches very irregularly ramified ; branches ultimately losing their fasciculation, and
then consisting of single capillary tubes, wliich may continue to branch before the
emission of the ultimate or hydranth-bearing ramuli, which are regularly alternate in
their disposition ; perisarc rigid, occasionally marked with nearly obsolete annulations
on the smaller branches. Hydranths with about twenty tentacles, frequently
atrophied in the male after the production of gonophores.

GONOSOME. — Male sporosacs two-chambered, borne upon the body of the
hydranth in a verticil immediately below the tentacles ; female sporosacs oval,
scattered on the hydrocaulus for some distance below the hydranth.

Colour. — Hydranths rose colour, hydrocaulus reddish brown, female gonophores brownish
orange.

Hahitat. — Attached to stones, old shells, &c., in the sea.

Bed hi/metrical dislnbution . — Deep-sea zone.

Localities. — Mediterranean, Pallas; coast of Norway, Sars; Firth of Forth, Sir J. G.
Dalyell and G. J. A. ; coast of Northumberland, Dr. Johnston and ]\Ir. Alder ; Shetland Islands,
Firth of Clyde, coasts of Lancashire and Cornwall, and east coast of Ireland, Mr. Hincks.

EudendriiiM ramcum, with its massive trunk and boughs, sending off smaller and smaller
branches, until these, in the wonderful profusion of their ramification, terminate in the delicate
hydranth-bearing twigs, is perhaps the most tree-like of all our hydroids, and might well have
suggested the generic name under which it is associated with other tree-like forms of these
beaxitiful animals.

It is certainly very nearly allied to the Eudendrium ramosum of Linna;us, more nearly,
indeed, than has been suspected by the systematic writers who have described it, and who have
followed Pallas in regarding its fascicled stems as affording the main grounds of distinction
between it and the species just named. Eudendrium ramosum, however, is also a fascicled form,
and the chief difference between the two species will be found in the nnich more massive stems
and main branches of Eudendrium. rameum, and in its more irregular ramification and rigid

EUDENDRIUxM CAPILLARE. 335

habit. The rigidity of the wliole hyih-ophyton is such that, on removal from tlie water, all the
branches, even to the finer twigs, retain their form and directions as completely as if the hydroid
were still immersed.

The first description of the present species is that given by Pallas, whose specimens were
obtained from the Mediterranean. It is a frequent hydroid on the Scottish coast, wliere it is
often brought up upon the lines of the fishermen from deep water.

Though I have many times seen specimens of this fine species, I have unfortunately not
had an opportunity of making a drawing from the living hydroid, which has, therefore, no place
among the plates of the present volume. The figures of Johnston, Dalyell, and Hincks, liowever,
referred to in the synonyms, will afford to the zoologist good aid in his determination of the
species.

3. EUDENDRIUM CAPILLARE, Alder, Sp.

Plate XIV, figs. 1—3.

EuDENDRiUM c.iPiLLARE, — Aider, Catal., p. 15, pi. i, figs. 9 — 12. Hincks, Brit. Hydr. Zoopli,,

p. 84, pi. xiv, fig. 2.
CoRVMBOGONiuM CAPILLARE, — Allman, in Ann. Nat. Hist, for August, 1861.
DicoRYNE CAi'iLLARis, — Aider, Suppl. Catal., p. 6.

TROPHOSOME. — Eydrocaulus forming stnall bush-like growths, much and
irregularly branched, rising to a height of from one half to three quarters of an mch,
and given off at intervals from a creeping anastomosing stolon ; stems and branches
capillary, invested with a thin pekisarc, which is annulated at the origin of the
branches. Uydranths with about twenty-four or twenty-six tentacles.

GONOSOME. — Male sporosacs two-chambered, supported in radiating, some-
what corjTubose clusters upon the bodies of hydranths from which the tentacles and
hypostouie have disappeared by atrophy, the stalks which carry them constituting
either uUimate ramuli of the hydrocaulus or simple stems, which spring du-ectly
from the hydrorhiza. Female sporosacs piriform, in clusters similar to those of
the male, and borne in the same way on the remains of atrophied hydranths whosf
supporting stalks are carried cither by the hydrocaulus or directly by the hydrorhiza.

Colour. — Hydranths and male gonophores a pale greyish olive, female gonophores reddish
orange ; perisarc light brown, becoming darker in the older parts.

Development of Gonosome. — June to September.

Habitat. — On rocks near extreme low-water mark, and on other hydroids, old shells, and
the tunics of Ascidians.

Bath y metrical distribution. — Laminarian and Coralline zones.

336 EUDENDRIUM ARBUSCULA.

Localities. — Coast of Nortluimberland, Mr. Embletoii ; Plymouth, ilr. Alder; Toihay,
Devonshire, Rev. T. Hincks; Firth of Forth and Torbay, G. J. A.

The first description of Eudendriiim capiUare is that given by Mr. Alder, who drew up its
characters from a specimen preserved in spirits. Some time after this I dredged the same
hydroid in the Firth of Forth from a depth of about five fathoms. The complete atrophy of the
gonophore-bearing hydranths led me to regai'd it as the representative of a new genus, which I
accordingly described under the name of Cori/mbot/onitan. A more thorough investigation of
the genus Eudendriiim, however, has since convinced me that this genus may present in the
hydranths which carry the gonophores, even in one and the same colony, every degree of atrophy,
from the fully developed hydrantli to the mere remnant of this body after mouth, hypostome,
and tentacles have entirely disappeared. The blastostyle-like condition of the hydranth, therefore,
in the genus Eudendriiim is a character of little importance, and can in no way justify a generic
separation. The genus Coryinboyoniuni must accordingly be absorbed into the older genus
Eudendriiim, in which Alder had originally placed it.

The Eudendrium capiUare is a small but elegant hydroid. It is abundant at extreme low-
water spring tides in some parts of the Firth of Forth, where it may be found attached to the
outer coarse tunic of some of the common Ascidians. Its distribution in Britain is a wide one,
for it is also abundant on the Devonshire coast, where I have dredged it from a depth of about
four fathoms, growing on ] Ij/drall mania falcata and other common hydroids.

4. EUDENDEITJM ARBUSCULA, Wright.

Eudendrium arbuscula, — SlretliiU IVriyht, in Edin. New Phil. Journ. for July, 1859, pi. ix,

figs. 5, 6. Hincks, Brit. Hydr. Zoopb., p. 84,
pi. xiv, fig. 1.

TROPHOSOME. — Hydkocaulus "forming a bushy tree of adnate stems ; branches
ringed near their insertions." Htdkanths "white, terminal, on very slender and
transparent branches, and with numerous alternate tentacles ; base of body surrounded
by a ring of large thread-cells."

* GONOSOME. — Male spouosacs " borne in clusters on short stems springing at
right angles from the branches. Summit of the double capsule with a tubercle con-
taining barbed thread-cells." — Strethill Wright.

I have never met with this species. It is described by Dr. Strethill Wright from a single
specimen, which he states to have been about two inches in height, and to have been obtained
at Queen's Ferry, Firth of Forth, in the month of September.

EUDENDRIUM INSIGNE. 337

5. EUDENBRIUM INSIGNE, Hilicks.

Plate XIV, figs. 4—6.'

EuDENDRiuJi HUMiLE, — AUmaii, in Ann. Nat. Hist, for January, 1863.

EuDENDRiuM INSIGNE, — Hiiicks, in Ann. Nat. Hist, for August, 1861 ; Brit. Ilj-dr. Zooph.,

p. 80, pi. xiv, fig. 3.

TROPHOSOME. — Hydrocaulus much ami irregularly branched, rising to a
height of about three quarters of an inch ; perisauc distinctly annulated throughout.
Htdranths with a shallow circidar groove near their base. Tentacles twenty or
twenty-three, with the alternate ones usually elevated and depressed in extension.

GONOSOME. — Male sporosacs two-chambered, forming a verticil round the
body of the hydranths, and springing each by a short stalk from the circular groove
which passes round the hydranth near its base. Female sporosacs piriform, borne
both by the body of the hydranth and by the hydrocaulus immediately below it.

Colour. — Hydranths \ellowish vermilion. Male gouopliores piuk; female gonophores
reddish orange. Perisarc reddish brown.

Development of Gonosome. — September.

Habitat. — Rock pools near low-water spring tides.

Batht/metrical distribution. — Laminarian zone.

Locality. — Torbay, Mr. Hincks and G. J. A. ; Ilfracombc and Swanage, ]\Ir. Ilincks.

Some years ago I described as a distinct species, under the name of Eudcndrium Iiumile, a
little hydroid from Torquay. I was aware at the time that Mr. Hincks had already described,
under the name Eudcndrium insiyne, a species of Eudendrium which, judging from his description,
resembled in some points my E. hiimile, though in other more important ones it differed from it
so widely as to forbid the association of the two in a single species.

Subsequent observations have induced Mr. Hincks (' Brit. Hydi'. Zooph.,' p. S5) to correct
his original description, and in its present form it will easily apply to my Eudendrium himUe. I
am therefore willing to accept Mr. Hincks' view, that the two species are identical, and to
suppress the name of Jiumile in favour of insiyne.

I met with this pretty little Eudendrium on the rocky shore of Torba\', where it occurred
rooted to the bottom of the clear pools near the limit of the lowest tides. ^Mr. Hincks found it
in the same place, and in some other parts of the southern shores of England.

' Tlie species is there figured under the name of Eudendrium liumile.

33S EUDENDRIUM ANNULATUM.

*^* 6. EuDENDEiUM DISPAE, Agassk.

EuDENDRiUM iiisPAR, — Affossiz, CoTitr. Nat. Hist. U. S., vol. iv, p. 285, pi. xxvii. A/e.v.
Ai/assiz, Illustr. Catal. of North American Acalophre, p. 159.

TROPHOSOME. — Htdrocaulus irregularly branched, slender and flexile, attain-
ing a lieiglit of about two inclies ; perisakc ringed throughout. Hydrantii with
twenty-eight tentacles.

GONOSOME. — Male sporosacs two- or three -chambered, springing from the
body of the hydranth in an irregular verticil just below the tentacles. Eemale sporo-
SAcs globular, scattered over the body of the hydranth and the distal portion of the
stem.

Development of Gonosome. — ]\Iay to September.
Bathymetrical distribution. — Deep-sea zone.
Locality. — Shores of Massachusetts, Agassiz.

The diagnosis here given is derived from Agassiz's description of the species. In this
description, however, the author seems chiefly to have had in view its distinctive points when
compared with Boiigainvillia, so tliat generic rather than specific characters are for the most part
referred to, and it is by no means easy to find characters available for a technical specific diagnosis.
Those selected above, however, will, I think, be sufficient to secure the American Eudendriiim
from being confounded with any of the other species of the present Monograph.

The same species is referred to by Alexander Agassiz in his ' Illustrated Catalogue of North
American Acalepha;,' where he tells us that " the male and female conniuuiities are readily recog-
nised by the different colour of the medusa-buds (sporosacs), the male medusa-buds being of a
bright orange, while the female are of a dull pink."

No mention is made of the atrophy of the gonophore-beaiing hydranths, so common in
other species.

7. EUDENDRIUM ANNULATUM, Normail.

EuDENDRiuM ANNULATUM, — Normaii, iu Ann. Nat. Hist, for January, 1864, pi. ix, figs. 1 — 3.

Hincks, Brit. Ilydr. Zoopli., p. 83, pi. xv, fig. 1.

TROPHOSOME. — Hydrocaulus shrubby, attaining a height of about four inches,

EUDENDRIUM VAGINATUM. 339

main stems fascicled, very thick, smaller branches extremely numerous, composed of a
single tube, which is closely and re2;ularly ringed throughout. Htdeanths with
sixteen to twenty tentacles.

GONOSOME. — Female gonophores in spike-like clusters borne round the body
and stalk of atrophied hydrantlis. Male gonophores not yet observed.

Habitat. — In rock pools and caves at low water.
Bathijmetrical dldriljiition. — Laniinarian zone.
Locality. — Burrafortli, Shetland, Rev. A. M. Norman.

I have never met with this fine species. It was discovered by the Rev. Alfred IMerle Norman at
Burraforth, in North Unst, the most northern of the Shetland Islands, where it was found attached
to the perpendicular sides of a cavern about a foot beneath the water at the lowest spring tides,
as well as in some adjacent rock pools. Its dense shrubby habit, deeply annulated branches,
and the spike-like groups of gonophores on atrophied hydrantlis in the female colonies, form
a combination of characters which distinguish it from all other described species.

The outermost portion of the fascicled stem appears to consist of a layer of remarkably
contorted tubes, which are figured by Mr. Norman, and referred to by him in his description of
the species.

The male gonophores have not been observed.

8. EuDENDEiuM VAGINATUM, Allman.
Plate XIV, figs. 7 and S.

EuDENDRiuM VAGINATUM, — Allman, in Ann. Nat. Hist, for January, 18G3. Hincks, Brit.

Hydr. Zooph., p. 86.

TROPHOSOME.— Htdbocaultjs much and irregularly branched, rising to about
one inch and a quarter in height ; peuisarc rigid, deeply and regularly annulated
throughout. Htdkanths with about eighteen tentacles, and having the body as far
as the origin of the tentacles enveloped in a loose corrugated membranous sheath,
which loses itself below on the hydrocaulus.

GONOSOME. — Not yet observed.

Colour. — Hydrantlis vermilion, perisarc deep reddish brown.
Habitat. — In rock pools at extreme low-water spring tides.

44

340 EUDENDRIUM TENUE.

Balliymefrical dislrihution. — Liuiiinanaii zone.
LocalUi). — Shetland Islands, G. J. A.

This beautiful little Eudendrium is conspicuous by its large, bright-vermilion Iwdranths. It
occurred in considerable abundance on the " Out Skerries," and other exposed rocky islets of
Shetland, but was found only at the extreme edge of low spring tides, where it grew rooted to
the bottom of shallow rock pools.

It was in the month of August that I met with it, but its season of greatest perfection
must have been then passed, for in none of the specimens could any trace of the gonosome be
found, while in many of them the hydranths themselves had disappeared from the summits of
the ranuili.

*at* 9. Eudendrium tenue, Alex. Agassiz.

Eudendrium tenue, — Aler. Agassiz, Illus. Catal. N. A. Acalepli*, p. KJO, fig. 220.

TROPHOSOME. — Hydkocaulus slender, profusely and irregularly branched,
rising to the height of an inch and a half; perisaec annulated on the ultimate ramuli.

GONOSOME. — Male GONOPnoKES in clusters upon the remains of atrophied
hydranths.

Colour. — General colour light pinkish.

Development of Gonosome. — June.

Locality. — Massachusetts Bay, Mr. A. Agassiz.

The diagnosis given above is derived from Mr. A. Agassiz's account of the species, aided
by the woodcut which accompanies his description, and which represents a male specimen of this
hydroid. He adds that the "species can at once be distinguished from the Eudendriurn dispar,
Agass., by its large clusters of medusae (sporosacs), while in the Eudendrium dispar the medusae-
buds are always somewhat scattered and never clustered together, as in Eudendrium tenue"

I can scarcely understand this passage, as the male gonophores of Eudendrium dispar have
been described and figured by Professor Agassiz as forming a verticillar cluster round the base
of the hydranths, the female gonophores only being scattered — the usual condition in Eudendrium.
It would seem as if Mr. A. Agassiz had by some confusion compared the male gonophores of his
Eudendrium tenue with the female gonophores of Eudendrium dispar.

Eudendrium tenue is evidently a form closely allied to the Eudendrium capillare of the
European side of the Atlantic. From this species, indeed, it would seem to differ chiefly in its
larger size, and in the fact (judging from Mr. A. Agassiz's figure) that the annulation of the
perisarc is not confined to the origin of the branches. There are slight differences, and I am by
no means convinced that the American hydroid is specifically distinct from the British.

EUDENDRIUM RACEMOSUM. 341

'*^^ 10. EUDENDRIUM RACEMOSUM, CaVoHui, Sp.

Sertularia racemosa, — Cttvolini, Mem. dei Pol. Mar., Sprcngel's German translation, p. 73,

tab. vi, fig. 1, &c.

TROPHOSOME.— HrDROCAULUs much and irregularly branched, attaining a
height of six or seven inches ; perisarc annulated on the ultimate ramuli for some
distance just below tlie hydranths. Htdranths with about thirty tentacles.

GONOSOME. — Male gonophores three-, four-, or five-chambered in umbel-like
clusters on atrophied hydranths ; female gonophores scattered in racemose groups on
the stems of atrophied hydranths.

Colour. — Hydranths reddisli ; perisarc reddish brown ; female gonophores purpHsh red.

Development of Gonosome. — Spring and summer.

Habitat. — On submerged rocks.

Localiti/. — Caves of Gajola, Bay of Naples, Cavolini.

After Trembley and Ellis there is no naturalist of the last century whose claims as an original
and accurate observer of the Hydroida rank so high as those of Filippi Cavolini; for like Ellis
he studied the animals in their native haunts along the sea-shore, and described and drew them,
not from dead and desiccated specimens, but in all the perfection and marvellous beauty of their
living forms.

Among the hydroids studied by the celebrated Neapolitan naturalist was the present
species, of which he has given us a careful and elaborate description and excellent figures. It is
true that in some points he falls into errors of interpretation, as may well be expected when we
bear in mind the state of hydroid zoology at the period when he wrote. Thus, having observed
in his hydroid both male and female gonophores, and, in accordance with the general views of
the zoologists of that day, believing those bodies to be eggs, he describes the species as producing
eggs of two kinds — eggs in racemes and eggs in umbels. His description, however, leaves little
to be desired. It is from it that I have selected the characters out of which the diagnosis here
given has been framed.

There can be no doubt that Eudendrium racemosnm is closely allied to the Eudendrium
rnmoaum of Linnaeus, the " small ramified tubular coralline " of Ellis. It would appear to differ
from it chiefly by its more irregular ramification, and by the more numerous chambers of the
male gonophores.

342 HYDRACTINID^.

INDETERMINABLE SPECIES.

Under the name of "Eudendrium cinrulatum," Stimpson (' Marine Invertebrate Fauna of
Grand Manan,' p. 9) gives the following description of a hydroid from the Atlantic shores of
North America :

" Polypidom small, very irregularly branched, somewhat as in Eudendrium rameicm, but not
so thickly; branchlets strongly ringed sometimes throughout their length, always near their
origin ; polypes small, with long tentacles and broad blunt proboscis. It differs from Eudendrium
rameum in the mode of branching."

" Dredged in twenty fathoms on a shelly bottom off Duck Island."

The description is unaccompanied by a figure, and it is otherwise too vague for an adequate
diagnosis. It is not even apparent whether it is intended to apply to a true Eudendrium or to
one of the other genera with which Eudendrium was at that time confounded.

HYBRACTINIDvE.

TROPHOSOME. — Hydrophtton forming a continuous adherent expansion, its
deeper parts consisting of freely intercommunicating tubes of ccenosarc, each invested
by a chitinous perisahc, and all adnate to one another by their sides, its free surface
overspread by a layer of naked ccenosaec. Hydeanths with filiform verticillate
tentacles.

GONOSOME. — GoNOPHOUES in the form of fixed spokosacs.

HYDRACTINIA. 343

HYDRACTINIA, Van Boieden.

Name. — A compound of the two generic names Hydra and Actinia, so named from a supposed
union of tlie external features of botli tliese genera.

EcHiNocHORiuM, — Hctssall.
Synhydra, — De Quatrcfai/cs.
Dysmorpiiosa (?), — Philippi.

TROPHOSOME. — Hutdkanths daviform, developed at intervals from the free
naked surface of tlie hydrophzton ; tentacles filiform, forming a single circlet round
the hase of a conical hypostome.

GONOSOME. — Sporosacs supported on blastostyles, which arise like the
hydranths from the free naked surface of the hydrophyton, each carrying round its
distal extremity globular clusters of thread-cells, which take the place of the tentacles
in the hydranth.

The generic name of Ilydradinia without a specific designation was given by Van Beneden
in 1841 to a hydroid obtained on the coast of Ostend.' The short memoir in whicli this name
is thus for the first time used contains no zoological description of the hydroid, beyond what is
involved in an account of the gonophores, to which the description is confined, and which, in
accordance with the views generally entertained at that period, are regarded as eggs. A figure,
however, is given of one of the hydranths, as well as of a blastostyle loaded with its gonophores,
and there is therefore little difficulty in identifying the animal to which the name of Ilydradinia
was intended to apply.

Nearly at the same time Dr. Hassall obtained in Dublin Bay specimens of a production
with which zoologists had been familiar, as forming a horny spinous crust on the surface of empty
univalve shells, and which Fleming, De Blainville, and Johnston, had regarded as the horny ectocyst
of a polyzoon. Hassall, however, from an examination of living specimens, became aware of the
incorrectness of this interpretation, and, recognising its hydroid affinities, assigned to it the name
of Eddnocorium claviyerum." We now know that the Echinocoriiim of Hassall is the animal to
which Van Beneden had just before given the name of Hydradinia.

In 1842 Dr. A. Philippi' noticed under the name of Dysmorpiiosa conddcola a hydroid
from the Bay of Naples. He tells us nothing of the gonosome, and his description, so far as it
goes, will apply in all points to the Hydradinia of Van Beneden ; but it is also just as applicable

' ' Bui. de I'Acad. Roy. de Bru.x.,' t. viii, 1841.
- ' Ann. Nat. Hist.,' 1841, vol. vii.
' ' Wiegman's Archiv,' 1843.

344 HYDRACTINIA.

to another genus, namely, Podocori/ne, sliortly afterwards instituted by Sars, and we have at
present no chie to enable us to say which of these two forms was intended.

In the following year the Hydractinia of Van Benedcn became the subject of an elaborate
and beautifully illustrated memoir by j\I. de Quatrefages, in which this animal was described
under the name of Synhydra parasites} De Quatrefages here shows that the zooids which
support the gonophores are destitute of mouth, and are otherwise of a different fomi from the
alimentary zooids of the colony. It is true that Van Beneden, in the memoir just referred to,
has figured a gonophore-bearing zooid as deprived of tentacles, but he has made no allusion to it
in the text, while De Quatrefages has not only fully described it, but has insisted on its constancy
and its value as affording a character of generic importance. De Quatrefages must therefore be
fairly regarded as the first who has made us properly acquainted with the occurrence of a blasto-
style among the Tubularian hydroids, a form of heteromorphism of great importance in its bearing
on the morphology of this group of animals.

De Quatrefages' memoir, notwithstanding some erroneous views, which could scarcely have
been avoided in the state of hydroid zoology when he wrote, contains by far the best account
which had up to that time been given of this interesting hydroid, and is perhaps the most valuable
which had appeared on any hydroid since the publication of Loven's famous memoir on Campanu-
laria and Syncorync.' The name of Synhydra, however, must yield to that of Hydractinia,
whose acceptance among zoologists has been secured by the fact of its prior publication.

In 1844 Van Beneden defined for the first time his genus Hydractinia, under which he
included two species with the names of Hydractinia rosea and Hydractinia lacteal He has
since, however, admitted that these two forms are only different sexes of one and the same
species. Indeed, there is no reason for supposing that the animal on which Van Beneden
founded his genus Hydractinia is other than the Alcyonium echinatum of Fleming ; and notwith-
standing Fleming's entire misapprehension of the affinities of this animal, his specific name must
stand instead of either of those assigned to the hydroid by Van Beneden, a change to the justice
of which the distinguished Belgian naturalist himself assents in a subsequent memoir.

In 1S5G an excellent account of Hydractinia ecldnata was published by Dr. Strethill
Wright.* His observations on its structure are very full and accurate, and his memoir supple-
ments in many important points what had hitherto been published on this hydroid. He calls
attention to the curious spiral zooids which occur near the margin of the colony, structures
with which I had myself become acquainted several years before, though Dr. Wright was the
first to publish an account of them.

In 1862 Agassiz described under the name of Hydractinia polyclina a North American
representative of this genus, which differs very slightly from the European form. He makes it
the subject of one of the elaborate and finely illustrated memoirs in his great work on the
natural history of the United States.^

Finally, in 18GG, Van Beneden gives us a second and mere extended description, with new

^ 'Ann. des Sc. Nat.,' 181-3, vol. xx.

- ' Wiegnian's Arcliiv,' 1837.

^ ' Reclierches sur les Tubulaiies.'

* ' Edin. New Phil. Journ.,' April, 1857.

» ' Contr. Nat. Hist. U.S.,' vol. iv, p. 227.

HYDRACTTNIA ECHINATA. 345

figures of liis genus Ilydractinia, which, however, he mixes up with Podocoryiie} He rephiccs
ills former specific names of lacle<t and rosea by Fleming's name of echinafa, while he describes
from imperfect data what he regards as three new species of the genus. Whatever these may
be, tiiere are no grounds for referring them to Hydradhiia in the sense in which this genus is
limited in the present Monograph, though in the wider sense in which it is understood by Van
Beneden, they may legitimately 'n\\i\ a place in it.

1. Hydractixia ECHINATA, Flemhuj .
Plate XV, and Plate XVI, figs. 10 and 1 1 .

Alcyonium ECHiNATUM, — Fleming, British Animals, 1828. Blainville, Actinologie, p. 525.

Alcyonidrum ECHINATUM,. — Johiiston, British Zooph., 1838, p. 304, pi. xlii, figs. 3, 4.

CoRYNE sQUAMOs.\, var., — Johnston, British Zooph., 1838, pi. ii, figs. 4, 5.

Hydractinia, sp., — Van Beneden, Bui. de I'Acad. Roy. de Brux., torn, viii, 1841.

EciNOCHORiuM CL.wiGERCM, — Hossall, Ann. Nat. Hist., 1841, vol. vii, p. 371, pi. x, fig. 5.

Dysmosphosa conchicola (?), — PA;///?/ji, Wiegman's Archiv, 1843.

Synhydra parasites, — De Qnatrefages, Ann. des Sci. Nat., 1843, vol. xx, p. 230, pis. 8, y.

Hydractinia lactea et H. rosea, — Van Beneden, Rech. sur I'Embryogenie des Tubulaires,

Mem. de I'Acad. Roy. de Brux., torn,
xviii, p. 104, pl. ix.

Hydractinia ecuinata, — Johnston, Brit. Zooph., 1847, p. 34, pl. i, figs. 4, 5. Strethill Wright,

Edinb. New Phil. Journ., April, 1857. Van Beneden,
Recherches sur la Fauna littorale de Belg., 1866, p. 134,
pl. xl, figs. 1 — 8 (exclusive of synonymes). Hincks,
Brit. Hydr. Zooph., p. 23, pl. iv.

TROPHOSOME. — Hydkanths crowded on the surface of the hydkophyton, and
attaining a height of about half an inch ; tentacular circlet formed by two or more
closely approximated series, the tentacles of the distal series being longer, and when
extended carried more erect than those of the proximal series. Cylindrical spirally
contractile appendages developed from the common basis of the colony near its
margin. Common basal expansion attaining a thickness of about a line and a half,
closely set with blunt conical spines which attain a height of about one line, and
which are traversed fi-om base to apex by longitudinal jagged ridges.

GONOSOME. — Blastostyles shorter and thinner than the hydranths ; spoeosacs
oval, almost sessile on the blastostyle, near whose distal extremity they are borne in a
cluster more or less dense.

' 'Recli. sur la Fauue lit. de Bels-,' p. 131.

346 HYDRACTINIA ECHINATA.

Colour. — Various shades of brownish red.

Development of Gonophores. — March to November.

Habitat. — Investing the surface of dead univalve shells, chiefly such as are inhabited by
hermit crabs, but occasionally occurring on other submerged bodies.

Bathy metrical distribution. — Littoral to deep-water zone.

Localities. — Atlantic shores of France and Belgium, Van Beneden, De Quatrefages. Gene-
rally distributed round the shores of the British Isles.

Hydractinia echinata is the only well-established European species of Hydractinia. It is
an abundant hydroid in the British seas, and may be easily kept alive in the confinement of our
tanks. Among some hundreds of specimens which have come under my observation, I have
never met with it except upon some gasteropodous shell, and then always with the shell
inhabited by a hermit crab ; or, if empty, affording evidence by abrasions of its surface that it
had at one time been similarly tenanted. A''an Beneden, however, has found it attached to
pieces of timber.

In the singular structure of its hydrophyton, and in the polymorphism of its zooids, it
possesses a deep significance for the student of hydroid morphology.

The spiral zooids constitute a vei-y striking feature in a coXouy oi Hydractinia echinata. They
occur close to the margin of the colony, and may be easily watched in a living specimen, when
they will be seen in a constant state of activity, rolling and unrolling themselves, and bending
backwards and forwards on the slightest irritation. Besides these Wright has described certain
long, contractile, non-spiral tenticula-like filaments, which also spring from the ccenosarcal base
near its margin. These have been also described by Hincks, but they are certainly not constant
in their occurrence, and I regard them as an abnormally modified form of some of the other
zooids.

Wright's description of sporosacs, borne directly by the basal expansion without the inter-
vention of blastostyles, I regard as founded on an error of observation, Dr. Wright having been,
as I believe, deceived by the shortness of the blastostyle which carried them.

I have made Hydractinia echinata the subject of one of the special studies of hydroid
anatomy given above (Part I, p. 220), where these and other morphological features of the
hydroid are discussed at length.

Gegenbaur figures under the name of Hydractinia a hydroid in which the sporosacs are
borne on ordinary hydranths instead of on blastostyles.' I believe that there is here some con-
fusion between Hydractinia and Stylactis.

^ Gegenbaur, ' Gruudziige der vergleich. Anatomie,' p. 90.

HYDRACTINIA POLYCLINA. 347

2. Hydractinia POLYCLINA, Agassiz.

lIvDRACTixiA POLYCLINA, — Afjassiz, Contrib. Nat. Hist. U.S., vol. iii, (il. xvi, and vol. iv,

p. 227, pi. xxvi, fig. 18.
IIvDRACTiNiA EciiixATA,^ — Leichj, Mar. Invert, of New Jersey, &c., p. 3, pi. xi, fig. 35.

TROPHOSOME. — IIydrocaulus crowded on the surface of the iiydrophyton,
tentacles tapering, equal or sub-equal ; margin of the colony carrying cylindrical
spiral appendages. Common basal expansion set with blunt conical spines traversed
by longitudinal jagged ridges.

GONOSOME. — Blastostyles with conspicuous hypostome and mouth. Spoeo-
SACS forming a more or less dense cluster near the distal end of the blastostyle.

Development of Gonosome. — July to February.

Habitat. — On univalve shells inhabited by Iierniit crabs, and on rocks in tide pools.

Batlaj metrical distribution. — From litoral to probably deep-water zone.

Locality. — Atlantic shores of North America, Agassiz.

It is not without great hesitation that I regard the Hydractinia polycUna of Professor
Agassiz as distinct from the European //. echinata. I have in vain sought in the description
given by Agassiz for characters sufficiently well marked to remove all doubt as to the justice of
the separation. The strongest arc included in the above diagnosis, which I have compiled from
Agassiz's description, and these will hardly be deemed altogether satisfactory. The apparently
more tapering form of the tentacles when extended, the greater equality in their length, and the
more decided development of an oral orifice in the blastostyle, are all that we can find ; and even
these do not agree with the characters expressed in Leidy's figure of a species which Mr. A.
Agassiz regards as distinct from the European form to which Leidy had referred it, and identical
with the Hydractinia jiolyclina of the American coast.

Professor Agassiz informs us that the //. polyclina is not only found on the shells of
gasteropods inhabited by hermit crabs, but that it occurs in great abundance in tide pools, where
it covers the rocks " for several square feet with a rosy, velvet-like carpet, presenting a delicacy

' It is on the authority of A. Agassiz (' Illustr. Catal.,' p. 198) that I include the Hydractinia
ecldnata of Leidy among the synonymes of Hydractinia polyclina. I take for granted that Mr. A.
Agassiz, who refers all the recorded North American species to the H. polycUna, has compared Leidy's
hydroid ^Tith the true H. polyclina of Prof. Agassiz, and has satisfied himself of the identity of the
two forms. I must confess, however, that I can find nothing either in the description or in the figur;-
given by Leidy to justify a belief that the H. ecldnata of this zoologist is distinct from our European
form,

45

348 PODOCORYNIDyE.

and vividness of tint which can hardly be described." This haliit is certainly so different from
that of our European form as to give support to the view which would regard the two as
specifically distinct.

PODOCORYNIDjU.

TROPHOSOME. — Hydkophyton a continuous adherent expansion formed by
adnata and inosculating canals ; the deeper part with its component canals invested
by a chitinous perisaec while a layer of naked ccenosarc spreads over the free
surface. Hydeanths with verticillate filiform tentacles.

GONOSOME. — GoNOPHOEEs phanerocodonic.

PODOCOHYNE, Sars (in part).

Name. — From ttouc, a foot, and Cori/ite, a genus of hydroids, in allusion to its broad basal
expansion.

HvDRA, — Rud. Wagner.
DrsMORPHOSA (?), — Philippi,
Hydractinia, — Chr. Loven.

TROPHOSOME. — Hydranths claviform, with a single circlet of filiform tentacles
surrounding the base of a conical hypostome.

GONOSOME. — GoNOPnoREs phanerocodonic, borne on the body of the hydranth
at the proximal side of the tentacles. Planoblast with a deep bell-shaped umbrella,
a small four-lipped manubrium, four radiating canals, and four or eight marginal
tentacles with bulbous bases which are destitute of oceUi.

We are indebted to Sars for the institution of the genus Podocorytie, founded on a hydroid
which he discovered in the Norwegian seas. The peculiar condition of its hydrophyton, con-
sisting as this does in its fully developed state of a continuous hydrorhizal expansion, formed by

PODOCORYNE CAllNEA. 349

adnate chitinous tubes, but covered superficially by a naked layer of ccenosarc, constitutes one of
its most remarkable characters, a cluu-actcr, lioAvever, which it has in common with lli/dractinia.

Sars believes that the hydroid described a short time previously by Philippi,* under the
name of Di/smorphosa conchicola, is identical with his Podocoryne carnea ; and the same view is
entertained by Krohn.' This may be true, liut, as already said, Philippi's description will just
as closely apply to IL/dracfniia:''

1. PODOCOETNE CAENEA, SarS.

Plate XVI, figs. 1—9.

PoDocoRYNE CARNEA, — Suis, Fauna lit. Norvegiffi, 1816, p. 1, pi. i, figs. 7 — 18. Krohn, Wieg.
Arch., 1851, p. 263. Allman, in Ann. Nat. Hist., July, 1859.
PoDocoRVNE ALBiDA, — ScTs, Fauua lit. Norvegise, 1846, p. 7,
Hydractima echinata, — Chr. Loven, Ofversigt af Kong. vet. Akad. Forhandl., 1757, p. 305,

tab. iv.
PoDOcoRYNE TUBULARi.E, — Savs, Middelliavets Literal Fauna, p. 36, note.

TROPHOSOME. — Hydrouhizal expansion forming a tliin layer, from whose free
surface minute blunt cliitinous spines project at intervals. Htdranths springing at
short distances from one another over the surface of the hyclrorhiza, and presenting
two different conditions of development; those destitute of gonophores attaining a
height of about two lines, and having usually about twelve tentacles, while those
which carry gonophores are about half the size of the former, and possess no more
than from four to six tentacles.

GONOSOME. — Gonophores on very short peduncles, forming a verticillate cluster
at a little distance below the tentacles of the smaller hydranths ; medusa with the
outer surface of the umbrella dotted with scattered thread-cells, velum well developed ;
lips of manubrium narrow, each terminated by a little pencil-like cluster of stalked
thread -cells.

Colour. — Hydranths varying from white to flesh colour, medusa with the aianubriuni and
bulbous bases of marginal tentacles brownish red.
Development of Gonosome. — j\Iay to August.

1 'Wiegmann's Arch./ 1812, p. 37.

= Id., 1851, p. 263.

^ See the general remarks on the genus Stylaclis given above, p. 303, note.

350 PODOCRYNE CARNEA.

Ilahiiat. — On stones in the sea and on old sliclls, chiefly of Gasteropoda.
Bathymetrkal didrihution. — Literal to deep-sea zone.

Localities. — Coast of Norway, Sars ; Bay of Naples, Krohn, Sars, Costa ; Firth of Forth,
G. J. A.

A hydroid which, in all respects, except in colour, resembles that described by Sars as
Podocoryne carnea, was found by me in the Firth of Forth growing on stones in rock pools a
little above low-water mark, and also on old Buccinum shells in similar situations, and on others
brought up from deeper water on the lines of the fishermen.

My specimens, instead of being pale red, like the Podocoryne carnea of Sars, were nearly
colourless ; but I consider this difference of colour as only varietal, and I believe that a white
Podocoryne described by Sars as a distinct species imder the name of Podocoryne allnda is, like
mine, only a variety of his Podocoryne carnea.

Sars' account of his Podocoryne carnea is very full, and possesses a special interest in having
been given at a time when as yet but few instances were known of the production of free medusas
by fixed trophosomes.

Hincks notices the occurrence in Podocoryne carnea of spiral appendages, like those of
Hydractinia echinata, and describes them as occupying the same position near the margin of the
colony as the corresponding bodies do in the latter hydroid. He also observed on certain parts
of the basal expansion tentacula-like slender filaments, which he compared to similar filaments
which have been described in Hydractinia.

Neither kind, however, was present in any of the specimens that came under my observation.
Whatever be the nature of the spiral bodies observed by Hincks, they certainly do not possess
the constancy which characterises the spiral appendages of Hydractinia, and it is difficult not to
regard both the spiral bodies and the tentacula-like filaments observed by Hincks in Podocoryne
as merely abnormal alterations of some of the ordinary hydranths. In well-preserved spirit
specimens of a Podocoryne from the Bay of Naples, for which I am indebted to Professor Costa,
and which I am unable to distinguish from P. carnea, very delicate capillary filaments existed
here and there on the coenosarcal base. These were, however, undoubtedly vegetable growths,
though, without careful examination, they might easily have been mistaken for component parts
of the colony.

Krohn, in his account of Podocoryne carnea, published some years after that given by Sars,
informs us that the medusa; in his specimens had attained to sexual maturity, and that tiie
generative elements, male and female, had been developed in the walls of the manubrium even
before the liberation of the medusa from the trophosome ; while a similar observation as to the
early sexual maturity of the medusa was made by Chr. Loveu in the specimens described by him.

Neither in the specimens examined by Sars nor in those which had come under my own
observation was there any trace of generative elements. I should feel tempted to regard this
difference as indicating a difference of species, but, as no other characters are mentioned by either
Krohn or Loven which would justify the separation, I believe it will be better for the present to
consider the difference in question as depending on local circumstances, favorable in the one case
and not so in the other to an advanced development of the medusa.

The shells on which Podocoryne carnea is found arc, like those selected by Hydractinia,
inhabited in almost every instance liy a hermit crab.

rODOCORYNE TROBOSCTDEA. 331

In a note to his account of Slj/Iadis {Podocoryne) fuckola, contained in liis ' Literal Fauna
of the Mediterranean/ Sars describes under the name of Fodocuryne Ticbidarice a species whicli
he discovered on the Norwegian coast, where it occurs in abundance on the stems of Tabidaria
indivisa from a depth of between tliirty and forty fiithoms. Tiiere is nothing, however, in Sars'
description of this species to distinguish it from Podocoryne car/tea.

3. Podocoryne proboscidea, Hincks.

Podocoryne proboscidea, — Hincks, Brit. H\'dr. Zoopli., p. 317, pi. xxiii, fig. 1.

TROPHOSOME. — Hydrantus " tall and rather stout, with a very long and some-
what columnar proboscis, and with about fourteen tentacles, some of which are tall
and erect, and others short and borne at right angles to the body." " No apparent
difference between the prolific and barren " hydranths.

GONOSOME. — GoNOPnoRES "forming a large collar round the polypite (hydranth)
at a short distance below the tentacles, disposed in two rows, and borne on small
tubercles;" plaxoblasts permanently attached, umbrella deep bell-shaped, with the
marginal tentacles in the condition of eight short conical processes, manubrium
surrounded by the mass of generative products very large, and projecting beyond the
codonostome.

Colour. — Hydranths orange brown, with opaque-white proboscis ; manubrium of planoblast
(male) surrounded by its generative elements orange with purplish base.
Development of Gonosome. — September.
Habifat. — On Laniinaria roots and the stones in rock pools.
Bathymetrical distribution. — Laminarian zone.
Locality. — llfracombe. Rev. T. Hincks.

Podocoryne jjroboscidea is stated by j\Ir, Hincks to be a larger species than P. earnea, and
to be at once known by its long, cylindrical, and very conspicuous proboscis, Avhich has the
appearance of being fluted down the sides. The planoblasts develope their generative elements
while yet attached to the trophosome, and, in accordance with the general rule in such cases,
appear somewhat arrested, the marginal tentacles presenting the form of small tubercles. Hincks
regards this condition of the planoblasts as depending on the season of the year during which
they are produced, and believes that, if observed in an earlier month, they would be found to be
of the usual form. I cannot see, however, any evidence in favour of this view, and must regard
the conditiiMi in which Hincks has witnessed the gonosome in the present species as its constant
one.

jVo observations were made as to the nature of the conunon hydrorhizal expansion, and it is
taken for granted that it belongs to the podocorynal type.

352 PODOCORYNE ACULEATA.

3. PoDOCORTNE ACULEATA, liud. Wagner, sp.

Hydra aculeata, — Rudolph Wagner, in Iris, 1833, p. 250, pi. xi, figs. 1 — 10.

TROPHOSOME. — 'Those hydranths which are destitute of gonophores attain-
ing a height of about two lines, and with usually from eight to twelve tentacles ; those
which, carry gonophores smaller, and with about five tentacles.

GONOSOME. — GoNOPnoRES on short peduncles springing in an imperfect verticil
from the hydranth at a little distance below the tentacles ; medusa never attaining
complete development, having a broad, flattened summit and four stunted marginal
tentacles.

General colour. — Yellowish.

Development of Gonosome. — Observed during May.
Habitat. — Investing various univalve shells in the sea.
Batltymetrical diHirihution. — Literal zone.
Localiiy. — Adriatic Sea, near Trieste, Rud. Wagner.

Towards the end of the last century Cavolini described the medusa of a Pennaria as loaded
with ova while still attached to the trophosome,' and though the Neapohtan zoologist failed to
recognise the true import of this observation, a phenomenon of vast significance in the life-
history of the IIydroida became thus for the first time recorded.

From that period to the time when Rudolph Wagner discovered his Hydra acideafa no
further observations were made tending to show that the medusiform buds of the fixed hydroids
were destined to reproduce the species by the formation of fertilised ova.

Wagner, however, saw that the medusa3 of his hydroid had, before attaining their freedom,
given rise to ova, which, after a time, were discharged into the surrounding water ; and though,
just as in the case of Cavolini's observation, the state of our knowledge of the Hyuroida at that
time caused the real significance of this fact to escape him, it was capable of affording to suc-
ceeding observers a strong argument in favour of the new views of hydroid development — views
which soon began to exert their influence throughout the whole study of the Hydroida.

That the " Hydra acideata " of AVagner is a true Fodocoryne there can be little doubt. Its
chief difference from the Todocoryne carnea is to be found in the fact that its medusa, though
referable to the same type as that of the species on Avhich Sars founded his genus, appears to be
somewhat arrested in its development.

Eor the description given above I have no authority beyond the original account of the

' Cavolini, ' Mem. Polyp.,' Sprengel's translatiou, p. G5, tab. 5.

PODOCORYNE AREOLATA. 353

hydroid published by Wagner. As may be easily expected, no notice is taken in that account
of certain points which arc now found to be necessary for a satisfactory diagnosis, and the species
has not been met with by any one since it was found by Wagner, many years ago, on the shores
of the Adriatic.

4. PODOCOEYNE AREOLATA, AlcUr, sp.

Hydractinia AREOLATA, — Aider, Suppl. Catal., p. 1, pi. ix, figs. 1 — 4.
Rhizocline AREOLATA, — Allmau, Ann. Nat. Hist, for May, 1864.
PoDocoRYNE AREOLATA, — Hiucks, BHt. Hydr. Zooph., p. 33, pi. vi, fig. 1.

TROPHOSOME. — Common basis of the colony set witli " simple linear spines in
irregular^ groups." Htdranths attaining a height of about one tenth of an inch,
sub-cylindrical, with from six to ten tentacles.

GONOSOME. — GoNOPiiORES sessile on the common basal expansion, large, globular,
or slightly pear-slia])ed. Radial tentacles of medusa rather short, three still shorter
tentacles developed in each inter-radial space of the margin ; manubrium with each of
its four lips supporting a tuft of thread-cells.

Colour. — Hydranths white ; radiating canals of medusa golden yellow.

Habitat. — On dead univalve shells.

Localities. — Coast of Durham, Mr. Alder; Shetland, Rev. A. ■\I. Norman.

Some years ago, when attempting a revision of the genera of tubularian hydi'oids, I saw that
it was impossible to retain the hydroid named Hydractinia areolata by Alder in the genus to
which this naturalist had assigned it, and, not being able to find a place for it in any published
genus, I constituted for it a new one under the name of RMzocIine.

I do not, however, desire to insist upon the retention of the genus Rhizocliiie. It is highly
probable that the basal expansion in jMr. Alder's hydroid resembles that of Hydractinia, and
that its true natiu'e had been overlooked when he described it as " consisting of a solid chitinous
expansion, from which arise simple chitinous spines." I have further satisfied myself that there
is no essential difference between the common base of Hydractinia and that of Podocoryne ; and
then the only distinction between Alder's hydroid and the admitted forms of Podocoryne will be
found in the position of the gonophores, which in the former arc sessile on the common base.
This can hardly amount to a generic difference, though its very exceptional character might lead
us to expect such a diflcrcnce in other parts of the hydroid. We have accordingly — though as

^ In Alder's description the word is " regular," but this is plainly a misprint.

;554 CORYNOPSIS.

yet by no means sufficiently acquainted with the " Ifj/dracfinia arcolala " — no alternative but to
allow the absorption of RInzodine into Podocoryne. This view is taken by llincks, and the
present state of our knowledge will justify it. It is possible, however, that a fuller accjuaintancc
with Mr. Alder's hydroid may require the restoration of the genus Rldzodine.

The resemblance between the medusa of Fodocoryne areolata and that of Podocoryne carnea
has been noticed by Alder. Indeed, the two medusae scarcely differ from one another, except in
the fact that while the medusa of Podocoryne carnea has at the time of its liberation only one
marginal tentacle in each inter-radial space, that of Podocoryne areolata has at the same period
three inter-radial tentacles between each of the radials. The middle one of these three inter-
radial tentacles is longer than that placed at each side of it, the latter being as yet scarcely
developed beyond the condition of a mere tubercle.

Of the two specimens which had come imder Mr. Alder's observation, one " was obtained
parasitical on a dead shell of Natica Alderi, brought in by the fishing boats at Cullcrcoats."
The other was " a dead and rather worn specimen upon Natica Granlandica among the zoophytes
collected in Shetland by the Rev. A. M. Norman."

COEYNOPSIS, AUman.

Name — Kopvvi], a club, and oi/.(c, face (resemblance), in allusion to the club-like form of
the hydranths.

Podocoryne, — Hodge.

TROPHOSOME. — Htdkanths claviform, with a single circlet of filiform tentacles
round the base of a conical liypostome.

GONOSOME. — Planoblasts borne on the body of the hydranth at the proximal
side of the tentacular circlet. Medusa deep bell-shaped; manubrium having its
mouth surrounded by four short capitate tentacles; radiating canals four, each
terminating distally in a bulb, from which are developed two tentacles, each with a
distinct ocellus at its base.

We are indebted to Mr. Hodge for the tirst notice of the hydroid on which the genus
Corynojjsis has been founded. He described it, however, as a Podocoryne, under the name of
Podocoryne Alderi; and though he does not give any definite description of the common base,
we may assume, until further observation proves the contrary, that this is similar to that of the
true Podocorynes. Taking for granted that there is no error of observation in ]\Ir. Hodge's
account of his Podocoryne Alderi, it is plain, however, that this hydroid must Ijc separated from
Podocoryne, for the medusa which he assigned to it is of an entirely different type from that of

CORYNOPSIS ALDERl. 355

the latter genus. It cannot, indeed, at the time of its liberation, be distinguished from that of
Jiou(/ainvUHa at the same stage of its development. The further progress of the Corynopsis
medusa has not been traced, but it is highly probable that it corresponds to that of the medusa
of Boiif/ainvUlia, and that we have thus two genera with similar gonosomes, though with different
trophosomes.

Corynopsis Aldeki, Hodge, sp.

PoDocoRVNE Aldbri, — Hodijc, in Trans. Tyneside Nat. Field Club., vol. v, pi. ii, figs.

10—15.
CoaYNOPsis Alderi, — Allman, in Anu. Nat. Hist, for May, 18G1. Hincks, Brit. Hydr.
Zooph., p. 31, pi. vi, fig. 2.

TROPHOSOME. — Htdranths attaining a height of from one quarter to half an
inch, gradually tapering from the distal to the proximal extremity, tentacles six to
twelve in number.

GONOSOME. — Medus^e springing from the hydranths in an imperfect verticil at
a short distance behind the tentacular circlet.

Colour. — Hydranth pale pink, medusa having its manubrium pale green, and tentacular
bulbs red.

Development of Gonosome. — August and September.

Habitat. — On shells from deep water.

Haihymetrical dhtrUmtion. — Deep water zone.

Locality. — Seaham Harbour, Coast of Northumberland, j\Ir. Hodge.

Considering how large an amount of caution is necessary in order to avoid referring to their
wrong trophosomes the free medusaj developed in our aquaria, it is impossible not to wish for
further observations which might tend to verify Mr. Hodge's reference of a Bonyainvillia-We
medusa to a trophosome, so very different as that of his Podocoryne Alderi is from the trophosome
to which this form of medusa had been hitherto traced. Coming, however, as the account does
from an observer so trustworthy as Mr. Hodge, we should not be justified in refusing to accept
it. This assent is all the more easy when we recollect that the present is not the only case of the
kind which has been recorded, for Hincks has shown that the medusa of Stauridia producta
cannot be distinguished, at least at the time of its liberation, from that of Coryne cxlinia.

46

356 CLADONEMID^.

CLJBONEMIDzE.

TROPHOSOME. — Htdrocaulus developed, invested by a peeisaec. Hydkanths
with two kinds of tentacles, one filiform, the other capitate.

GONOSOME. — GoNOPnoEES phanerocodonic with the radiating canals, more than
four, and with the marq-inal tentacles ramified.

CLADONEMA, DnjanUn.

Name. — From k\ucoq, a branch, aud vnm, a thread ; so named from the l)ranching tentacles
of the planoblast.

Cladonema (planoblost), Stauridie (trophosome), — Biijardin.

TROPHOSOME. — HYDEOPnYTON consisting of a branching or simple hydeocaultjs
arising from a creeping filiform hydeoehiza, the whole invested with a cliitinous
PEEisAEC. Hydeanths clavate with two verticils of tentacles, each verticil consisting
of four tentacles disposed in a cross, — the tentacles of the proximal verticil filiform,
those of the distal verticil capitate,

GONOSOME. — Planoblasts developed from the body of the hydranth. Mature
PLANOBLAST with a deep bell-shaped umbrella ; manubrium with the mouth surrounded
by five lips ; radiating canals, eight or ten, eacli continued at the umbrella-margin
into a tentacle with a bulbous base which carries an ocellus ; branches of tentacles of
two kinds, one kind very extensile and destitute of suctorial organs, the other situated
at the proximal side of these, scarcely extensile, and provided with terminal suctorial
capitula.

The beautiful observations of Dujardin on the production of free medusas from their hydroid
trophosomes were among the first which made us acquainted with this remarkable phenomenon,
and among these his observations on Cladonema are perhaps the most interesting.'

^ 'Ann. des Sc. Nat.,' vol. xx, 1843, p. 370, and vol. iv, 1845, p. 271, pis. xiv and xv.

CLADONEMA RADIATUM. 357

When Dnjardin wrote tlio nomenclature of tlic IIyuroida liad not yet acquired that definite-
ncss which subsequent investigations had conferred upon it, and the French zoologist designated
the trophosome of the present genus " Stauridie " {Staiiridium), while to the medusa which he
observed to be budded from his Stauridiiim he gave the name of Cladonema.

Dnjardin observed not only the gemmation of the medusa from trophosome, l)ut he saw its
manubrium become loaded with eggs, which were afterwards deposited on the sides of his vase,
where they became developed into young /S'/a?«-«/«««-trophosomes. This is the first observation
by which the whole life series of a hydroid from trophosome to medusa by gemmation, and back
agoin by true generation to the trophosome, has been directly followed.

The observations of Dujardin were afterwards confirmed and supplemented in some
important details by Krohn.^

Cladonema eadiatum, Dnjardin.
Plate XVII, figs. 1—10.

Cladonem.\ RADIATUM (plaiioblast), — Dnjardin, in Ann. Sci. Nat., vol. x.'s, 1843, p. 370 j and

vol. iv, 1845, p. 271, pis. xiv and xv. Krohn,
in Miiller's Archiv, 1853, p. 420, pi. xiii.
Gegenhaur, in Zeitschr. f. Wissensch. Zool.,
vol. viii, 1857, p. 230. Van Beneden, Fauna
lit. de Belg., p. 159, pi. xii. Keferstein
and Elders, Zool. Beitr., p. 85, pi. xiii, fig. 5.

Stauridie (trophosome), — Dnjardin, loc. cit.

CoRYNE STAURIDIA (trophosome), Gosse, Devonshire Coast, p. 257, pi. xvi, figs. 1 — G.

Cladonema radiatum, — Allman, in Ann. Nat. Hist, for May, 1864. Hincks, Brit. Hydr.

Zoopb., p. 62, pi. xi.

TROPHOSOME. — HroKOPHXTOX in the form of a very slender, creeping, ramified
filament, from which short, simple hydranth-bearing branches about one tenth of an
inch in height are given off at intervals, or presenting a free branching, slender stem,
which attains a height of half an inch or an inch ; peuisauc smooth. Htdeanth with
the tentacles of the proximal verticil much smaller than those of the distal.

GONOSOME. — Planoblast (when mature) with the vertical and transverse
diameters of its umbrella nearly equal, and with a slightly prominent boss upon its
summit ; velum very wide, oral appendages, in the form of short cylindrical processes,
terminated each by two or three small spherical clusters of thread-cells, the marginal
tentacles with about four extensile branches nodidated with clusters of thread-cells,

^ ' Miiller's Archiv,' 1853, p. 420, pi. xiii.

358 CLADONEMA RADIATUM.

and terminated each by a spherical cluster larger than those which are borne along
their length ; the sucker-bearing appendages, usually three in number, smooth.

Colour. — Body of hydranth very pale reddish ; perisarc light yellowish brown ; manubrium
of medusa very pale reddish, marginal tentacles and their branches crimson brown ; ocellus deep
crimson.

Development of Gonosome. — Spring and summer.

Habitat. — Attached to stones, &c., in the sea.

Bathymetrical distribution. — Unknown, probably confined to the litoral zone.

Localities. — Northern shores of France, Dujardin ; Coast of Belgium, Van Beneden ;
Coast of Devonshire, Mr. Gosse; Coast of Kent, Mr. Dovvker; Messina, Krohn, Gegenbaur,
Kefferstein and Ehlers.

I take for granted that the synonymy given above refers to one and the same species, namely,
the original Cladonema radiatum of Dujardin, because, though some slight differences may be
noticed between the descriptions, these do not appear to be sufficiently marked to justify ovu"
regarding them as of specific value. In some cases they are certainly the results of mere
differences of age.

Though the medusa of the present species has been occasionally met with in the open sea,
the trophosome has as yet been found only in the confinement of tanks appropriated to the
preservation of living marine animals, and in all those cases it has shown itself spontaneously
without any attempt being made to introduce it. Whether the form of the trophosome has
undergone any change in consequence of the artificial conditions to which it has been thus
exposed it is impossible to say. It presents itself, at all events, under two modifications ; in one
the hydrophyton runs over the supporting surface as a very slender branching filament, from
which equally slender simple branches, about one tenth of an inch in height, are given oft' at
intervals, each carrying a hydrantli ; in the other form (that represented in the plate) there is a
free branching, slender stem, which attains a height of from half an inch to an inch.

Since Dujardin the principal original observers of this hydroid have been Mr. Holdsworth
and Mr. Hineks in this country (see Hincks, loc. cit.), and M. Van Beneden on the Continent.
Some years ago I obtained specimens of the trophosome from one of the tanks belonging to the
zoological survey of London, in which it had appeared in abundance, but no planoblasts were
developed from them. I am indebted, however, to Mr. Dowker, of Stourmouth House, Kent,
for living specimens of Cladonema radiatum, presenting both trophosome and gonosome. They
showed themselves in a large tank filled with water from the neighbouring coast, and have
afforded me an opportunity of making a careful study of this remarkable hydroid (see above,
page 216).

There can be little doubt that Van Beneden is right in regarding a medusa, which made its
appearance in an aquarium supplied with water from the coast of Belgium, as the medusa to
which Dujardin gives the name oi Cladonema ; but it is not easy to say why he could have
thought himself justified in referring it to a naked //y^/rrtc/iw/a-like trophosome, which he noticed
at the same time in his aquarium.' In favour of this association there are not the slightest

' 'Faun. lit. de Bclgique,' p. 110.

CLADONEMA RADIATUM. 359

grounds, and the true tro[)liosome has been entirely overlooked by the learned Belgian zoologist.
Van Beneden's description of the medusa, however, is a valuable contribution to our knowledge
of the species. He has shown that the marginal tentacles are at first comparatively simple, and
that their ramification increases up to a certain point with the age of the medusa — statements
which I can confirm from my own observations.

The three smooth appendages which are given off" near the base of the tentacle serve the
medusa as organs of attachment, the terminal capitida of these appendages having the power of
adhering, sucker-like, to any surface to which they may be applied, and the little medusa may be
often seen attaching itself by means of them to the sides of the glass jar in which it is confined.

It is impossible to grow tired of watching this beautiful little medusa ; sometimes while dashing
through the water with vigorous systole and diastole, it will all at once attach its grapples to the
side of the vessel, and become suddenly arrested in its career, and then after a period of repose,
diu'ing which its branched tentacles are thrown back over its umbrella, and extended into long
filaments which float, like some microscopic seaweed in the water, it will once more free itself
from its moorings and start off with renewed energy.

When the medusa is in the act of swimming, the tentacles are contracted and curved
upwards round the margin of the umbrella. In this state the smooth appendages of attachment
undergo little or no contraction, but the nodulated appendages are strongly contracted, and
assume the appearance of short thick clubs.

The development of Cladonenia is through planula;, as has been shown by Krohn. Dujardin
has also seen the stauridioid trophosome developed from the egg of the medusa, but he over-
looked the intermediate stage of planula.

Dujardin has noticed a retroversion and ultimate absorption of the umbrella near the time
when the ova are about to be deposited, and this observation has been, according to Mr. Ilincks,
confirmed in great part by Mr. Holdsworth. ' The phenomenon may be compared to what has
been noticed above in Syncoryne pulchella (see p. 203), while similar observations have been
made liy Hincks and otliers on the medusa3 of Podocoryne and of Tarris (see Hiucks, op. cit..
Introduction, p. xxix).

Dujardin informs us that he has seen the ova of Cladonema adhering to the glass vessel in
Avhich the medusae were confined, and he states that he has seen the medusa remove the ova
from the cavity of its manubrium by means of its marginal tentacles, which are introduced into
the mouth for this purpose; while he further describes these tentacles as employed in glueing the
ova to the glass — statements which there can be little doubt are founded on some deceptive
observation.

360 NEMOPSID^E.

NEMOPSIDvE.

TROPHOSOME. — Htdeanths with a proximal and distal circlet of filiform
tentacles.i

GONOSOME. — GoNOPHOUES medusiforra planoblasts, with four radiating canals,
and with the marajinal tentacles clustered and dissimilar.

NEMOPSIS, Jffassh.

Name. — From vjj^iia, a thread, and oy^iQ, cye-siglit; in allusion to the resemblance of the
two clavate tentacles in each marginal ckister of the planoblast to the eye-bearing tentacles of a
snail.

TROPHOSOME. — Hydrophtton not known. Htdranth conical ; the proximal
circlet of tentacles surrounding the base of the hydranth, the distal one situated at a
short distance behind the mouth.

GONOSOME. — Planoblasts borne by the hydranth between the proximal and
distal circlets of tentacles. Medus.e, when mature, deep bell-shaped ; umbrella-walls
thick ; radiating canals, terminating each at the margin of the umbrella in a bulb
which supports one of the clusters of tentacles ; two of the tentacles in each cluster
clavate and but slightly contractile, every tentacle carrying a distinct ocellus at
its root ; manubrium with four dichotomously branched oral tentacles ; generative
elements produced in four lobes, whose basis of attachment partly rest on the walls of
the manubrium, and partly extend for a greater or less distance along the length of
the radiating canals.

The genus Ncmojms was founded by Professor Agassiz" for a medusa captured on the

' The hydi'ophyton of this family has not yet been discovered.
' ' North American Acaletha;/ part i, p. 289.

NEMOPSIS. 361

Atlantic shores of North America. He knew nothing, however, of the trophosome, while he fell
into the mistake of supposing that the two clavate tentacles which occnr in every marginal
cluster of the medusa, carry each an ocellus on its summit, an error which was afterwards
rectified by M'Crady, to whom we are also indebted for having been the first to make us
acquainted with the trophosome of a congeneric species.

A medusa, nearly allied to the Nemopsis described by Agassiz, was believed hy M'Crady to
be developed by budding from a free hydranth which he captured in the open sea, and he was
thus led to the belief that Nemopsis presented the very exceptional condition of possessing a free
floating trophosome. I feel, however, almost sure that the body here regarded by M'Crady as
the entire trophosome of Nemopsis is only a hydranth detached from its stem. We know that
this phenomenon is common in Tuhularia, in ■which we find the hydranth periodically detached
from the hydrocaulus, carrying with it its burden of gonophores, and continuing for some time
to live in this detached condition ; and I consider it highly probable that not only the free
hydranth of Nemopsis, but the supposed free trophosome of Stimpson's Acaulis, affords merely
another example of this phenomenon.

M'Crady appears to entertain no doubt that the medusa to which he gives the name of
Nemopsis Gibbesii, and which he found free in the open sea, is derived from this free trophosome,
which, at the time of its capture, carried numerous medusa-buds ; but here again, though there
is no reason against the truth of this conclusion, the evidence is not complete, for it does not
seem that M'Crady traced directly the development of the medusa-buds into the adult Nemopsis-
medusa, but that he merely inferred the relation of the buds to the adult form from the observa-
tion of intermediate stages, which he discovered in the open sea, and through which he convinced
himself that the free adidt medusa could be connected with the buds detached from the tropho-
some. Coming, however, as this opinion does from so excellent and conscientious an observer as
M'Crady, we may, with all safety, accept it in anticipation of further confirmation.

The presence of two different kinds of tentacles in each marginal cluster of the medusa is
an interesting feature in the genus ; and another very important one is the way in which the
generative lobes, which are at first confined to the walls of the manubrium, gradually extend
down the radiating canals towards their marginal termination.'

The medusa of Nemojjsis is intimately allied to that of BougainvilUa, which it especially
resembles in the fact of each marginal bulb carrying a cluster of tentacles with ocelli, in that of
the mouth being provided with bifurcating tentacles, and in the great thickness of the umbrella
walls.

See above, p. 63, where the significauce of this character is discussed.

362 NEMOPSIS GIBBESII.

Nemopsis Gibbesii, H'Crady.

Nemopsis Gibbesii, — i)f'Cr«f/y, Gymnoplitlial. of Charleston Harbour, p. IGO, pi. x, figs. 1 — 7.

TROPHOSOME. — Htdeanths having the form of an elongated cone ; tentacles
about ten (?) in the proximal circlet, and six (?)' in the distal ; those of the proximal
circlet in two closely approximate alternating serieso

GONOSOME. — MEDUSA-buds on short, simple peduncles, which are scattered
irregularly over the space between the proximal and distal sets of tentacles, Plaxo-
BLAST, at time of liberation, with three or four tentacles in each marginal cluster,
and with its manubrium destitute of oral tentacles. Mature medusa with its vertical
shghtly exceeding its transverse diameter ; sixteen filiform and two clavate tentacles
in each marginal cluster ; oral tentacles very much branched, springing from a point
at some distance above the mouth ; generative lobes extending from the manubrium
along the radiating canals for about two thirds of the length of the canal.

Development of Gonosome. — Winter.

HaUtat of detached (?) hydrantb. — Free, floating at large in the open sea.

LoeaJify. — Charleston Harbour, M'Crady.

Our sole knowledge of the hydroid just descril)ed is derived from the observations of
M'Crady, no other naturalist having yet met with it. It is full of interest, even though we may
be permitted to doubt the normal freedom of the complete trophosome. " The appearance of the
medusa,'' says M'Crady, "is at once singular and beautiful. The conspicuous crescentic outline
of the pale orange-coloured sexual ribbons, the vivacioxis movements of the mouth and its
appendages, the graceful waving outline of the flapping disc, with the clavate tentacula carried
erect, as if always on the watch, the others floating in xarious curves, or lightly curled at their
extremities, make it an unusually remarkable object even in this remarkable group."

The ocelli of the two clavate marginal tentacles of the medusa are, according to M'Crady,
" carried on their lower surface turned somewhat inwards, that is, towards each other ; the
ocelli of the other tentacles on their inner surface."

M'Crady believes that he has been able to follow the development of the medusa from an
early condition as a bud still attached to the trophosome to that state in which the generative
lobes have attained their complete form ; and his observations have been repeated and in many
respects confirmed by A. Agassiz in planoblasts captured at large in the open sea. He has

^ M'Crady makes no mention of the number of tentacles; the numbers here given are such as
appear to be indicated by his figures.

PENNARIDiE. 363

shown that at first the marginal tentacles are few, and consist of the filiform tentacles alone, the
clavate not making their appearance until a later period, while the oral tentacles are also entirely
absent in the earlier stages. When first noticed these were in tlie condition of bifid filaments,
and then gradually assumed the greatly ramified form which ultimately distinguished them. He
has also shown that as development proceeds the generative lobes gradually extend themselves
from the base of the manubrium downwards along the length of the radiating canals.

Only one specimen of the trophosome has been observed. " It was floating at large, and
taken with the dip-net. It lived five days, developing medusae, but never fixed itself, only
gradually dwindling away as the medusae were developed. The tentacula were all at last retracted,
especially those around the broad base. In its first activity it was incessantly moving about l)y
means of its tentacula, mouth downwards." ^

Nemo2)sis Bachii is the name given by Agassiz to the medusa on which he founded the
genus Nemopm. As we know nothing of its trophosome, it docs not come within the range of
the systematic portion of the present monograph.

PENNARIDjE.

TROPHOSOME. — IlrDiiocAULirs developed or not. Htdkaxths with two kinds
of simple tentacles, one filiform, the other capitate.

GONOSOME. — GoxopiiORES [where known] mednsiforra planoblasts, with four
radiating canals, and one or four more or less developed simple marginal tentacles.

Peknauta, Goldfuss.

JSfame. — From penna, a feather ; in allusion to the symmetrical, feather-like disposition of
the branches of the trophosome.

Sertulakia, — Cavolini.

TROPHOSOME. — HrDROPnTTON composed of a symmetrically ramified htdeo-
CAULUS, rooted by a creeping filiform nrDuouHiZA, the whole invested with a chitinous

^ M'Cradv, op. cit., p. 1G3.

47

364 PENNARIA CAVOLIXII.

PERiSARC. IlrDRAXTns flask-sliaped, with the filiform tentacles constituting a proximal
set, and arranged in a single verticil round the base of tlic hydranth, and the capitate
tentacles a distal set scattered on the body of the hydranth.

GONOSOME. — GoNOPHORES developed in a more or less perfect verticil between
the proximal and distal sets of tentacles. Umbrella deeply ovate ; manubrium large,
destitute of oral appendages ; marginal tentacles four, rudimental, no ocelli.

The genus Pennuria was constituted by Goldfuss for the Scrfiihiria j)e»iiaria of Cavolini, a
well-marked generic form, so well marked as to justify its assumption as the type of a distinct
family of Gymnoblastic hydroids.

*:j^* 1. Pexnaeia Cavolinii, EJirenhcrg.
Woodcut, fig. 80.

Serti'laria PENNARIA, — CavoUni, Mem. Pol. Mar., Sprengel's trans., p. 61, tab. v.
Pennaria disticha, — Goldfuss, Handbuch der Zool., p. 89.

Pennaeia Cavolinii, — E/irenberff, Corallenthiere, Abhandl. Akad. Berlin, 1832, p. 297.

K'6lliker, Zeitschr. f. Wiss. Zool., vol. iv, 1853, p. 303.

TROPHOSOME. — ^IIyduophxton attaining a height of six or seven inches ; main
stem slightly zig-zag, and with a uniform and gentle curve from base to apex ;
branches given off alternately from each side, with a regular distichous arrangement,
PERISARC annulated at the origin of the branches, and on the ultimate hydranth-
bearing ramuli. IIxdrantus borne on the summits of the main stem and primary
branches, and on those of very short simple nearly cylindrical ramuli, which spring at
regular intervals from the distal side of the primary branches; body of hydranth
tumid at the base, and thence rapidly tapering to the mouth, so as to assume the form
of a flask ; filiform tentacles about twelve, terminating in blunt, slightly swollen
extremities, and springing from a zone at some distance above the base of the hydranth ;
capitate tentacles about twenty, much shorter than the filiform.

GONOSOME. — MEDUSiE long ovate, borne on short peduncles, and springing from
the body of the hydranth just within the base of the verticil of filiform tentacles ;
marginal tentacles in the form of four small tubercles.

Coluiir of liyclrocaulus brown, becoming paler at the apices of the branches ; thick basal

PENNARIA CAVOLINII.

365

portion of hydraiith white, speckled with brown; tliiu apical [loilioii white; gonophorcs \rdh
rose red. (Cavoliui.)

Development of Guiwiiome. — Observed during the siunincr (Cavoliui) and autnnni (Kolliker).

Habitat. — Attached to rocks in the sea.

Bathymetrical distribution. — Zone of Cystoseira or Lauiinai'ian zone.

Localities. — Bay of Naples, Cuvolini; Messina, Kolliker. Gulf of Genoa, Trinchesi.

Fig. so.

Though the specific name of disticha assigned by Goldfuss to Cavolini's hydroid may lay
claim to priority, I have not hesitated to retain that of Cavolinii, by which Ehrcnbcrg designated
the species ; not only because this
name has already obtained accep-
tance, but because Goldfuss's name
is manifestly faulty, for, while claim-
ing to possess a specific significance,
it is founded on a character which
cannot be regarded as specifically
distinctive. And in this I feci my-
self all the more justified, as the
name of one of the most original
and accurate observers of the Hy-
DRoiDA thus remains connected with
an important hydroid form, which
he was the first to make known to
the zoologist.

The species of Pennaria are
rendered very remarkable by the
regular distichous disposition of their
primary branches, conferring on the
hydrosoma that symmetrical plume-
like form which, though the usual
condition among the Sertularians, is
quite exceptional among the Tubu-
larians.

I have never had an oppor-
tunity of seeing a living example of
this fine hydroid, of which there has
been no complete original description
since that of Cavolini, who found

the species growing upon submarine rocks in the Bay of Naples. I am indebted, however, to
Professor Trinchesi, of the University of Genoa, for a well-preserved spirit specimen of what I
have no doubt is Cavolini's hydroid. It was obtained in the Gulf of Genoa, and has enabled me,
by supplementing the description left us by Cavolini, to frame the diagnosis just given. It has
also afforded me the subject from which the accompanying woodcut has seen made.

Cavolini saw and described the medusiform gonophores, with their umbrella, radiating canals,

Fragment of a colony of Pennaria CavoVuw
preserved in spirits. Magnified about e

from a specinie
jlit diameters.

866 PENNARIA GIBBOSA.

and mamiljrium, tlioiigli lie overlooked their relation to free medusa;, and thus failed to recog-
nise the importance of an observation which nearly half a centiuy later became an essential
element in all true conceptions of hydroid morphology.

The filiform tentacles are very remarkable in the possession of blunt swollen extremities.
This condition did not escape Cavolini, who has represented it in his figure, and it was very
distinct in Professor Trinchesi's specimen. It is plainly a persistent feature indicating an
approach to true capitate terminations, and must not be confounded with the temporary enlarge-
ment frequently noticed in other hydroids as the result of certain states of contraction. It is here,
on the contrary, connected with a difference of structure, the enlarged e.xtremity being loaded
with minute thread-cells as in a true capitulum. A very similar condition is presented by the
proximal circlet of tentacles in the Actinulaj of TiihuJaria.

The body of the hydranlh is borne on a little disc, which intervenes between it and the
summit of the supporting ramulus, whose diameter the disc but slightly surpasses.

Though the gonophores of Pennaria Cavolinii are truly phanerocodonic, it is probable that
they never become free. Those described by Cavolini were female, and were seen to be loaded
with ova while still attached to the trophosome, and the same was the case in the specimen
examined by myself. The codonostome is very much contracted, and the ova appear to be
retained within the cavity of the umbrella for some time after their escape from the walls of
the manubrium. Indeed, it would seem from Cavolini's figure that they undergo their change
into planuke before liberation from the medusa. The male gonophore has since been described
by Kolliker,^ from specimens obtained at Messina. According to him, the codonostome in these
is like that of the female provided with marginal tentacles in the form of four short lobes. No
notice is taken of these processes by Cavolini ; they were, however, very evident in Prof.
Trinchesi's specimen, which, like that figured by Cavolini, was a female, and there is little doubt
that the rudimental tentacles have been overlooked by the Neapolitan zoologist.

*^* 3. Penxaki.a. GIBBOSA, Agassiz.

Pennaria gibbosa, — Agassiz, Contr. Nat. Hist. U.S., vol. iii, pi. xv, figs. 1 and 2, and vol. iv,
p. 378.

TROPHOSOME. — Hydkocaultjs attaining a height of at least four inches ; main
stem slightly zig-zag, and with a long and gentle curve from its base to its summit,
branches given off alternately at regidar intervals from each side with a distichous
arrangement ; main stem annulated at its base, and with a few annulations in each
interval between the branches ; primary branches annulated at their base and summit,
and with a few annulations at the distal side of the point of origin of each ultimate
ramulus. Hybrantus borne on the summits of very short ultimate ramuli, and on

1 'Zeltsclu-. f. Wiss. Zool.,' 1853, p. 303.

PENNARIA GIBBOSA. 367

those of the main stem and primary branches, the ultimate ramuli given off
at regular intervals from the distal side of the branch, each ramulus increasing in
thickness from its base to its summit, and annulated in its entire length ; the
hydranths which terminate the main stem and primary branches larger than the
others ; body of hydranth bulging obliquely on one side, very tumid at the base, and
thence rapidly thinning away towards the mouth in the form of a flask ; filiform
tentacles twelve, springing in a verticil from the extreme base of the hydranth ;
capitate ; tentacles much shorter than the filiform, about thirty-two in the matm'e
hydranlli.

GONOSOME.- — GoxopiiORES springing just within the verticil of filiform tentacles,
each borne on a short peduncle. Medus.e oval oblong, with four globular papillate
marginal tentacles. — Ai/assh.

Habitat. — Growing on submarine objects.
BatUymetrkal distriljution. — Lamiiiarian zone ?
Locality. — Gulf of Florida, Agassiz.

I have never seen a specimen of this liydroid, and have compiled my diagnosis from tne
description published by Agassiz, who has, moreover, given us a beautiful figure of it, and there
can be no doubt that the characters which he assigns to his hydroid must separate it from the
European species. It will be seen that while in Gavolini's hydroid the long filiform tentacles
spring from a zone, separated by some distance from the summit of the supporting ramulus, they
are described as springing from the extreme base of the hydranth in the American species.
Agassiz refers to the termination of these tentacles in slightly globular lips, but no indication of
this is given in his figure, so that it can scarcely be so well marked as in P. Cavolinii. It will
be further noticed that while in P. Cavolinii the ultimate hydranth-bearing ramuli are nearly
cylindrical, being but a very little contracted at their origin ; in the American species they rapidly
increase in thickness from the base to the summit.

A remarkable want of symmetry is described by Agassiz as existing in the hydranths of his
P. (jihhosa, for he tells us that these " bulge to such an extent on the side facing towards the
main stem as to render them strongly gibbous ;" a feature, however, which is scarcely repre-
sented in his figure.

The oldest gonophorcs observed by Agassiz had four marginal, rudimental, knob-like
tentacles, but as the generative elements had not made their appearance we have no evidence of
the sex of the gonophorcs. On each of the radiating canals, near their distal extremities, Agassiz
describes a fusiform swelling, which he regards — though not without some hesitation — as genera-
tive sacs like those of the Campanuiarian medusae. There can be no doubt that these bodies are
nothincr of the kind ; what they really are, however, it is impossible, with our present knowledge
of them, to say. A comparison with the enigmatical swellings already described on the radiating
canals of Slabberia, and which have also been erroneously taken for generative sacs, will suggest
itself.

368 HALOCORDYLE.

HALOCOEDYLE, AUmm.

Name. — From aXc, the sea, and KopcvX>i, a club; in allusion to the club-like form of the bydrauths.

Globiceps, — Arjres.
EucoRYNE, — Leidy.

TROPHOSOME.— Htdkopiitton composed of a symmetrically ramified iitdro-
CAULTJS rooted by a creeping filiform htduohhiza, tlie wliole invested with a cliitinous
PERISAEC. Htdranths flask-sliapcd, with the filiform tentacles forming a single
verticil at the base of the hydranth, the capitate much shorter than the filiform,
arranged in one or more distinct verticils towards the distal extremity of the hydranth.

GONOSOME. — GoNOPHORES phanerocodonic, developed between the filiform
tentacles and the proximal verticil of the capitate tentacles. Umbrella deeply
ovate, manubrium large, destitute of oral apendages ; marginal tentacles four, rudi-
mental; ocelli absent.

For the institution of the present genus we are indebted to Ayres, who founded it on a
hydroid discovered by him upon the eastern shores of North America. To his new genus he
gave the name of Globicejjs.

Shortly afterwards, without any knowledge of the previous observations of Ayres, the genus
was defined by Leidy, under the name of Eucoryne, with an excellent figure of the only species
as yet known. As Agassiz, however, has pointed out, both these names had already been appro-
priated, Glohicejis being in use for a genus of hyraenopterous, and Eucoryne for one of coleopterous
insects. There is, therefore, no alternative but to employ some other generic designation for
Ayres' and Leidy's hydroid, and as Agassiz proposes none, the name of HalocorJyle has here
been substituted.

HalocorJyle has close affinities with Pennaria, from which, however, it is at once separated
by the verticillate arrangement of its capitate tentacles, which in Pennaria, instead of being
verticillate, are scattered on the body of the hydranth.

HALOCORDYLE TIARELLA. 369

*;j(.* IIalocobdtle tiakella, Aijres.

Globiceps TIARELLA, — Ayves, in Proc. Bost. Soc. Nat. Hist., 1852, vol. iv, p. 193.
EucoRYNE ELEGANs, — Leidy , Marine Invert. Fauna of Rhode Island and New Jersey, 185,5,

p. 4, pi. X, figs. 1 — 5.
Pennaria TIARELLA, — M'Crafhj, Gymnophtlial. Medusa;, Proc. Elliott Soc. Nat. Hist., vol. i,

1857, p. 153. Alex. Ayassiz, IJlustr. Cntal., p. 187, figs.

311—315.

TROPHOSOME. — Hydrocaulus attainiug a height of four or five inches, main
stem nearly straiglit, alternately branched, branches distichous, sending off short,
sim])le, or slightly branched ramuli from their distal side ; main stem and branches
annulated at intervals, the ultimate ramuli annulated at their origin and termination.
Hydranths with about twelve fihform tentacles, in two verticils, having about six
tentacles in each.

GONOSOME. — GoNOPHORES on short peduncles, developed just within the verticil
of filiform tentacles. Umbrella deep and narrow, marginal tentacles in the form of
four short conical processes.

Colour. — Main stem and primary branches shining black, ultimate ramuli ochre-yellow ;
body of hydranth between the filiform and capitate sets of tentacles pink, encircled by two red
bands, the thin distal portion of the body translucent white. j\Iedus£e with the external surface
of the umbrella marked by four longitudinal red bands, each band lying directly over a radiating
canal; manubrium deep opaque red.

Development of Gonosome. — Observed in August.

Habitat. — Attached to Fuci shells, &c., in the sea.

Bathymetrical distribution. — Laminarian zone.

Locality. — Atlantic shores of North America, Ayres, Leidy, M'Crady, and A. Agassiz.

Leidy has published an excellent description and figure of this beautiful hydroid. I have
never seen a specimen of it, and have availed myself of Leidy 's memoir in drawing up the
diagnosis here given. A very complete description of the medusa is also given by M'Crady,
who, however, refers the species to the genus Pennaria, in which he is followed by A. Agassiz,
■who has also fully described the medusa and given us figures of it and of the hydranth. For
reasons already stated I cannot agree with M'Crady and A. Agassiz in referring it to Pennaria ;
I agree, on the other hand, with Professor L. Agassiz* in believing that it must constitute a

^ 'Contr. Nat. Hist. U.S.,' vol. iv, p. 28].

370 STAURIDIUM.

distinct genus. Indeed, the principles which have guided us in the definition of genera among
the Hydroida would not justify the association of the present hydroid, in a single genus, with
the " Sertidaria pennaria" of Cavolini.

As in Pennaria CavoKnii, the generative elements are not only developed in the walls of
the manubrium, but the ova may even attain the condition of planulae before the liberation of
the medusa. M'Crady, however, has seen the medusa become free before the escape of the
planulae.

STAUEIDIUM, JDiijardin.

Name. — From aTav^uq, a cross ; in allusion to the cross-like disposition of the tentacles in

the hydranth.

TROPHOSOME. — Htdrophyton consisting of a simple or irregularly branched
nTDROCAULUs arising from a creeping filiform nTBROUHizA, the whole invested with a
PEEiSAEC. Hydranths clavate, with one or more verticils of capitate tentacles, and
one verticil of filiform tentacles, the tentacles in each being four in number and
disposed in a cross ; the verticil of filiform tentacles placed at the proximal side of
the others.

GONOSOME. — Planoblasts developed from the body of the hydranth. Umbrella
deep bell-shaped ; manubrium with the mouth simple ; marginal tentacles four,
nodulated with clusters of thread-cells, and having a distinct ocellus on the basal bulb.

The name Slauridium is the same as " Staiir'uUe" of Dujardin, only with its termination
altered so as to adapt it to the ordinary form of zoological nomenclature, a form in which
Dujardin's name has been used by subsequent writers, as Krohn, Gegenbaur, and Wright.

Dujardin and the writers who have followed him have given this name to a hydroid whose
trophosome is distinguished by the characters just enumerated ; but as it has been shown by
Hincks that this form of trophosome may have tw^o very different forms of gonosome,* it is
necessary to break up Dujardin's genus into two, one of which may retain his original name for
the trophosome, while to the other we may give the name of Cladonema, that employed liy
Dujardin for the only form of medusa whicli he succeeded in tracing to a Stauridioid
trophosome.

We are indebted to Ilincks for a description of the medusa of the present genus. He has
called attention to the important morphological fact that it is in every respect identical with that
of the very different hydroid, Sj/ncorj/ne exiutiur

1 Hincks, in ' Ann. Nat. Hist.' for Dec, 180,3.
" Loc. cit.

STAURIDIUM PRODUCTUM. 371

The filiform tentacles arc less contractile and somewhat more rigid than is usually the case
with the tentacles of the Hydroida. This character they possess in common with the similarly
formed tentacles in the hydranth of Chuhnema. Hincks names them " false tentacles," but their
structure docs not essentially differ from that of the filiform tentacles of most other marine
hydroids.

Stauridium peoductum, Wr'ujht.

Plate XVII, figs. 11 and 12.

Stauridia producta, — Sirethill Wright, in Proc. Roy. Phys. Soc. of Edin. for April, 1857,

pi. xix, figs. 7 — 9.
Stauridium PRODUCTUM, — Hhicks, Brit. Hydr. Zooph., p. 58, pi. xii, fig. 1.

TROPHOSOME. — IIxDiioPHYXON attaining a heiglit of about a quarter of an
inch ; HTDEOCAULUs Simple, or with a few irregularly disposed branches, shorter than
the hydranth, and springing at intervals from a creeping filiform htdrouhiza.
HroEANTHS cylindrical, the capitate tentacles in three verticils, the filiform tentacles
tapering to a blunt point.

GONOSOME. — GoNOPiiORES developed at " the base of the lower capitate
tentacles ;" umbrella of medusa set with scattered thread-cells, marginal tentacles with
the terminal cluster of thread-cells larger than the others.

Colour. — Hydranth white, perisarc pale yellowish-brown ; manubrium and marginal bulbs
of medusa rose coloured.

Development of Gonosome. — Summer.

Habitat. — Rock pools.

Batliymetrical distribution. — Literal zone.

Localities. — Firth of Forth, Dr. S. Wright ; Ilfracombe, Rev. T. Hincks ; Penzance,
G. J. A.

This elegant little hydroid is rendered very remarkable by the symmetrical disposition of
its tentacles in crucial verticils, situated at nearly equal distances from one another along the
cyhndrical body of the hydranth. I met with it at Penzance during the autumn in considerable
abundance, attached to the bottom of shallow rock pools, a little above low-water mark, but
neither my specimens nor those originally described by Dr. Wright carried gonophores, while

48

372 YORTICLAYA HUMILIS.

Mr. Hincks's specimens, oLtained during the summer in the north of Devonshire, were furnished
with them.^

Wrio-ht and Hincks describe the fihform tentacles which compose the proximal verticil as
directed away from the hydrorhiza. In my specimens they always pointed towards the
hydrorhiza — a difference, however, which may not be constant, and cannot in itself be regarded as
specific.

VOETICLAVA, Alder.

Name. — From vortex, a whorl, and clava, a genus of hydroids ; so called from the verticillate

disposition of the tentacles.

TROPHOSOME. — Htduocatjlus simple, destitute of a conspicuous peeisaec ;
HTDKOEiiiz.A. a filiform stolon. Htdkaktii claviform, with two verticils of tentacles ;
tentacles composing the proximal verticil filiform, those composing the distal verticil
shorter and capitate.

GONOSOME. — Unknown.

The genus Vorticlava was founded by Alder for a small hydroid, wliich he discovered in a
rock-pool on the coast of Northumberland. It seems to be a well-constituted genus, but as
nothing is yet known of the gonosome, we are compelled to regard it as, to a certain extent,
provisional. In the form of the hydranth, with its two dissimilar sets of tentacles, there is a
strong affinity between Vorticlava and the genera Acharadrimn , Heterodephanus, Stauridium,
Cladonema, Hahcordi/le, and Pennaria. The absence of a distinct perisarc, however, separates
Vorticlava from all these genera except Heterostephanus.

1. Vorticlava eumilis. Aider.

Vorticlava humilis, — Alder, Catal. Zoopli., p. 10, pi. i, figs. 1 — 3. Hincks, Brit. Hydr.
Zooph., p. 132, pi. xxiii, fig. 1.

TROPHOSOME. — Hydrosome attaining a height of about two tenths of an inch,

' Thougli without the gonopliores, the identification cannot be cousidered as absolute, the
resemblance of the trophosomes is so close as to justify our assuming, ^^■ith a certain provisional reserve,
an identity of species.

VORTICLAVA PROTEUS. 373

gradually tapering from its attached to its distal end. Htdranth witli five distal
and ten proximal tentacles, the proximal tentacles about three times the lengtli of
the distal.

GONOSOME. —Unknown.

Colour. — White.

Habited. — In a rock pool attached to Coralhna otficinalii.
BafJii/mefrical Distribution. — Litoral zone.

Localities. — Coast of Northumberland, Mr. Alder; Felixstowe, on the Suffolk coast,
Mr. Busk.

Though Mr. Alder regards the solitary condition of the single specimen on which he
founded hi.s genus Vortidava as a permanent character, Mr. Hiucks believes that this condition
indicates only an immature state of the hydroid. Hincks founds this view on a specimen of an
undetermined species of Vorticlava, which he obtained on the southern coast of Devonshire, and
which consists of two hydranths, one fully grown, the other immature, and both nnited by a
common adherent base.

I have never seen a specimen of Vortidava Iiuniilis. The diagnosis given above is compiled
from Mr. Alder's description of it, supplemented by some observations of Hincks.

2. Vorticlava peoteus, Slrethill Wright.

Vorticlava proteus, — Wriytit, in Quart. Joiirn. Micr. Sci., vol. iii, n. s., p. 5, pi. v, figs.

1 — 6. Hindis, Brit. Hydr. Zooph., p. 133, pi. xxiii,
fig. 3.

TROPHOSOME. — Htdrocaulus when extended cylindrical. Hydranths with
five distal and nine proximal tentacles.

GONOSOME. —Not known.

Habitat. — Attached to stones in the sea.

Batliymetrical Distribution. — Coralline zone.

Locality.— W\Q "Pluke Hole," Firth of Forth ; Dr. Strethill Wright.

Under the name of Vortidava proteus, Dr. Strethill Wright has described a hydroid which
he regards as distinct from Alder's Vortidava kumilis. In Dr. Wright's description, however,
which can hardly be regarded as a specific diagnosis, it is difficult to recognise any character
which can be accepted as pointing to a specific distinction from Vortidava humilis, for the great
mutability of form which he assigns to his hydroid may equally belong to the latter, though no
allusion is made to it in Mr. Alder's description. The slender cylindrical fonn of the stem when

374 HETEROSTEPHANUS.

fully extended contrasts with the relatively thick tapering stem of Vorticlava humilis ; and this,
with our present knowledge of Vorticlava proteus, wonld seem to be the only reliable character
for specific distinction. Under these circumstances it is with some hesitation that I retain the
Vorticlava jjroteus as an established species. Dr. Wright gives a series of figures showing his
hydroid in various states of contraction, and it woidd appear from these that the contractibility
and consequent mutability of form is possessed no less by the hydrocaulus than by the hydranth.
He informs us also that the hydroid "has the power of changing its place," a curious and
important observation.

Wright has not been more fortunate in meeting with the gonosome of Vorticlava proteus
than Alder has been in meeting with that of Vorticlava humilis.

HETEEOSTEPHANIJS, Allman.

I^ame. — From 'InpoQ, dissimilar, and aTi(j,m'o<:, a wreath ; in allusion to the dissimilarity of
form between the two verticils of tentacles in the hydranth.

TROPHOSOME. — Htduocaulus simple and solitary, destitute of peuisarc.
Htduanths with two verticils of tentacles, a proximal and a distal ; the tentacles
composing the proximal verticil filiform, those composing the distal verticil shorter
and capitate.

GONOSOME. — Planoblasts borne on peduncles which arise between the two
tentacular verticils (?). Umbrella in the form of a shallow bell, with one large
marginal tentacle and three rudimental ones.

The genus Heterosteplianus is constructed for the Corymorpha annulicor/iis of Sars, which is
certainly not a Cori/morpha, as, indeed, Sars himself suspects. It may possibly belong to the
genus Vorticlava, as already suggested by Hincks ; but, from our present knowledge of it, and
the absence of all knowledge of the gonosome in Vorticlava, we should not yet be justified in
associating the two forms in a common genus. I had originally assigned to this genus the name
of Heteractis ('Ann. Nat. Hist.' for May, 1864), overlooking the fact that Heteractis had been
already appropriated as a generic name by the botanist, but soon afterwards becoming aware of
the inadmissibility of this name, I changed it to that of Heterostephanus.

ACHARADRIA. 875

*^* Heterostephanus annulicok^^is, Sars, sp.

CoRYMOKPHA ? ANNUncoRNis, — Stirs, Forliaucll. i vid. Selsk i Christiania, 1859; trans.

in Wiegm. Arch., 18G0, p. 311 ; trans, in Ann. Nat.

Hist, for Nov., 1861.
Heteractis annulicornis, — Allman, in Ann. Nat. Hist, for May, 1861.

TROPHOSOME. — Htduosome attaiuing a height of two thirds of an inch.
HxDRANTH with the proximal tentacles twenty in number and " annulated ;" the
distal tentacles eight to ten, very short.

GONOSOME. — Peduncles of the gonophoees very short.

Habitat. — On a muddy bottom in the sea.
Bathymetrical Distribution. — Coralline zone.
Locality. — Floroe, coast of Norway ; Sars.

In referring the present species to the genus Corymorpha, Sars expresses a doubt as to the
correctness of this association, but ventures no opinion as to an alternative. It is not easy to say
what may be the character which Sars desires to express by the term " annulated," which he
applies to the tentacles composing the distal verticil. He found the species on only one occasion,
when he obtained two specimens at a depth of from thirty to forty fathoms.

ACHAEADEIA, StretUll WrigU.

TROPHOSOME. — Hydrocaulus simple or irregularly branched, with a well-
developed PEEiSAEC. Htdeanths with two verticils of tentacles, those composing
the proximal verticil filiform, those composing the distal verticil shorter and capitate.

GONOSOME.— Unknown.

If a perisarc were developed in Vortidava this genus would become Acharadria, so far at
least as the trophosome is concerned, for we know nothing as yet of the gonosome. In Achar-
adria, moreover, the only descril^ed species is branched, while Vortidava is simple.

The affinity of Acharadria to Pennaria and Halocordyle is too obvious to be overlooked.

376 ACIIARADRIA LARYNX.

Indeed, the hydranths of Acharadria and Halocordyle with their verticils of filiform and capitate
tentacles present no difference of generic value, and in the absence of all knowledge of the
gonosome in Acharadria, we can find nothing but the symmetrical ramification of Halocordyle to
contrast with the simple or irregularly i-amified hydrocaulus of Acharadria as grounds of generic
distinction. The whole habit of Acharadria, however, is widely different from that of Halo-
cordyle, and it is probable that if the gonosome of Acharadria were known, it would be found
to differ no less than the trophosome does from that of Halocordyle. It is possible, however,
that Acharadria is but the immature state of some already described form of pennaridan
hydroids.

AciiAKADRiA LARYNX, StrelMll Wright.

Woodcut, fig. 81.

Acharadria larynx, — Strethill IVright, in Micr. Journ., vol. iii, n. s., p. 50, pi. v,

figs. 7, 8. Hincks, Brit. Hydr. Zooph., p. 131,
pi. xxiii, fig. 3.

TROPHOSOME. — Htdrosome attaining a height of ahout a quarter of an inch ;
HYDROCAULUS " Spirally twisted." Hydranths with from four to twelve tentacles in
the proximal verticil, and from two to eight in the distal.

GONOSOME.— Unknown,

Colour. — Hydranths pale orange.
Habitat. — On stones in the sea.
JBathymetrical disfridutiofi.—hammavian zone (?).
Locality. — Ilfracombe ; Dr. S. Wright.

The little hydroid on which Dr. Wright founded his genus Acharadria was discovered by
him on stones along with Caryophyllia Smifhii, but from what part of the very wide bathy-
metrical range of the Caryophyllia his specimens were procured he does not tell us. He informs
us that the Acharadria resembles Tiibularia larynx in habit.

I am indebted to Mr. Rotch for an opportunity of examining a little solitary hydroid which
had made its appearance in his aquarium, and which has afforded me the subject of the annexed
woodcut (fig. 81). It corresponds in all characters of generic value with the Acharadrium
larynx as described by Wright. Prom AVright's hydroid, however, it differs in its absolutely
simple stem, and in the absence of the spiral twisting given as a character of A. larynx. The fili-
form tentacles were ten in number, and the capitate tentacles five. The former possessed con-

ACUALIS.

377

Fig. 81.

SKlcral)le contractility, and niiglit at one time be seen thrown forward (r) towards the mouth, and at
another thrown back {i) so as to expose the whole of that
part of the hydranth-bodj which lay above them. The very
delicate transparent perisarc which invested the stem be-
came, just below the hydranth, so thin as to be with
difficulty detected ; here the stem w as irritable and con-
tractile, and the hydranth would frequently be seen to
incline from side to side, or would assume a nodding
attitude, as represented m the right hand figure (c).
The hydranth was of pale vermilion colour and, the
stem nearly colourless. No trace of a gonosome was
present.

Though this little hydroid differs from Acliaradria
larynx in characters which may be deemed of specific
value, I am not disposed to assign to it the rank of a
distinct species. It has all the appearance of an immature
form, and I regard it as the young state of A. larynx.
Indeed the mature state of Acliaradria has probably not
yet been seen.

The aquarium in which it occm'red was stocked bv o am na arym (yonng s^mmeii).

1 -J rt, Natural size : 6, magnihed; c, hyaranth and

Mr. Rotch from the shores of the Channel Islands. distal extremity of stem ; the hydranth has

here assumed a droopmg posture on its stem.

ACAULIS, St'mpson.

Name. — From a negative, and /oauXoc, a stalk, in allusion to the supposed absence of a

hydrocaulus.

TROPHOSOME. — Hxdeocaultjs unknown. Hydranth sub-cylindrical; ten-
tacles of the proximal set long, filiform, disposed in a single verticil near the
proximal extremity of the hydranth ; those of the distal set, short, capitate, scattered
upon tlie body of the hydranth.

GONOSOME. — GoNOPnoRES sessile, scattered, springing from the body of the
hydranth, between the proximal and distal sets of tentacles ; form of medusa unknown.

Our ignorance of the form of the gonophore will not allow ns to assign the genus Acaidis
to any special family otherwise than provisionally, and it is only with such reservation that it is
here placed in the Fennaridce}

' By au oversight Acaidis is placed under the family of Corymorphidce in page 240.

378 ACAULIS PRIMARIUS.

In the above diagnosis I have confined myself to one of the forms described by Stirapson, as
two different states of the same hydroid, and I have done so because I feel sure that Stimpson
had two entirely different hydroids under examination when he believed that he had only two
different phases of a single one.

After describing an animal with the characters given above, and which he found floating free
in the open sea, Stimpson goes on to say : — " At a subsequent time I met with several of these
animals, which presented a diff'erent appearance. The tentaculae were larger, especially in the
region of the mouth, at the now blunt extremity of the body, and the medusa-buds were in an
advanced state of development, soon to become free-swimming individuals. The inferior appen-
dages had disappeared, and the body was firmly attached by a broad base, and bore much resem-
blance to one of the ordinary Cori/itulce deprived of its stalk. In strong contractions it assumed a
shape resembling that of an hour-glass. The length of the animal in this latter stage was half an
inch, the breadth two tenths. In the earlier stage the dimensions were one half these.

" It was dredged in the Laminarian zone, from five to fifteen fathoms, attached to various
Rhodosperms, as Ptilota, Chondrus, and Hhodymena"

The animal thus described is a Suncorifne-XAsS, hydroid with undeveloped stem, and has
assuredly nothing to do with the Tuhidaria-\'k% free hydranth previously described. If I am
correct in this supposition, the Syncorjjne-\k.Q fixed hydroid presents us with a distinct generic
form, characterised by its sessile hydranths with scattered capitate tentacula and phanerocodonic
gonophores. It is not improbably identical with the genus IlaJocharis of Agassiz ('Contr. Nat.
Hist./ N. S., vol. iv, p. 239, pi. xx, fig. 10), a genus from which he afterwards withdrew the
name of Halocharis in favour of M'Crady's name of Cori/nites. (See the description of Conjnifes
given above.)

Stimpson found the animal for which he constituted his genus Acaidis in the condition of a
free floating hydranth, and derived fi'om this fact the leading character of the genus. I am
strongly of opinion, however, that, with the similar condition attributed by M'Crady to
Nemopsis, this free state is only accidental, and that the floating hydranth, with its gonophores,
had been detached from a stalk in a way we know to be so common in Tulularia. I have
therefore not included the freedom of the trophosome as a character in the diagnosis.

*^* AcAULis puiMAEius, Stimpson.

AcAULis PRi.MAKius, — Stimpson, Marine luvertebrata of Grand INIanan, pi. i, fig. 4, in Smith-
sonian Contributions, vol. vi, 1854, Agassi::, Contr. Nat.
Hist. U.S., vol. iv, p. 315. Allman, Ann. Nat. Hist, for
May, 1861.

TROPHOSOME. — Hydranth, with the tentacles of its proximal zone, eight in
number ; distal tentacles numerous, very small, commencing a little below the mouth,
and thence scattered over about two tliirds of tlie surface of the body.

CLADOCORYNlD^ii. 379

GONOSOME.— GoNOPiiORES thickly scattered over tlic whole of the space which
intervenes between the proximal and distal sets of tentacles.

llahitdt. — Found fioating in tlie o|)eu sea.

Localify. — Moutli of tlie Bay of Fundy, New Brunswick, W. Stinipsoii.

The tloatiiig hydranth found by Stinipson measured about half an incii from its suimnit to
its l)asc. As ah-eady said, it is highly probable that the Acaulk primariiis is nothing more tlinii
the detached hydranth of a stalked form, whose hydropliyton lias not yet been discovered.

CLADOCORYNIILE.

TROPHOSOME. — Hydrantiis with both simple and ramified capitate tentacles.
GONOSOME.— Not known.

CLADOCORYNE, TV. D. Rotch.

Name. — /cXoSoc, a branch, and Kopxnn], a chdi ; so named from the branched condition
of the tentacles in the hydranth.

TROPHOSOME. — Htdeocaulus developed, invested by a chitinous perisarc,
and rooted by a creeping filiform iitdrorhiza. Hydranths claviform, with a single
verticil of simple capitate tentacles round the mouth, and several verticils of branching
capitate tentacles on the rest of the body.

GONOSOME.— Xot known.

The genus Cladocorijne is separated from all other known hydroid genera by the ])resence of
ramified tentacles in the hydranths. Tins condition may be compared with that of Cladonewa,
in which the planoblast is distinguished by its ramified marginal tentacles.

49

380

CLADOCORYNE FLOCCOSA.

CL.iDOCOUTNE TLOCCOSA, Botcll.

Woodcut, fig. 82.

Cladocorv.ve FLOCCOSA, Rotch, in Ann. Nat. Hist, for March, 1871.

TROPHOSOME. — Htduocaulus attainini

Cladocoryne Jloccosa.
A. A colony of the natural i
-B. Hvdranth magnified.

a lieight of about half an incli,
slender, simple or sparingly branched;
PEEiSARC smooth, or slightly and
irregularly annulated. Htdkanths
long, linear, slender, tentacles very
long ; the simple tentacles forming
a verticil of four to eight imme-
diately round the mouth, the
branching tentacles forming three
or four verticils round the body,
with three or four tentacles in every
verticil, each tentacle giving off
numerous short capitate ramuli.

GONOSOME. — Not known.

Colour. — Hydranth reddissh-bi-owu,
with opaque white oral extremity, and
with its proximal end merging into the
light straw-colour of the perisarc.

Habitat. — On stones at low tides.

Balhymetrical Butribtdion. — Literal
zone.

Locality. — Herm, near Guernsey,
Mr. W. D. Rotch.

For our knowledge of this beautiful little Iiydroid we are indebted to Mr. Rotch, who
obtained it in the Channel Islands. Among the whole range of hydroid trophosonies there is not
one more remarkable. It is absolutely unique in the ramified condition of the tentacles of the
hydranth. Unfortunately no gonosome had been developed in any of the specimens collected,
and we wait with much impatience for the discovery of this important element.

The accompanying woodcut has been made from a specimen in Mr. Rotch's aquarium.

MYRIOTHELID^. 381

MYRIOTHELID^.

TROPHOSOME. — Hydrantii solitary, attached ; tentacles scattered, capitate.
Htdrocaulus not developed.

GONOSOME. — ^GoNOPnoEES fixed sporosacs borne on special processes which
spring- from the body of the hydranth.

MYEIOTHELA, Sars.

Name. — From /.wploc, numl)crless, and OijXi), a ni[)[)le ; in alhision to the great number of
nipple-like tentacles in the hydranth.

LucERNARiA, — Fubricius.
Candelabrdji, — -De Bhdnville.
Arum, — Vigors.
Spadix, — Gosse.

TROPHOSOME. — Htdraxth claviform or sub-cylindrical, springing from a broad
adherent htdrorhiza, which is invested with a peris arc ; tentacles very small,
papilliform.

GONOSOME. — Processes which support the sporosacs naked, springing from
the hydranth at the proximal side of the tentacles.

I have never had the good fortune of meeting with an example of this remarkable genus.
It was first accm-ately defined by Sars,' to whom we owe the name of Myriothela ; but, as
Mr. W. Stimpson has pointed out,'' the first example of the genus was described by Fabricius,^
whose Lucernaria pimjgia is certainly a Myriothela, probably identical with the Myriothela
arctica of Sars.

De Blainville, seeing that Fahricius's animal had no relation with Lucernaria, constituted for
it, in 1S34, his genus Candelahrum , and as this name has priority over Myriothela, it is accepted
by Agassiz as the legitimate name of the genus.

' Sars, ' Beretning om en Zoologisk Reise i Loftoen,' 1850.
- See Agassiz, ' Coiit. Nat. Hist. U. S.,' vol. iv, p. 34'1, note.
' 'Fauna Gioeulandica,' p. 313.

382 MYRIOTHELA PHRYGIA.

Though I followed Agassiz in this view myself/ more mature consideration has induced me
to return to the name of Myriothela. De Blainville deserves no credit for seeing that the
Lucernaria pliry(/ia of Fabricius was not a Liicernaria. Any zoological contemporary of his
would have done the same, and though by withdrawing this hydroid from the genus Lucernaria
he has so far rectified the error of Fabricius, he has, nevertheless, totally misunderstood its rela-
tions. He asserts of it that it " certainement n'appartient pas au type des Actinozoaires," and
he concludes his allusion to it by affirming its affinity to Sijmnculus.^ Indeed, Fabricius had a
far more correct conception of the zoological position of his animal than De Blainville had, for
besides giving a very good description of it, he expresses doubts as to its relations with Lucernaria,
and plainly recognises its hydroid affinities.

Now Sars was well acquainted with the animal, for which he formed his genus Mi/riothela.
He had a true conception of its relations, and gives a legitimate generic diagnosis of it, so that I
have no hesitation in returning to the name of Myriothela, though the laws of priority, if rigidly
enforced, might justify the suppression of Sars's name in favour of De Blainville's.

I can find no reason for regarding Fabricius's species as distinct from that of Sars, and as
no reason can be urged against the use of the specific name given to it by the famous author of
the ' Faima Grcenlandica,' I lielieve that here the law of priority must prevail, and that Fabricius's
name oi phryyia must take the place oi Arctica given by Sars.

From tiie observations of Mr. Cocks it would appear that the young of Myriothela phrygia
escape from the gonophores under a form which has much resemblance to the young of Ttibularia,*
and Mr. Alder informed me that he had himself made a similar observation. It would thus
appear that the young of Myriothela are not 2)lanuIcB but actiniilce.

Myriothela phuygia, Fahricms.

Lucernaria phrygia, — Fabrichis, Fauna Gioenlandica, p. 343.

Candelabrum phrygium, — De Blamville, Mem. d'Actinologie, p. 318. Agassix, Cunt.,

Nat. Hist. U. S., vol. iv, p. 3U.
Myriothela akctica, — Sars, Beretning otn en Zoologisk Raise i Lofoten.
Arum Cocksii, — Vigors, in Report of Royal Polytechnic Society of Cornwall, 1849. Cocks,

in ditto, 1853, p. 34, pi. iii, figs. 7 — 12.
Spauix purpurea,- — Gosse, in Ann. Nat. Hist., vol. sii, 2nd ser., 1853, p. 126 ; and Man. of

Marine Zoology, p. 19, fig. 25.
Myriothela phrygia,— iZ/wc^s, Brit. Hydr. Zooph., p. 77, pi. xi, fig. 3.

TROPHOSOME. — Hydranth attaining a lieiglit of two inches when extended,
and then nearly cyUndrical in form ; tentacles densely crowded.

^ " Genera of Hjdioida," 'Ann. Nat. Hist.' for May, 1864.
"" De Blainville, ' Actinologie,' p. 318.

^ W. P. Cocks, in ' Report of the Roy. Polytechnic Soc. of Cornwall,' 1853, p. 34, pi. iii, figs.
7 — 12 ; and in ' Ann. Nat. Hist.,' vol. xii, 1853, p. 365.

CLAVATELLID^. 383

GONOSOME. — GoNOPnoRES spherical on alternately branched processes.

Habitat. — Attached to stones in the sea.

Bathymetrieal Distribution. — Latuinarian to decpwater zone.

Localities. — Coast of Greenland, Fabricius ; coast of Norway, Sars ; Falmouth, Mr. Cocks;
coast of Devonshire, Mr. Gosse and Mr. Hincks ; Grand Manan, Atlantic shores of North
America, Mr. W. Stimpson.

If I am right in believing that all the examples of Myriotlida hitherto found belong to a
single species, the distribution of Myriotlida phryyia in the North Atlantic is a wide one, for the
species is found both upon the European and American shores.

Not having succeeded hi obtaining a specimen of Myriothela, I am unable to record any
observations of my own on this most interesting hydroid, and the account here given has been
derived from the notices of other observers, more especially of Cocks, Alder, and Hincks. Hincks
compares the processes vvliich carry the gonophores to the lalastostyles of Hydractinia. To this
view, however, the origin of the processes from the body of the hydi-anth is opposed. Their form
and situation, indeed, judging from the publislied and other figures which I have seen, would
suggest rather a comparison with the gonophore peduncles of Tiibularia and Coryniorpha.

CLAFATELLIDjE.

TROPHOSOME. — Htdranths with simple verticillate capitate tentacles.

GONOSOME. — GoNOPiioBES in the form of ambulatory medus.e, with undeveloped
umbrella and branchina: marafinal tentacles.

CLAVATELLA, ninch.

Name. — A diminutive noun, formed from clava, a club, in allusion to the form of

the hydranth.

Eleutheria, — Kro/in.

TROPHOSOME. — Htdrocaulus rudimental, springing from a creeping filiform
HTDRORHiZA, the whole invested with a chitinous perisarc. Htdranth elongated.

S84 (JLAVATELLA PROLIFERA.

with its tentacles in a single verticil, which surrounds the base of a conical
hypostoine.

GONOSOME. — GoNOPiiORES developed in clusters on branched peduncles from the
body of the hydranth. Rudimental umbrella not fitted for natation. Manubrium
short, conical, destitute of oral appendages ; radiating canals six ; marginal tentacles
six, bifurcated, the outer branch of the bifurcation terminated by a capitulum of large
thread-cells, the inner by a claviform enlargement, which carries a suctorial disc of
attachment ; an ocellus at the root of each tentacle, but no distinct marginal bulbs.

For the establisliment of the genus Chtvatella we are indebted to Hincks, who gave this
name to a hydroid trophosome, which he discovered on the south coast of England. From this
trophosome he observed tlie development of gonophores having a close resemblance to the free
medusiform body which had previously been described by Quatrefages under the name of
Eleuiheria dichotoma, but which had not been traced to a trophosome. Its chief difference from
mieutJieria consists in the fact that the inner branch of the marginal tentacle is terminated by an
apparatus for suctorial attachment, while in Eleufheria it ends in a capitulum of thread-cells like
that of the outer branch.

This difference, however, I regard as of generic value, so that notwithstanding the close
resemblance between Quatrefages's EleufJieria and the gonophore of Clavatella, the two will
scarcely admit of being associated in a common genus, however closely the trophosomes may
ultimately be found to resemble one another. Clavatella and Eletdheria must, therefore, stand
as two distinct genera, though the latter will not admit of complete definition until the discovery
of its trophosome.^

Clavatella prolifera, EiiicJcs.

Plate XVIII.

Clavatella riiOLiFERA, — Hincks, \n Ann. Nat. Hist, for Feb., 18G], pis. vii and viii,

Brit. Ilydr. Zooph., p. 73, pi. xii, figs. 2, 2a. AUma,n.
ill Brit. Assoc, Rep. for 1862, p. 100; and in Aun. Nat.
Hist, for May, 1S6L
Eleutheria (Plauoblast), — Krohn, in Wiegmanu's Archiv, 1861, p. 157, trans, in

Ann. Nat. Hist, for Jan., 1862.' FUippi, in Mem.
della Reale Academia d. Scienze di Torino, ser.
ii, tome xxiii.

TROPHOSOME. — Hydrocaulus consisting of very short cylindrical processes,

' Elentlierhi, being as yet known only by its free planoblast is not included among the genera of
the descriptive part of the present monograph.

CLAVATELLA PROLIFERA. 385

springing at intervals from the creeping filaments of the hydrorhiza. Htdranths
very much elongated, cylindrical in extension, with a dilated base, where they spring
from the summits of the rudimental hydrocaulus ; when fidly extended attaining a
height of about half an inch ; tentacles six to eight, with well-defined spherical
capitula.

GONOSOME. — GoNOPnoRES in one or two clusters, each cluster containing two
or three gonophores, which are borne on a very short, branched peduncle which
springs from the dilated base of the hydranth. Planoblast dome-shaped, with its
margin continued into six short cylindrical tentacles, which divide, with their two
branches of nearly equal length, a little to the distal side of their middle point.

Colour. — Hydranths wliite, tinged with a pale jjiiik at the distal extremity. jMedusaj with
pale reddish-yellow eiuloderni a])pcariiig through the translucent colourless ectoderm ; ocelli
red.

Development of Gonosome. — Observed during the suamier and autunui.

Habitat. — Attached to the sides of roek-pools.

BaihymetricaJ Distribution. — Litoral zone near its upper limit.

Localities. — Coast of Devonshire, Rev. T. Hincks and G. J. A. ; coasts of Cornwall, Cork,
and west of Scotland, G. J. A. ; Gulf of Genoa (free planoblast), Prof. Trinchesi ; Mediterranean
sea (free planoblast), Krohn, ]''ilippi (?).

Among the observations which have of late years so greatly advanced our knowledge of the
Htdroida, one of the most important is that of Hincks, who, as already stated, found a hydroid
trophosome giving origin to gonophore-buds, having an intimate affinity with the free hydroid
organism described by Quatrefages, some years previously, under the name of Eleutheria diclio-
toma. This discovery renders it almost certain that Quatrefage's Eleutheria is also a planoblast,
originating as a bud from some hydroid trophosome, a fact which the excellent description
given by Quatrefages had already led more than one zoologist to suspect, but which, until
Hincks's discovery, had received no further confirmation.

A free gonophore, apparently undistinguishable from that of Hincks's Clavatella prolifera, has
also been well studied by Krohn, who obtained it in the neighbourhood of Nice ; and by Fdippi,
who found what would seem to be the same organism in the marine aquariums of the Zoological
Museum of Turin ; while a nearly allied form, though apparently belonging to a different species,
has been examined by Claparede, who discovered it on the coast of Normandy.^ In none of
these, however, any more than in the original Eleutheria of Quatrefages, have the observers seen
the trophosome which has as yet been witnessed only by Hincks and by myself. For my first
opportunity of examining this remarkable little hydroid I am indebted to Mr. Hincks, who
directed me to the spot on the coast of Devonshire, where he originally discovered it. Since
then I have found the Clavatella in other localities, and have been enabled to make both tropho-
some and gonosome the subject of a careful study. (See above, p. 212.)

' I am indebted to Professor Trinchesi, of the University of Genoa, for a drawing of the free
planoblast of a Clavatella, which is probably identical with the British species.

386 CORYMOEPHIDiE.

One of the most striking features in the trophosouie of ClmatcUa prolifera is the compara-
tively great development and the elongated form of the hydranth, which, just as in Clava,
contrasts strongly with the rudimental and inconspicuous hydrocaulus. When extended its
form is that of a cylinder, slightly dilated at its proximal end, where it springs from the summit
of its supporting hydrocaulus, and at its distal end, where it carries its verticil of tentacles.
When contracted its form varies, according to the degree of contraction, from that of an elongated
cone to that of a flask.

The tentacles are usually seven or eight in number, springing in a single verticil from the
dilated summit of the hydranth, and surrounding a short bluntly conical hypostome. When
extended they slightly taper from their base towards their summit, where they terminate in a
well-defined spherical capitulum, formed by an accumulation of large thread-cells in the thickened
ectoderm.

The fully formed planoblast is very remarkable in the non-development of a free umbrella.
It is thus destitute of an organ of natation ; but, to compensate for this deficiency, the marginal
tentacles are peculiarly developed and adapted to the function of creeping, so that the free
gonophore of Clavatella, instead of being natatory, like other hydroid medusae, is truly ambulatory.
The generative elements are produced between the endoderm and ectoderm on the dorsal
or proximal side of the medusa. In all the specimens I have examined the medusa had six
marginal tentacles. The variation from this number noted by Filippi, who occasionally found
seven, is probably abnormal. The planoblast may frecpiently be seen repeating itself by buds,
which are developed in the interradial spaces of the margin.

Clavatella prolifera is a singularly beautiful little hydroid. It grows in small scattered
groups, which the eye will easily detect in the shallow reservoirs of clear water left behind by the
retiring tide, and whose gracefully bending hydranths, with their coronals of globe-tipped
tentacles, and budding clusters of medusae, render it one of the most attractive and interesting of
all the smaller tenants of the rock-pool.

It appears to be confined to a zone just below the upper limit of the range of neap tides,
and seems to prefer the smaller and shallower pools, more especially such as have their sides
overgrown with an incrustation of nullipores.

CORYMORPIIIDjE.

TROPHOSOME. — Hydrocaulus solitary, destitute of perisarc. Hydranths with
a proximal and a distal set of filiform tentacles.

GONOSOME. — GoNOPnoKES in the form of medusiform planoblasts, with four
radiating canals, and one or more simple marginal tentacles.

CORYMORPHA. 387

CORYMOEPHA, Sars (in part).

Name. — From Kopm-i), a club, and lio^ipi], lonii, in allusion to the form of the hydranth with
its stem.

TROPHOSOME. — Hydrocaulus emitting towards its proximal extremity tubular
fleshy processes ; perisahc replaced by a delicate filmy pellicle. Hydraxths flask-
shaped, abruptly distinct from the hydrocaulus ; proximal tentacles imperfectly
contractile, larger than the distal, and arranged in a single verticil near the base of
the hydranth, the distal tentacles very contractile, forming several closely approxi-
mate alternating verticils round the base of a conical hypostome.

GONOSOME. — -Plaxoblasts borne on branched peduncles which spring from the
body of the hydranth between the proximal and distal sets of tentacles, with a deep-
belled umbreUa, a well-developed simple-mouthed manubrium, and a single marginal
tentacle ; each of the radiating canals terminates at its junction with the circular
canal in a bulbous expansion without distinct ocellus ; one of these bulbs is larger
than the other, and from this alone the solitary tentacle is developed.

The genus Cori/inorjiha was instituted by Sars for the Cori/morj)ha nutans, a beautiful
hydroid discovered by the celebrated Scandinavian zoologist on the coast of Norway. Since
then several species have been admitted into the genus, but I believe that of these one only has
any claim to be oongenerically associated with the original Cori/morpha nutans.

Among the characters which especially strike us in the planoblast of Corymorplta is the
absence of symmetry, as shown in the development of only a single marginal tentacle. This
must not be confounded with that deficiency of marginal appendages which in other planoblasts
is merely the result of an immature condition. Here it is a permanent feature, having an im-
portant morphological significance, and undoubtedly entitled to a place among the essential
characters of the genus.

388 CORY.MORPHA NUTANS.

CORYMORPHA NUTANS, S(irs.
Plate XIX.

CoRYMORPHA NUTANS, — Sofs, Beskrivelser, 1835, p. 6, pi. i, fig. 3. Forbes and Goodsir,
Ann. Nat. Hist., 1840, vol. v, p. 309. Johnson, British Zooph.,
1847, p. 54, pi. vii. Sars, Forhandl. i Vid. Selsk. i Christiania,
1859 J trans, in Wiegm. Archiv, 1860, p. 3341, and in Ann.
Nat. Hist., 1861, p. 353. Hodge, in Trans. Tynesdale
Naturalist's Field Club, toI. v, p. 80, pi. ii, figs. 1 — 9. Atlman,
in Ann. Nat. Hist, for May, 1864. Hincks, Brit. Hydr. Zooph.,
p. 127, pi. xxii, fig. 2.

TROPHOSOME. — Htdrocaulus attaining a height of from two to three inclies,
and having in its tliickest part a diameter of about two lines, sub-cylindrical, usually
enlarging towards the base, and again contracting and tapering away to a blunt point at
the proximal end, the proximal pointed end bent at nearly a right angle to the rest ;
behind the short papillary projections which are emitted from the stem near its proximal
end are numerous very fine long capillary filaments ; the whole stem marked with
narrow longitudinal bands ; the pellicle which takes the place of the perisarc very thin,
filmy, and colourless, closely embracing the ectoderm in the greater part of its extent,
but towards the proximal extremity separated from it by a considerable interval, and
here forming a loose corrugated sac in which the pointed extremity of the stem is
enveloped. Htdrantiis with a proximal zone of about thirty-two long tentacles, and
a distal one of about eighty, which are very much shorter and thinner than the
proximal, and form a brush-like group, composed of six or seven closely approxi-
mated verticils.

GONOSOME. — Peduncles of gonophores fifteen to twenty in number, springing
in two alternating verticils from the body of the hydranth immediately within the
proximal zone of tentacles, and carrying the gonophores in clusters at the extremities
of their branches. Gonophores with the summit of the umbrella continued into a
short conical projection ; marginal tentacle rendered moniliform by clusters of thread-
cells.

Colour. — Hydranth light red from the mouth to the proximal set of tentacles ; stem very
pale red. Peduncles of gonophores, manubrium, bulbous terminations of radiating canals, and
nodules of marginal tentacle light red.

Bevelopuent of Gonosome. — June to September.

CORYMORPHA NUTANS. 389

Habifai. — On sandy sea bottoms.

Baflii/melrical distribution. — Coralline zone.

Localities. — Coast of Norway, Sars ; Orkney Islands, Forbes and Goodsir; Shetland Islands,
G. J. A. ; Firth of Forth, Mr. M'Fie and G. J. A. ; Coast of Nortliunibcrland, Mr. Hodge ;
Coast of Cornwall, Mr. Alder and Mr. Peach ; Isle of Man, Mr. Alder.

Corymorpha nutans is one of the most beautiful of the many beautiful acquisitions of the
dredge. When the naturalist is fortunate enough to carry his dredge over the sandy ground
inhabited by it, he will usually l)ring up numerous specimens, which may be all derived from so
limited an area as to force upon him tiie conviction that the species is gregarious in its habit,
though the trophosomes are never united so as to form composite colonies. The specimens on
being brought to the surface are always found with sand adhering to their proximal end, which
is constantly bent on the rest of the stem, and it is almost certain that while in their natural
habitat they live with this end plunged into the sandy sea bottom.

When freed from the sand, thus adhering to the lower end of the stem, and transferred into
a jar of sea water, the Curi/niorjjJia soon begins to fix itself to the bottom of the vessel, and at
the termination of about twenty-four hours its base is seen to be surrounded by a delicate web,
which closely adheres to the glass, and in a few days has spread itself over a surface of a square
inch or more in extent.

Under the microscope this web is found to be composed of a multitude of fine tubular
filaments, which are given off from the stem all round close to its lower end, and then by repeat-
edly crossing one another form an entangled web-like tissue. It is almost certain that similar
filaments existed in the hydroid while yet undisturbed beneath the sea, where they must have
served to fix it to its sandy bed, and that in the act of detaching it they had become torn off", to
be speedily renewed on the specimen being again allowed to rest in the confinement of our
jars.

There is no true perisarc, but the stem is invested by a delicate pellicle, so delicate, filmv,
and colourless indeed, that it may easily be overlooked; and when the animal is removed from
the water the stem, destitute of the support which the Hvdroida usually receive from their
firm perisarc, appears soft, flaccid, and gelatinous. Towards the proxinui! end of the stem, how-
ever, this pellicle becomes separated from the ectoderm by a considerable interval, and here
constitutes a loose filmy sac, in which this portion of the stem is enveloped, and which allows
the passage across it of the delicate filaments of adhesion already described.

By the naked eye the stem is seen to be traversed from one end to the other by uari'ow
longitudinal bands. Under a low magnifying power these bands are seen to inosculate with one
another here and there, while towards the base of the stem they become fewer and broader by
coalescence. They indicate the canals which are excavated in the endoderm of the stem, and
which thus show themselves through the more superficial tissues.

The peculiar short conical papillae which are given off from the stem near its proximal end
are arranged in regular longitudinal series, which foUow the course of the longitudinal bands, the
stem immediately over each band bearing two alternate rows.

The planoblasts when they become free are about -^i\\ of an inch in diameter. Their
solitary tentacle constitutes one of their most striking features, and has acquired a great develop-
ment even before the detachment of the medusa. It consists of a very extensile mojnliform

390 HALATRACTUS.

chord, and when extended presents the appearance of a cylindrical string with ten or twelve little
spherules distributed upon it at equal distances. The last of these spherules exactly terminates
the string, and is larger than the others. The medusa undoubtedly belongs to the form to
which Edward Forbes had given the generic name of Steenstrupia. Of this relation between
Steensfrupia and Cori/morplia Forbes himself had a suspicion ; indeed, he expresses a belief that
his Steenstrupia rubra will turn out to be the free medusa of Corj/morpha nutans.

A medusa, which I regard as the adult planoblast of Corymorjjha nutans, was obtained in the
open sea, near the spot from which the complete hydroid was dredged. It has been fully described
above (p. 211), and from the account there given, it will be seen that the changes undergone by
the medusa between the time of its liberation from the trophosome and its attainment of sexual
maturity are of little importance. It is especially to be noted that the marginal tentacle always
remains solitary.

When the animal is allowed to assume its natural position in an aquarium, Avith its proximal
extremity rooted in the sandy bottom, the stem rises vertically from the point of attachment, while
the hydranth usually droops gracefully from the opposite end, its long tentacles forming a beautiful
coronal of curved rays round its base, and the short ones forming a dense brush-like cluster
round its oral extremity. But there is no part of this beautiful hydroid which tends so forcibly
to impress the observer as the gonophores. Nestling at the base of the great coronal of tentacles
we may see them in every stage of development, from the nascent bud, in which no medusoid
structure can yet be detected, to the fully-formed medusa ready to break away from its restraining
stalk, some — as if the life of the hydroid was specially concentrated in these wonderful buds —
palpitating in rapid systole and diastole, seemingly eager for their approaching freedom ; while
others, less restless, but no less actively engaged in ministering to the welfare of their existence,
are casting their long tentacles into the water, marvellous fishing lines, loaded with deadly thread-
cells, and sensitive to the slightest touch of the passing prey.

Corpnor^jha nutans has afforded to me one of the subjects of detailed hydroid study con-
tained in the former part of this IMonograph. (See above, p. 208.)

HALATRACTUS, Allman.

Name. — From a'Ac, the sea, and arooKroc, a spindle; so called from the form of the

hydrosome.

TEOPHOSOME. — Hxdrocaulus surrounded towards its proximal extremity with
tubular fleshy processes. Htdranth abruptly distinct from the hydrocaulus ; the
proximal set of tentacles in a single verticil, and larger than the distal, which are
scattered or subvertillate round the base of a conical hypostome.

GONOSOME.— Planoblasts sessile, springing from the body of the hydranth
between the proximal and distal sets of tentacles ; umbrella at the time of liberation

HALATRACTUS NANUS. 391

bell-shaped, witli one of the four radiating canals continued into a short club-shaped
tentacle, while each of the others terminates at the margin of the umbrella in a bulb
destitute of tentacle ; manubrium long-, with a simple mouth.

The genus Ilulutradus has been instituted for the Cori/morplia nana of Alder, whose sessile
gonophores at once distinguish it from the true Corymorphas, whose gonophores are always pedun-
culated. Alder, however, in his description of this bydroid, mentions another character, which,
if established, would be one of great importance ; for he tells us that while the gonophores are
jdiauerocodonic in some individuals, they are adelocodonic in others, and he figures irrcgularly-
lobed oval bodies as tlie adelocodonic form, suggesting at the same time that this condition may
depend on a difference of sex.

The peculiarity thus indicated I cannot accept without the confirmation which the examina-
tion of a greater number of specimens may afford. The phenomenon of phanerocodonic and
adelocodonic gonophores being borne by one and the same species of hydroid rests upon no valid
evidence, though its occurrence has been maintained in more than one instance, and I am far
more inclined to regard the unsymmetrical, irregularly- lobed bodies of Alder's Corymorplia nana
in the light of misshapen monstrosities, occm'ring in an individual instance, than as a constant
and normal phenomenon ; and this view I am the more disposed to adopt as Alder's figures
convey no idea of true sporosacs.

It is also a point for further investigation to determine whether the peculiar club-shaped
form attributed by Alder to the solitary marginal tentacle of the planoblast is not the result of a
contracted condition of this organ. Alder evidently thinks not.

Halatractus nanus, Alder.

CoRYMORPH.i NANA, — Alder, Catal. Zooph. of Northumberland, Durham, p. 80, pi. vii, figs.
7 and 8; Suppl. Catal., p. 9, pi. xi. Allman, in Ann. Nat.
Hist, for IMay, 1864.. Hincks, Brit. Hydr. Zooph., p. 130,
pi. xxii, fig. 3.

TROPHOSOME. — Htdrocaultis about one half an inch in height, marked by
longitudinal opaque white bands, and tapering towards its proximal end, where it
terminates in a blunt point, which is surrounded by a loose sheath of filmy perisarc.
Htdranth with from fifteen to twenty tentacles in the proximal zone, and about
sixteen or eighteen in the distal, where they are disposed in two imperfect rows.

GONOSOME. — GoxopiiORES forming a verticil close to the proximal zone of
tentacles ; medusa with the summit of the umbrella rounded.

392 AMALTH^A.

Habitat.— {^).

Bathymetrical distrihution. — (?) .

Locality. — Coast of Noitliumbeilanc], ]\Ir. Alder.

The remarkable little hydroid to which Alder assigns the name Corymorplia nana, is by far
the smallest of the species which have been referred by authors to the genus Corymorpha. It
was obtained from the refuse of the fishing boats on the Northuinljerland coast, but we know
nothing of its habitat or of its bathymetrical distribution.

AMALTHJIA, Oscar Schmidt.

Name. — Amallkeia — a mythological name : the goat that suckled Jupiter.

TROPHOSOME. — Hydkocatjlus emitting tubular processes near its proximal
end ; PERisAuc rudimental. Htdranth abruptly distinct from the hydrocaulus ; the
proximal tentacles larger than the distal, and disposed in a single verticil near the
base of the hydranth, the distal tentacles scattered (or else multiverticillate ?).

GONOSOME. — Plange'Ists borne upon peduncles whicli spring from the body of
the hydranth between the proximal and distal sets of tentacles, having a deep bell-
shaped umbrella, and four equal marginal tentacles with bulbous bases.

The genus Amalth(ea is closely allied to Corymorpha. It was instituted by Oscar Schmidt
for a hydroid which he names Amalthaa uvifera, and which was obtained off the Island of Loppen,
in Finland.

From Corymorpha it is distinguished by the form of its gonophore, the umbrella of whicli is
provided with four equal marginal tentacles instead of the solitary tentacle of Corymorpha, a
character which is certainly of generic imporance, though Sars, not recognising it as such, places
Schmidt's hydroid in the genus Corymorpha.

Admitting the generic value of this character, two other species placed by Sars in Corymorpha
must be removed to the genus Amallhaa, which will thus include three species, all distinguished
from Corymorpha by the possession of four equal marginal tentacles in the medusa. These are
the original Amalthcea uvifera of Schmidt, the Amalthma Sarsii, and the Antalthcea Januarii.

AMALTHJEA SARSII. 393

*:j^* 1. Amalth^a uvifeba, 0. Schmidt.

Amalth.ea uviFERA, — Oscar Schmidt, Hand-Atlas der Vergleich-Anatomie, 1854, p. 13,

pi. ix, figs. 2 and 2a. Allman, in Ann. Nat. Hist,
for May, 1864.
CoRVJioRPHA uviFERA, — SuTS, Foiliandl. i Vid. Selsk i Cliristiania, 1859; trans, in Wiegmann's
Archiv, 1860, p. 341 ; and in Ann. Nat. Hist, for Nov.,
1861.

TROPHOSOME. — Hydrocaulus scarcely exceeding an inch in height. Ht-
DKANTH chtb-shaped.

GONOSOME. — GoNOPHORE-peduncles much branched. Medusa with rounded
summit and four very Large marginal bulbs.

General colour of tropliosome red.

Habitat. — Sandy sea bottoms.

HatkymetricaJ distribiitmi. — Coralline zone.

Locality. — Off the Island of Loppen, Finland, 0. Schmidt.

The only information we possess regarding this species is derived from Oscar Schmidt's
figures, accompanied by a few explanatory words, in his ' Hand-Atlas der Vergleichenden
Anatomic.' These figures, however, leave ranch to be desired, while their deficiency are in no
respect supplied by the explanation, which, with a short note referring to the habitat of the species,
is the only account we have of the hydroid for which Schmidt constituted his genus AmalfhcEa.

The medusa is figured without any marginal tentacles proceeding from the marginal bulbs,
a condition which we must regard as indicating an immature state of the gonosome in the speci-
mens from which the drawings were made.

It is from such materials that the above description has been compiled, a description which
cannot be regarded as otherwise than deficient in many points which are essential for a satis-
factory diacrnosis.

*^* 2. Amalth^a Saksii, Skenstriq).

CoRYMORPHA NUTANS, — Savs, ReisB i Lofoten eg Finmarken nyt Magazia for Naturvidensk,
1850, vol. vi, p. 135.

CoRYMORPHA Sarsii, — Stt^etistrup, !Meddel fra den Naturhist. Foren. i Kjobenh., 1854, p. 48.

Sars, Forhaudl. i Yid. Selsk. i Cliristiania, 1859; trans, in
AViegni. Arcliiv, 1860; and in Ami. Nat. Hist., Nov.,
1861.

Amai.th^ea Sarsii, — Allman, in Ann. Nat. Hist, for May, 1864.

TROPHOSOME. — HTDEOCArLUS attaining a height of from two to three inches.

394 AMALTH^A JANUARII.

Htdeanth with the tentacles of the proximal zone very long, thirty to forty in
numher, those of the distal zone very short and numerous.

GONOSOME. — Medusiferous peduncles divided at the apex. Medusa having an
elongated bell-shaped form, with rounded summit.

Habitat. — Oil a muddy sandy sea bottom.

Bathpnetrical distribution. — Deeper parts of Coralline zone.

Locality. — Vestfjorden, near the Lofoden Islands, Sars.

The present species, which was discovered by Sars in 1849, has hitherto been fomid in only
one locality, where, however, as Sars informs us, it is very abundant, and is often taken up with
its long tentacles entangled in the meshes of the dredge.

It differs from Amalthaa uvifera by its greater size, by the longer proximal tentacles of its
hydranth, and the more numerous distal ones, and by its shorter and more sparingly branched
medusiferous peduncles.

*^'^ 3. AMAXTHiE.i Jantiaeii, Stecnsirwp.

CoRYMORPHA Januarii, — Steenstnq), Vidensk. Meddel fra d. Naturh. Foren. i Kjobenh., 1854,

p. 46, Sars, Forhandl. i Vid. Selsk. i Christiania, 1859 ;
trans, in Wiegm. Archiv, 1860; and in Ann. Nat. Hist.,
1861.

AsiALTH/F.A. Januarii, — AUmun, in Ann. Nat. Hist., May, 1864.

TROPHOSOME. — Htdeocatilus attaining a height of six inches. Hxdranths
with the tentacles of the proximal zone very long, above eighty in number.

GONOSOME. — Peduncles of gonophores branched, about forty in number.
Planoblast with the height of the umbrella about twice its width.

Colour. — Pale reddish.

Locality. — Rio Janeiro, Steenstrup.

The Amalthcea Januarii has been made known by Steenstrup, who described it from a single
specimen obtained in the harbour of Rio Janeiro, and sent in spirits to the Zoological Museum
of the University of Copenhagen. In the thickness of its stem, and in the size of its hydranth, it
surpasses all known hydroids, even the great Tubidaria regalis of the Spitzbergen seas, and the two
hydroids with largest hydranths which have as yet come to the notice of the zoologist have thus
been obtained, one from the icy ocean, the other from the regions within the Tropics.

MONOCAULTD^. 395

The stem oi xlmaUhcea Januarii measures six inclies in lieiglit, and one third of an inch in
thickness, wiiiie the hydranth is nearly two thirds of an incli in height, and one half of an inch
in diameter, having the tentacles of its proximal circlet nearly two inches in length, and about
eighty in number. The specimen appears to have been somewhat injured at the distal end of the
hydranth, and only a few of the oral tentacles have been here preserved, so that nothing can be
asserted as to the number and disi)Osition of these.

The stem is marked by longitudinal deeper coloured bands, which are mucli more numerous
than the corresponding bands of Corjmorpha nutans. It terminates below in a blunt cone, and a
little above this it presents a darker coloured zone, marked by longitudinal lines of small dark
bodies, which would seem to represent the fleshy processes given off near the base of the stem in
Corymoiyha .

The gonophore peduncles are very numerous. They appear to be about forty, and are
branched, and carry the gonophores on the extremities of the branches. The planoblasts are
remarkable for the height of the bell in proportion to the width. The margin of the umbrella
carries foui" nearly equal bulbs, but no tentacles appear to have been developed from these in
the specimen, an indication, doubtless, of an immature state.

The specific name of this hydroid has been given to it by Steenstrup after the locality iu
which it was discovered.

TROPHOSOME. — Htdeocattlus solitary, naked. Htdranths with a pro:simal
aud a distal set of filiform tentacles.

GONOSOME. — GoNOPiioEES in the form of fixed spokosacs.

MONOCAULUS, Allman.

Name.—YmA\ uoloc, solitary, and K-auXo'c, a stalk, in allusion to the single zooid of which the

trophosome consists.

TROPHOSOME. — Hydranth abruptly distinct from the hydrocaulus ; proximal

51

39G MONOCAULUS GLACIALIS.

tentacles longer than the distal, and disposed in a single verticil near the base of the
hydranth, the distal set scattered over a zone close to the summit of the hydranth.

GONOSOME. — Sporosacs borne upon peduncles, which spring from the body of
the hydranth between the proximal and distal sets of tentacles.

The genus Monocaulus is constituted for the Corpnorpha yhickdis of Sars, a form which,
though its trophosonie is that of a Corymorpha, is yet strongly distinguished from the true Cory-
morphas by its adeloconic gonophores.

The description given by Agassiz of his Corpnorpha pendula renders it necessary to place this
hydroid also in the genus Monocaulus.

*^* 1. Monocaulus glacialis, Sars.

CoRYMORPHA GLACIALIS, — Suvs, Forhaiidl. i Vid. Selsk. i Cliristiania, 1859; trans, in Wieg.

Arch., 1860, p. 341 ; and in Ann. Nat. Hist., 1861, p. 353.
Monocaulus glacialis, — Allman, in Ann. Nat. Hist, for May, 1864.

TROPHOSOME. — nmROCAULUs attaining a height of four or five inches.
Proximal tentacles of htdeanth very long, forty to fifty in number ; distal tentacles
very short and very numerous.

GONOSOME. — Peduncles of gonophoiies thirty to thirty-five, but slightly
branched. Sporosacs oval, destitute of tentaculiform appendage.

Habitat. — On a soft clayey and stony sea bottom.
Bathpnetrical distribution. — Deep sea, from 60 to 120 fathoms.
Locality. — The Varangerfjord, near Nadsoe, Sars.

The only account we possess of Monocaulus glacicdis is Sars's short description, from which
the above diagnosis has been compiled, and which is by no means so full as could be desired.
Sars's specimens were obtained in small quantity at a depth of from 60 to SO fathoms, and rather
abundantly at a depth of from 80 to 120 fathoms. The only locality where it has as yet been
found lies as far north as 70° N. lat. It thus possesses special interest, not only from the depth
at which it lies, but from its very high northern latitude.

Prevailing colour a clear bright pink.

# #

MONOCAULUS PEiXDULUS. 397

2. MoNOCAULUS PEXDULUS, Agassiz

CoRVMORPHA TENDULA, — Affossiz, Cont. Nat. Hist. U.S., vol. iv, p. 270, pi. .\xvi, figs. 7 — 17.
MoNocAULUS PEXDULUS, — Allinuti, in Aun. Nat. Hist, for May, 18G4. ■" ■!',/>■ 37o,

TROPHOSOME. — Htdrocaulus attaining a heiglit of four inches, and in its
thickest part a diameter of one quarter of an inch, from which part it gradually tapers
towards its proximal and distal ends, proximal end emitting filamentary hydrorhizal
rootlets. HroROCAULUS marked hy longitudinal hands. Hydranths pendulous.

GONOSOME. — Peduncles of gostophoues branched. Sporosacs (male) with a
thick tentaculiform process projecting from one side of the distal end, and with a
band of greatly developed cells over each radiating canal.

Habilat. — On a sandy or muddy sea bottom.

Bathy metrical distribution. — Coralline zone ?

Localities. — Massachusetts Bay, Agassiz aud Stimpson ; Cape Cod, Agassiz.

The present species is one with regard to whose generic position we are by no means free
from uncertainty. Agassiz gives us some beautiful figures of it, but the gonosome is hardly
figured in sufBcient detail. It would seem that radiating canals are visible in the gonophores ;
it appears pretty certain, however, both from the description and the figures, that the gonophores
are as truly adelocodonic as those of Tuhiilaria.

A short thick tentacle is described as being developed from one side of the distal end of the
gonophore ; and it is also stated that there exists " on the outer surface of the bell, over each
radiating tube, a narrow longitudinal band of enormously developed cells." It is to be desired
that we had more information regarding these bands than what is contained in the remark here
quoted. The presence of a voluminous spermatic mass, enveloping the spadix and completely
filling the cavity of the sporosac, shows that the male specimens — the only ones described — were
mature when examined.

Mr. Alexander Agassiz, on the other hand, figures (' Proc. Boston Soc. Nat. Hist.,' 1862,
and ' Illustrated Catal.,' p. 193) a medusa, Avhich he regards as that of Professor Agassiz's Cory-
morpha pendula. It resembles that of a true Corymorjjha, but besides the long solitary tentacle
has others in process of development from the margin of the umbrella. Since, however, he states
that it was taken in the open sea, and not du'ectly observed to proceed from the trophosome to
which he inferentially attributes it, we cannot consider Professor Agassiz's description of the
gonophore in his Corymorjjha pendida as superseded by the observation of his son.

The chitinous pellicle of Monocanlus pendula is described by Professor Agassiz as con-
spicuous over the proximal third of the hydrocaulus, though on the remaining two
thirds it has either disappeared altogether, or exists only as a mere film. This condition

398 TUBULARID^.

appears to be pretty much tlie same as what we find in Corymorjiha, where the thin pelHcle which
invests the stem becomes conspicuous near the proximal end by its separation from the ectoderm
at this place. It is totally different from that of the firm perisarcal sheath of other hydroids, and
cannot be regarded as invalidating the character of "naked" attributed to the genera of the
Cortjiaorplndce and MoiiocauIuJce.

TUBULARIDjE.

TROPHOSOME. — Hydeocaulus developed, invested by a cliitinous peeisaec,
Hydeanth with a proximal and a distal set of verticillate filiform tentacles.

GONOSOME. — GoNOPHOEES in the form of fixed spoeosacs.

TUBULARIA, Linnmis (in part).

Name. — From tubulus, a diminiiti\'e noun formed from tuhm a pipe, in allusion to the pipe-
like stalks which support the hydranths.

TROPHOSOME. — IIydeophyton consisting of a simple or branched hydeocaulus
and a filiform adherent nxDEOEniZA. Hxdeaxths flask-shaped, abruptly marked off
from the supporting stalk ; tentacles composing the proximal circlet larger than
those composing the distal one ; distal circlet surrounding the base of a conical
hypostome.

GONOSOME. — Spoeosacs developed upon branched pedmicles, so as to form
racemiform clusters, which spring from the body of the hydranth between the distal
and proximal circlets of tentacles. Embryonal development by Actinul^.

The genus Tiihularia of the older naturalists was an ill-assorted and heterogeneous group,
and at the close of the last century it still included, along with the true Tuhulance, not only other
hydroid genera, but various Actinozoa, Polyzoa, and certain members of the sub-kingdom
Annulosa. Even after it had been freed from its more glaring misassociations, it continued to be
imperfectly defined ; and it was reserved for Ehrenberg, in 1832, to make the first step towards a
reform of the group as it then stood, though from overlooking certain characters of importance
he failed to assign to it its natural limits as a well-defined zoological genus.'

^ Ehreuberg, Corallenthiere, Abliandl. der Kouig. Ak. der Wessens. zu Berlin, 183,3.

TUBULARIA. 399

Elircnberg's reform consisted in a (lisnicinliernicut of the genus Tiihiilaria as tlien accepted
into two groups, one of wliicli included the simple and the other the branched forms. Tiie
former he retained under the old name of Tubularia, while for the latter he constituted a new
genus under the name of EuJcndri/im.

Though Ehrenberg did not strike upon the true grounds of n philosophic revision when
he selected as the fundamental character of his proposed groups the simple or branched condition
of the hydrocaulus, neglecting the far more important characters found in the form of the
hydranth, and the presence of two circlets of tentacles in Tubularia, while there is only a single
one in Eudendrium, his establishment of the genus Eudendrium as distinct from Tubularia, never-
theless led the way to more accurate definitions of the two groups, and tiius became a most
important step in the systematic zoology of the Hydroida.

There is no one to whom we are so much indebted as to Agassiz for a knowledge of this
beautiful genus. He has made some of the North American representatives of it the subject of
laborious and conscientious study, and has illustrated his researches by the very finest hydroid
iconography in existence. I am unable, however, to accept in all points the systematic dismem-
berment of the genus Tubularia, as proposed by the celebrated zoologist who has adopted America
as his coimtry.

The genus Tubularia of modern systematists has been broken up by Agassiz^ into four
separate genera, for one of which he retains the name of Tubularia, while for the three others he
proposes respectively the designations of Parypha, Tkamnocnidia, and Eclopleura. Agassiz gives
no technical diagnosis of any of those genera, and it is by no means easy to discover the
characters upon which he would chiefly rely as the grounds of his division ; but from my own
knowledge of the European species, which he separates from Tubularia, as well as from the very
detailed descriptions and beautiful figures of the American species, which he now for the first
time records, and refers to his new genera, I can find only in one of these forms, namely, the
Tubularia Dumortieri, of Van Beneden, characters which would, in my opinion, justify the
proposed dismemberment. Eor Tubularia Dumortieri Agassiz constitutes a new genus, under
the name of Eclopleura, and relying on Van Beneden's account of this hydroid, I willingly follow
Agassiz in regarding it as the representative of a separate genus ; but the only common character
of importance by which Pari/jjha and Tkamnocnidia are separated from Tubularia, would seem to
be the non-development of distinct gastro-vascular canals in the sporosacs of the species referred
by Agassiz to these genera, while they are present in the sporosacs of those species to which
he would restrict the name of Tubularia.

Now, I cannot admit that the apparent absence of these canals, if unaccompanied by any
other ditference of importance, ought to be regarded as affording a character which would justify
the construction of a separate generic group ; for besides the practical objection that it is fre-
quently very difficult to detect them even when present, it should be borne in mind that though
they may exist in the younger sporosac, they may entirely disappear before the contents of
the gonophore are discharged, a fact which Agassiz himself notices in the case of his Tubularia
Contiioni/i.

Again, between Panjpha and Thavinocnidia, the only diff'erence alleged is in the structure
of the tentacula-like processes, which occiu- upon the distal end of the sporosac. I believe, how-

' ' Contr. Nat. Hist. U.S./ vol. iv.

400 TUBULARIA INDIVISA.

ever, that there is here no important difference. I have carefully compared the sporosacs of
Tuhularia larynx (= T. coronafa, Abildg.), a species which Agassiz refers to his genus Thamno-
cnidia with that of Tubularia mesembryanthemum, mihi, a Mediterranean species which would
certainly be referred by Agassiz to his genus Parypha, and the only difference of importance I
can find between them is the conical form of the apieal processes in T. larynx (see Plate XXIII,
figs. 21, 22, 24) and their laterally compressed crest-shaped form in the female sporosacs of
T. mesemhryanthenuuH (see woodcut, figs. 83, 84, page 419). In both cases these processes
contain a cavity, but the cavity is more conspicuous in the crest-shaped processes regarded by
Agassiz as characteristic of his genus Parypha than in the conical processes which characterise
his genus Thamnocnidia. Of the three new genera, then, which Agassiz has constituted for the
reception of certain forms included by other .systematists in the genus Tuhularia, I can only
admit the validity of one, and it is therefore to this one alone {Ectopleicra) that I have assigned a
generic rank in the present monograph.

I do not, however, wish to depreciate too much the value of the characters assumed by
Agassiz as distinctive of the others; and 1 believe that these characters may be conveniently made
the grounds of a subordinate grouping of the genus Tubularia. I shall, therefore, employ
Agassiz's names for the designation of the subordinate groups or sub-genera thus constituted.

Among the characters which must be regarded as of value in determining the limits of the
species included under the genus Tuhularia, is one which has been hitherto unrecognised. It is
afforded by a remarkable condition of the ccenosarc inmiediately below the summit of the stem,
and shows itself in the presence of a collar-like expansion of this part, convex on its upper
surface, which is always marked with rather deep, radiating flutings, and slightly concave on the
lower. I believe that it is only the ectoderm which participates in the formation of this collar.
It is entirely absent in Tuhularia indivisa, but in Tuhularia lanjinr, and all other species of this
genus which I have had an opportunity of examining, it is present.

So far as we can determine from the data before us, species 1 — 4 in the following
enumeration would be referred by Agassiz to Tuhidaria ; 5 — 15 to Thamnocnidia; and 15 — 17
to Farypha.

At' Sub-genus, Tubularia proper, Agassiz.
Sporosacs with conspicuous gastro-vascular canals.

1. TuBULAEIA INDIVISA, LinncBus.
PLATE XX.

Adianthi AUKEi MINIMI FACIE PLANTA MARINA, — RttU, Sjn. 31, 4. Jussieu, in Mem. Acad.

Eoy. des Sci., 1742, p. 296, tab. x.
Tubular coralline, like oaten pipe, — Ellis, in Phil. Tiaus. for 1754, p. 504, tab. xvii,

fig. D.
Corallina tubularia calamos avenaceos referens, — Ellis, Coral!., p. 31, tab. xvi, fig. C,

TUBULARIA INDIVISA. 401

TuBiiLARiA INDIVISA, — Liimifitis, Sjst., 1758. Lamarck, An. saus Vert., 1816, vol. ii, p. 110.

Lamonroux, Cor. flex, 230. Cuvier, Regn. Anira., iii, 299,

Johnston, Brit. Zoopli., 1847, p. 48, pi. iii, figs. 1, 2.

Dalyell, Rare and Remark. Anitn., vol. i, p. 2, pis. i — iv.

Mummery, in Trans. Mic. Soc., 1853. Hincks, Brit. Hydr.

Zooph., pi. 115, pi. XX.
TuBULARiA CALAMARis, — Palhis, Eleuclius, p. 81. Ehrenberr/, Corall. Roth. Mer. Abhandl.
Berl. Acad., 1832, p. 295.

TROPHOSOME. — Hydrocaulus consisting of a cluster of simple tubes, destitute
of annulation, separate from one another above, but forming an entangled mass
below, where the tubes become smaller in diameter, and more or less twisted together
and adherent to one another ; they vary in height in different specimens from about
three to nine inches, and in the upper part of their course have a diameter of about
one tenth of an inch ; ccengsakc longitudinally striated, not forming a collar-like
expansion below the hydranth. Htdeoehiza consisting of branched, sinuous, inos-
culating tubes. Htdranths with twenty to thirty proximal tentacles in a single
verticil, and with about forty distal tentacles in two or three alternating verticils,
which are so closely approximated as to form a single circlet.

GONOSOME. — GoNOPHOUES oviform in three or four pendulous racemes, which,
when mature, surpass the body of the hydranth in length, the racemes of the male
colony being somewhat longer than those of the female ; the gonophores are destitute
of tentaculiform tubercles, and have four radiating canals terminating in a circular
canal, which surrounds a perforation near the distal end of the gonophores. Actinula
with oral tentacles at the period of its liberation.

Colour. — Body of hydranth varying from a pale pink to a bright crimson or scarlet ; hydro-
caulus brown below, becoming light red or orange red towards the summit ; spadix and peduncle
of gonophore scarlet.

Development of Gonosome. — From April to October.

Habitat. — Attached to rocks, stones, shells, &c., in the sea.

Hathymetrical distribution. — Laminarian to deep-sea zone.

Localities. — English, Scottish, and Irish coasts abundant ; Coast of Greenland, Morch ;
Coasts of Scandinavia as far as the North Cape, Sars ; and the coasts of Belgium and Northern
France.

This magnificent hydroid — the type of the family of the TubularidcB — besides offering
features of great interest in the morphology and physiology of the Hydroida, has a special
historical value, for it was it which in the hands of Bernard de Jussieu afforded the first proof we
possess of the animality of the marine Hydroida.

During the early part of the last century the Tabularia indivisa had already by its large size
and conspicuous hydranths forced itself on the attention of the marine naturalist. Its nature.

402 TUBULARIA INDIVISA.

however, was entirely misunderstood ; no one doubted as to its being a genuine sea plant, and
with this impression we find it described by the naturalists of that period under designations
indicative of a belief in its vegetality, such as that of " Adianthi aurei minimi facie planta
marina," under which it was recorded l)y Ray.

In 1741 De Jussieu obtained it during a visit to the coast of Normandy, when along with the
actinozoal Alcyonium, and the polyzoal Flustra and Cellcpora, he made it the subject of iiis famous
memoir,^ in which, opposing himself to the general belief of the day, he not only supports the
views of Peysouel in favour of the animality of coral, but extends them to tlie plant-like Ilydroida
and Polyzoa. The figure of Tubularia indimsa which illustrates this memoir, though without
the manipulative skill of modern engravers, is admirable ; it is more exact than that subsequently
given by Ellis, and though the gonosome is not represented in it, it has remained up to the
present day unsurpassed in accuracy and expressiveness.

In the former part of the present monograph the structure of Tubularia indivisa has been
fully described ; it has been shown how the coenosarc of the stems, instead of presenting the
usual axial cavity, is excavated into numerous peripheral channels, first pointed out by AVright,
with their walls all richly ciliated, and with the somatic fluid circulating through them in
advancing and returning currents, and how the endodermal lining of the Jiydranth cavity is
thrown into pendulous lobes and marked by deep intersecting sulci. It has been shown that the
longitudinal fibrillse, conspicuous in the tentacles, are formed by series of very much elongated,
fusiform, nucleated cells, presentuig thus the essential structure of the non-striated muscle of
higher animals. The structure of the gonophores has also been fully detailed, with their four
conspicuous, radiating canals, opening into a small circular canal which surrounds an orifice in
the distal end of the gonophore, through which the contents of this body escape at the period of
maturity. It has been further shown that the ova have their origin in differentiated masses of a
granular plasma, which is developed as usual between the endoderm and ectoderm of the spadix,
and which in its early condition consists of nucleated cells; and that these ova are developed
into actimdcE, though no evident germinal vesicle nor any true process of segmentation has as yet
been detected in them. And further, the remarkable phenomenon first noticed by Dalyell of the
successive shedding and renovation of the hydranths has been described, and it has been shown
that the new hydranth is produced by a metamorphosis of the distal end of the decapitated stem
rather than by a true budding.^

Few more beautiful objects present themselves to the student of marine life than a well-
developed specimen of Tubularia indivisa. From a complicated and intertwining mass of stems
where the hydroid roots itself to some submarine rock or the surface of some old shell, it gradu-
ally becomes disentangled, and soon displays a group of flexile cylindrical stems, rising without a
branch to the height of many inches, and each crowned by a scarlet or crimson hydranth, with its
double coronal of tentacles. The longer tentacles now spread abroad like the petals of an ex-
panded flower, now closed in over the summit, like the same flower in its bud, and now again
thrown back in gentle curves round the summit of the supporting stalk, while the long pensile

^ Bernard de Jussieu, in ' Mem. Acad. Roy. dcs Sci.,' 1742.

■ The details of the morphological and physiological facts here referred to will be found in
pages 69, 121, 131, 205, &c., and pi. xxiii.

TUBULARIA COUTHOUIF. 403

clusters of berry like gonopliores, which droop gracefully from aiuong the tentacles, complete the
attractions of this beautiful hydroid.

During the spring and summer months Tuhaluria indivisa is in its greatest perfection,
though in some localities it may still be found in good condition until late in the autumn. It is
during the season of its most active growth, and when the hydranths with their racemes of gono-
phores have attained their greatest size and perfection, that these may be seen to be perpetually
cast off and renewed, the stem increasing in height with the formation of each successive hydranth.
But towards the end of summer the renewal of the hydranths after the casting off of the old ones
appears to cease, and we now usually find the upper parts of the perisarcal tubes empty, while
their lower parts and the hydrorhiza are still filled with the living coenosarc. In this state, I
believe, they generally remain during the winter, ready on the return of spring to throw out new
hydranths, and these hydranths, with their clusters of gonopliores, undergo in their turn successive
shedding and renovation, until the autumnal months once more put a check to the activity of
their functions. Each successive renovation of the hydranth, and consequent elongation of the
stem, is marked by a slightly elevated annular ridge on the surface of the perisarc.

Tubularia indivisa will continue to live for some time in the confinement of our aquaria, fre-
quently throwing off and renewing its hydranths, and giving origin to hundreds of embryos,
which during their actinula stage look like minute spiders creeping over the bottom of the tank
or floating passively in the water. Many of these actinulae will pass through subsequent phases
of their development and attain their fixed condition, in which they may be seen rooted in multi-
tudes to the sides of the vessel, a forest of diminutive Tuhularice. They often, also, attach them-
selves to the stems of the parent colony, where they will continue to grow as long as these
stems may afford them sufficient surface of support.

The species seems to range from the upper limit of the Laminarian zone to a depth of thirty
fathoms, or even more. At extreme low-water spring tides it may frequently be seen growing
luxuriantly in the rock pools or spreading over the rocks where these are still washed by the
sea.

It is, perhaps, generally distributed over the European shores of the Atlantic. Some of
the finest specimens have been obtained from estuaries where the influence of the fresh water has
not yet wholly ceased to modify the saltness of tlie sea.

Under the name of Tubularia giyaniea, Lamouroux (' Exposition Methodique,' tab. LXVIII,
fig. 5) figures the dead stem of a large Tubularia, which he informs us attains a height of from
1 2 to 1 5 inches. He gives no description which might aid in its determination, but it is probably
only a large form of T. indivim.

*^* 2. TUBUL.A.UIA CouTHOuir, Agassiz.

TuBULARLA. CouTHoi'vi, — Affussiz, in Contr. Nat. Hist. U. S., vol. iv, p. 226, pi. xxiii a, figs.

8, 9, pi. xxiv, pi. xxvi, figs. 1 — 6. Alex. Agassiz, in
Illustr. Catal. N. A. Acalcpli.-c, p. 196.

TROPHOSOME

— HTDEOCAULrs varvint? in lieiglit from about three inches to six

404 TUBULARIA REGALIS.

inches, and consisting of clusters of uubranclied stems, which are about one twelfth of
an inch in diameter, and spring from a nxDRORniZA composed of " closely tangled
knotty root-like tubes ;" peris^vhc " more or less ringed or jointed, sometimes very
regularly at intervals of an eighth of an inch, or constricted once or twice, and then
again smooth throughout." Htdranth with the tentacles of the distal set about
fifty in number, gradually decreasing in length from before backwards, and disposed
in three or four indistinctly defined closely approximated verticils ; the hydranth is
large, and, when fully expanded, measures from tip to tip of the proximal tentacles
about an inch and a half.

GONOSOME. — GoNOPiioBES in dense pendulous racemes, which surpass the
proximal tentacles in length, and are disposed one over the other, so qs to form three
or four superposed verticils ; the gonopliores are elongate-oval or piriform in the male,
broadly oval or globular in the female ; in both destitute of tentaculiform tubercles
and with four radiating canals, which open into a small circular canal which
surrounds a perforation near the distal end of the gonophore.' Actintjla with oral
tentacles at the period of its liberation.

Development of Gonosome. — From March to December.
Habitat. — Attacbecl to suljmerged bodies in brackish water.
Locality. — Massachusetts Bay, Professor Agassiz.

This fine species is described and beautifully illustrated by Agassiz, from whose account of it
I have obtained the characters embodied iu the above diagnosis. It apparently comes near to the
Tuhularia indivisa of the European side of the Atlantic, differing from this species chiefly in its
more or less annulated stem, the larger size of the hydranths, the more numerous racemes of
gonophores, and the disposition of these in several superposed verticils.

Tiibidaria Couthouii has been studied by Agassiz, who has given us a very valuable chapter
on its structure. He has described in its stem a large-celled imperforate axis, and a system of
peripheral tubes, similar to those met with in Tubitlaria indivisa.

*^* 3. TuBULARIA REGALIS, Bocch.
T0BULARIA REGALIS, — Boec/c, iu Forhand. Videusk. Selsk Christiania, 1859, p. 115, tab. iii.

TROPHOSOME. — Hydrocaulus simple, attaining a height of from six to seven
inches, and with a thickness of about one eighth of an inch ; ccenosarc exhibiting
longitudinal vermilion bands, which are visible through the clear horny perisarc.

^ In some instances fine radiating canals were observed; plain!}' an abnormal ii-regularity.

TUBULARIA INSIGNIS. 405

HrDRANTn measuring across tlic proximal circlet of tentacles about two and a half
inches ; proximal tentacles twenty-eight in number.

GONOSOME. — GoNOPnoRES oval, destitute of tentaculiform tubercles, and dis-
posed in simple pendulous^ racemes, which alternate with the i)roxiraal tentacles, and
are equal to them in length.

Local i/^. — Spitsbergen.

With the exception of Amalthaa Januarii, there is no known hydroid whose hydranths
approach in size to those of the colossal Tuhularia just described. For our knowledge of it we
are indebted to a paper by Boeck, but further details are much to be desired. It approaches in
many respects to Tubidaria indivisa, and, like that species, presents the longitudinally striated
condition of its stem, which indicates a channeled structure of the coenosarc. We are not
informed whether radiating canals exist in the walls of the gonophores. These, after the escape
of the AdinulcB, present a wide orifice through which the spadix projects for a considerable
distance. The Adinula appears to resemble very nearly that of Tuhularia indivisa.

Tubidaria regalis was obtained on the coast of Spitsbergen, and is thus the most northern
hydroid whose tiophosome has been yet discovered.

*^* 4. TuBULARiA INSIGNIS, Allnuin.

TROPHOSOME. — Hydrocaulus simple, attaining a height of seven inches (or
more), gradually increasing in thickness from below upwards, until it attains a
diameter of one eighth of an inch ; ccenosarc longitudinally striated ; perisarc quite
smooth. Hydranth borne on a collar ; proximal tentacles about thirty, distal
tentacles more than 200 in a dense brush, formed by numerous closely superimposed
verticils ; length of proximal tentacles about nine tenths of an inch ; of distal about
three tenths.

GONOSOME. — GoNOPHOKES barrel-shaped, with terminal aperture, destitute o
tentaculiform tubercles, and with four obvious parietal canals ; peduncles of gono-
phores in six or seven imbricated verticils, with about twenty in each verticil, not
pendulous, each peduncle dividing near the summit sub-dichotomously into short
ultimate pedicells.

Lncalili/. — Dieppe, M. L. Rousseau.

The characters enumerated in the specific diagnosis just given are those of a tubularian,
which far surpasses in size every British representative of the genus. A specimen consisting of a

' Boeck's figure represents them as erect, an attitude which their slenderness and length
renders it impossible for them to as.^nrae.

406 TUBULARIA LARYNX.

single stem, with its li^-dranth and gonophores, exists in the Miiseuiu of the Jardin des Plantes.^
This sjjecimen, so far as I know, is unique, but it has been fortunately preserved in spirits, so that
some of its most important characters are determinable. It is desirable, however, that we know
more of this fine species than what can be satisfactorily made out from the preserved specimen, and
we must still wait for an opportunity of inspecting other examples, more especially living ones.
I have not been able to determine anything regarding its development, nor, indeed, even the sex
of the specimen, for though this was with great liberality placed at my disposal, I did not feel
justified in undertaking an examination which would have exposed to injury the only example as
yet preserved of this remarkable hydroid.

The height of the hydranth in the specimen, from its base to its summit, is about half an
inch, while the diameter of the basal portion is about three tenths of an inch. Tubularia insignis
nuist, indeed, when in a living state, be a magnificent hydroid. Nothing, however, can be deter-
mined from the specimen regarding the colour of either hydranth or gonophores, the original
colour of these parts having been entirely discharged by the action of the spirits ; that of the peri-
sarc, however, has probably remained unchanged. It is of a light brown in the distal parts of the
stem, becoming darker towards the base. Where the transparency of the perisarc allows a view
of the contained coenosarc, numerous longitudinal parallel striae may Ijc seen, doubtless indicating
the presence of a system of longitudinal coenosarcal canals.

The gonosome has scarcely attained complete maturity, only two or three gonophores being
sufficiently advanced, among the multitude of immature ones, to allow of even the approximate
determination of their proper form.

The only known facts regarding the history of this unique specimen are found in a short
appended note, which states that it was obtained at Dieppe, by M. L. Rousseau.

W Sub-genus Thamnocnidia, Agassiz.
Sporosacs without evident gastrovascular canals, apical processes conical.

5. Tubularia laetnx, Ellis and Solandcr.
Plate XXI.

FuCUS DEALENSIS FISTULOSUS LARYNGiE SIMILIS, RciH, Syn., i, 39.

Tubular coralline, wrinkled like the windpipe, — Ellis, in Phil. Trans, for 1754, vol.

xlviii, p. 504, tab. xvii, fig. l.
CoRALLiNA tubularia LARYNGi-siMiLis, — EUis, Coral., p. 30, pi. xvi, fig. b. Bast., Opusc.

Subsc, p. 28, tab. ii, figs. 3, 4, aud tab. iii,

figs. 2 — i.

' I must here express the obHgations I am under to Prof. Milne-Edwards, and Prof. Lacaze du
Thiers, for the opportunities they have afforded me of examining this aud other specimens of the
museum.

TUBULARIA LARYNX. 407

TuiiULARi.v MuscoiDEs, — Pul/us, Eleucli., p. 82 (ncc Llniueus, Giuel. Lin., 3832).

Tuiu'L.uuA L.iRYNX, — SuldiKkr's ElUs, p. 31. Lamarck, Auim. s. Vert., 1816, vol. ii.

Duhjell, Rare and Rem. Anim., vol. i, p. 42, pi. v. Johnston,

Brit. Zooph., second edit., p. .50, pi. iii, fig. 3, and

pi. V, figs. 3, 4. Alder, Catal., p. 16.

TuBUL.iRiA coRO.VATA, — AhUdguard, in Zool. Dan., pi. 141. Van Beneden, Tubulaires, p. 49,

pi. i, figs. 7 — 19, and Faune lit. de Belg., p. 106, pi. iv.
Hincks, Brit. Ilydr. Zooph., p. 119, pi. xxi, fig. 2.
EuDENURiUM BRVoiDEs, — Elirenhertj , Coralleuthiere, Abhandl. Ak. Wissen. Berl., 1832,

p. 296.
EoDENDRttiM SPLEXDIDU.M (?), — Elircnberg, Corallcntliierc, Aljhandl. Ak. AVissen. Berl.,

p. 296.
TuBULARiA GRACILIS, — Hurveij, in Proc. Zool. Soc. Lond., 1836, No. 41, p. 51. Johnston,

Brit. Zooph., second edit., p. 52, pi. iv, figs. 3, 4, .5.
Alder, Catal., p. 17.

TROPHOSOME. — Hydrocaulus consisting of numerous brandling stems liaving
a diameter of about -hih. of an inch, and rising from a creeping stolon to a height of
from one inch and a ludf to two inches, or even more ; stems presenting at intervals
more or less distinctly marked transverse annulations ; cct;nosarc forming a collar-
like expansion just below the hvdrauth. Htdrantiis about one fifth of an inch
across the widest part of the body, with a circlet of from fourteen to twenty distal
tentacles in two closely approximate alternate series, and with the proximal tentacles
about twenty in number, and about two fifths of an inch in length.

GONOSOME. — GoxopiiORES in pendulous clusters, forming in the male long
simple racemes, which, when mature, surpass the proximal tentacles in length, while
in the female the much shorter clusters do not equal these tentacles in length, and
the peduncle is here branched, so as to form a sort of panicle or compound raceme.
The gonophores are destitute of gastro-vascular canals, and are crowned with four
conical tentaculiform tubercles, larger in the female than in the male ; female gono-
phores somewhat more globular than the male, which are of an elongated oval form.
Oral tentacles of Actinula not developed at the time of its liberation.

Colour. — Body of liydranths and peduncle and spadix of gonopliore.s rose colour; perisarc
straw colour.

Development of Gonosome. — Observed from April to October.
Habitat. — Attached to rocks, stones, old shells, other hydroids, &c.
Bathymetrical distribution. — From Laminarian to deep-sea zones.
Locality. — British, Scandinavian, and Belgian shores.

As long as naturalists confined themselves in their description of Hydroida to the dry peri-
sarc, or when, without neglecting the soft parts, they still failed to see that certain differences in

408 TUBULAR! A LARYNX.

the gonosome depend merely on diSei'ent degrees of maturity and on differences of sex, it was
vain to expect correct specific diagnoses. The present species is exactly in this case ; much con-
fusion exists with regard to it, and it has been described under different names by different
observers, a confusion which, I believe, results in great part from regarding the differences
between the male and female colonies as indicative of two different species.

I have no doubt of the identity of our species with the Tabularia figured more than a
hundred years ago by Ellis, under the appellation of " Corallina tubularia laryngi similis," and in
his posthumous work, edited by Solander, described under the binomial designation of Tubularia
larynx. Ellis's hydroid is referred by Pallas to Tubularia mtiscoicles of the ' Fauna Suecica,' an
identification which Liimeus's short diagnosis, " T. culmis sidjdichotomis totis annuloso-rugosis,"
will hardly allow us to accept. Lideed, when we consider the very imperfect knowledge of the
Hydroida at the time of the publication of the ' Fauna Suecica,' we must see that there is more
reason to regard the Tubularia muscoldes of that work as an annulated species of Cori/ne than there
is to view it as a true Tubularia. The sufficiently characteristic figure of ElHs, however, though
without any representation of the gonosome, and his short but accurate description, leave no
doubt as to the hydroid the celebrated English observer had before him, and gives us a fixed and
definite point in the determination of its synonymy.

I believe, also, that the Tubularia coroiiafa of Abiklgaard, figured in the ' Zoologica Danica,'
is identical with Ellis's species, notwithstanding the unbranched condition and tortuous marking
of the stems and the very short peduncles of the gonophores in the otherwise excellent figure of
the Danish naturalist, while I have little doubt that it is this same species which is described by
Mr. J. B. Harvey, in the ' Proceedings of the Zoological Society,' under the name of Tubularia
gracilis.

Hincks maintains the specific distinctness of Tubularia larynx and T. coronafa. I cannot
see, however, in the characters contrasted sufficient grounds for distinction. The more important
of them appear to me to be merely sexual differences.

If the criticism just offered of the synonyms of our species be accepted, we have no alterna-
tive but to adopt for this hydroid Ellis's specific name of larynx, as given in his posthumous work
published under the care of Solander.

When Tubularia larynx is examined in its living state, obscure annular corrugations may
usually be observed at short intervals on the stems ; but it is in dried specimens that these cor-
rugations become distinctly marked, and afford a character which, in the various descriptions of
the species, has taken a prominent place, and has suggested the specific designation given to it by
Ellis.

Under the name of Tuhdaria coronafa the present species has been made the subject of
some of Van Beneden's researches, published in his ' Faune Littorale de Belgique,' and in which
his attention has been especially directed to the structure of the sporosac and the development of
the Actenula.

I have obtained Tubularia laryux in abundance fi'om the lines of the deep-sea fishing boats
on the east coast of Scotland. It is also, however, an inhabitant of the Laminarian zone, and,
like some other tubularian hydroids, it appears to delight in an admixture of fresh water with
the sea, some of the finest specimens having been obtained from the lower reaches of estuaries.
Though a much humbler species than Tubularia indivisa, it is yet one of the most charming of
our British hydroids. When well developed it forms dense bushy tufts, which, with their rosy

TUBULARIA BELLIS. 409

astei--like hydrantlis expaiuliiig from tlie suniiiiits of the Ijraiiclics, suggest, witli a vividness
unsurpassed hy any other liydroid, some of the most faniihar forms of the flower garden.'

6. TuBUL.vKiA Bellis, Alhion.

Plate XXII, figs. 5 and G.

TuBULARiA Bellis, — Alhnuii, in Anu. Nat. Hist, for Jan., 1863. Hinck.i, Brit. Hydr.

Zooph., p. 122, ()1. x.\i, fig. 3.

TROPHOSOME. — Hydrocaulus consisting of short sparingly branched stems,
which are mostly prostrate at the base, where they spring from the creeping stolon,
and then becoming erect attain a height of three quarters of an inch or even one
inch ; PERiSARC towards the basal portion marked with wide but distinct annulations,
which disappear towards the distal extremity of the stem. Hyduanths supported on
an annular collar-like dilatation of the coenosarc, and with the breadth of its base
exceeding its entire height; about twenty tentacles in the proximal circlet, and
fifteen or twenty in the distal ; diameter of proximal circlet when extended about five
lines.

GONOSOME. — Clusters of gonophoues short, erect, with four or five gonophores
usually in the cluster, the gonophores, both male and female, crowned with four
conspicuous conical tubercles.

Colour. — Hydranth scarlet ; coenosarc reddish-orange, becoming deeper in tint towards the
base ; spadix scarlet.

Development, of Gonosome. — .July and August.
Habitat. — Attached to the sides of rock pools.
Bathymetrical disfribtttion. — Laminarian zone.
4oca/^^.-— Shetland Isles.

Tubularia Bellis is an exquisite little hydroid. It occurs in considerable abundance round
the shores of the " Outer Skerries" and some of the other small rocky and more exposed islands
of Shetland, where it grows in shallow rock pools, exposed only at extreme low-water spring
tides, amid luxuriant meadows of Laminaria, and where the bright colour of its hydranths render
it a conspicuous and beautiful object beneath the pure transparent water of the rock pool.

1 lu the coUectiou of the Jardin des Plantes is a dried specimen, labelled in Lamarck's nriting,
" Tubularia larynx." It is a sparingly branched, irregularly rugose, strong and somewhat coarse form.
It is impossible, however, to determine it from the dried perisarc, but it is apparently very different
from the true Tubularia larynx.

410 TUBULARIA ATTENUATA.

7. TuBULARIA ATTENUATA, Allman.

Plate XXII, figs. 1 and 2.

TuBULARiA ATTENUATA, — Alhnan, iu Aun. Nat. Hist, for July, 1864. Hincks, Brit. Hydr.

Zooph., p. 122.

TROPHOSOME. — Htdrocaulus attaining a height of three or four inches,
slender, ohscurely corrugated, very irregularly branched, with the branches given off
at a veide angle. Htdeanth supported on an annular expansion of the ccenosarc,
and with the tentacles of the proximal circlet about three times as long as those of
the distal one.

GONOSOME. — GoNOPiiORES (male)^ borne on short erect branched peduncles,
with usually five to eight in a cluster, and with four conical tentaculiform processes,
which equal in length nearly half that of the mature gonophore.

Colour. — Body of hydranth deep vermilion between the two tentacular verticils, paler in the
enlarged base ; coeiiosarc pale pink, with light straw-coloured perisarc ; spadix vermilion.
Development of Gonosome. — June.
Habitat. — Attached to old shells and stones in the sea.
Batliymetrical distribution. — Deeper pai'ts of Coralline zone.
Localities. — Shetland Islands and Firth of Forth, G. J. A.

Tubularia attenuata is a deep-water species. I have dredged it from about fifteen fathoms in
the Firth of Forth, and from about fifty fathoms in the seas round the Shetland Islands. It has
a diffuse and somewhat straggling habit. Its nearest congener would appear to be the Tubularia
simplew of Alder, from which, however, it differs by its branched hydrocanlus, and apparently,
also, by the greater length of its distal tentacles.

' I have had no opportunity of examining female specimens.

TUBULARIA HUMILIS. 411

8. TUBULAKIA SIMPLEX, Aider.

TuBULARrA SIMPLEX, — AlcJcr, Catal. Suppl., pi. viii, figs. 3, 1. Hincks, Brit. Hydr. Zooph.,

p. 121, pi. .Kxii, fig. 1.
TuBULARiA Du.MORTiEui, — JoluistoH, Biit. Zoopli., sccoud edit., p. 50. Alder, in Trans.,

Tynes. Club, vol. iii, p. 106.

TROPHOSOME. — Hyurocaulus attaining a height of from two to two and a
half inches, slender, unbranched, smooth, generally a little bent or geniculated at
intervals, and tapering towards the base, usually solitary. Htdranths slender, distal
circlet of tentacles bi-serial, shorter and less numerous than in Tuhularia indivisa,
the proximal with about twenty or twenty-four moderately long tentacles.

GONOSOME.— Not observed.

Colour. — Hydraiith rose colour ; crenosarc orange or scarlet, with the perisarc horn coloiu-.
Habitat. — On shells, &c., in the sea.
Hathy metrical distribution. — Deep sea zone.
Locality. — Coast of Northumberland, Mr. Alder.

I have never met with this species, which is, doubtless, nearly allied to Tubularia attemiata,
and, like it, is a deep-water species. Its unbranched hydrocaulus, however, and the angular
flexures of its stem, will ati'ord characters by which it may be distinguished.

Mr. Alder, who at first referred this species to the Tubularia Dumortieri of Van Beneden,
subsequently recognised its distinctness, and described it as a new species.

9. TuBULAKiA HUMILIS, AUmcin.
Plate XXII, figs. 3 and 4.

TuuuLARiA HUMILIS, — Allinun, ill Ann. Nat. Hist, for July, 18G4-. Hincks, Brit. Hydr.
Zooph., p. 123.

TROPHOSOME. — Hydrocaulus attaining a height of about an inch, sparingly
branclied ; perisarc delicate, with nearly obsolete transverse corrugations. Hydraxth
supported on an annidar expansion of the coenosurc ; greatest breadth of the basal

53

412 TUBULARIA CALAMARIS.

portion exceeded by the entire height ; about twenty proximal and fifteen distal
tentacles ; diameter of proximal circlet from tip to tip of the tentacles, when extended
in fidl-sized specimens, about two lines.

GONOSOME. — GoNOPiiORES (male)^ borne on very short branching peduncles
forming erect clusters, with usually about three gonophores in each cluster ; summit
of gonophore with four rather large conical tentaculiform tubercles.

Colour. — Hydranth scarlet, cociiosarc reddish yellow, perisarc liglit straw colour, spadix
scarlet.

Development of Gonosome. — Autumn.
Habitat. — On exposed rocks in the sea.
Batlii/metrical didrihution. — Lamiuarian zone.
Locality. — Kinsale Harbour, Ireland, G. J. A.

I obtained Tahularia Immilis during the autumn attached to rocks close to the level of
low-water spring tides near the mouth of Kinsale Harbour. It is a very elegant little hydroid,
resembling Tubularia Bellis in its mode of growth and in the shortness of its gonophore clusters,
but is at once distinguished from this species by the absence of distinct annulation, and by the
smaller size and less appressed form of its hydranths.

*^* 10. TuBTJLAUiA CALAMARIS, Van Bencdeii.

Tubularia calamaris, — Van Benedeu, Recherches sur I'Embryogeuie des Tubulaires, p. 44

(exclusive of synonymes), pi. i, figs. 1 — 6 ; Recherches
sur la Faune littorale de Belgique, p. Ill (exclusive of
synonyraes).

Thamnocnidia CALAMARIS, — Aycissiz, Contr. Nat. Hist. U. S., vol. iv, p. 343.

TROPHOSOME. — Hydrocaulus attaining a height of about two inches, simple
and straight towards the summit, tortuous and irregularly ramified at the base, with
the tubes here united to one anotlier, so as to form a cluster ; perisarc annulated at
the base. Htdraxtus with about thirty tentacles in the proximal circlet, and about
twenty in the distal.

GONOSOME. — Gonophores borne on short erect peduncles ; summit crowned
with four well-developed conical processes; walls of gonophore with longitudinal
contractile rib-lilie bands.

' All tlie specimens e.xainined were male.

TUBULARIA POLYCARPA. 413

Colour. — llvdraiitli niid C(ciu)s;irc red ; in some specimens nearly or quite colourless,
Habitat. — Attached to submerged bodies of various kinds.
Bathpiietrical Dhtrihulion. — Laminarian to deep-sea zones?
Localifj/. — Coast of Belgium, Van Benedcn.

The TiiJjiilarla cnhimnrh of A'an Benedcn is a species with regard to which there is much
confusion. The Belgian naturalist assumes it to be identical with the Tubidaria calamaris
of Pallas, and the synonymes attached to his description are selected in accordance with this
view.

Van Beneden's sjiecies, however, is altogether different from Pallas's Tidtutaria calamaris,
which is the same as the true Tubularia indivisa of Linnaeus. From this species it is at once
distinguished by its much smaller size, branched hydrocaulus, and short erect clusters of gono-
phores — supposing this condition to be that of mature individuals — as well as by the occurrence
of tentaculiform appendages, and of meridional bands on the gonophores.

It approaches more nearly to the Tubularia larynx, but from this again it is separated by
the short, erect clusters of gonophores, and by the peculiar contractile, meridional bands of the
gonophore.

It is to be regretted that we do not know more of these contractile bands than what we learn
from Van Beneden's figure and the descriptive paragraph which refers to it. According to Van
Beneden the bands are five in number, equidistant, and running in meridional lines along the
sides of the gonophore ; their normal number, however, is proba])ly only four. "When contracted
they give a lobed outline to the gonophore.

There can be little doubt that two or more species have been confounded in Van
Beneden's description of his Tubularia calamaris. It is at all events certain that — led astray,
apparently, by an incorrect identification — he refers to it observations which various naturalists
have made regarding the true Tubularia indivisa.

Professor Van Beneden states that his Tubularia calamaris is very aljundant on the Belgian
coasts. I have never met with any Tubularian referable to it on the British or other coasts
which I have explored.

*^* 11. TUBULAUTA POLYCARPA, Allmau.

TROPHOSOME. — Hydrocaulus attaining the height of al^out an inch and a
half, and a thickness of about half a line, unbranclied, or sparingly branclied near the
base, each stem crowned by a fluted coUar for the support of the hydranth, and
having its perisarc marked here and there by a few indistinct annulations ; hydro-
RUiZA consisting of an entangled plexus of thin tubes. Hydranth measuring about
half an inch from tip to tip of the basal tentacles, which are about twenty-four in
number.

414 TUBULARIA SPECTABILIS.

GONOSOME. — GoxopiiORES oval, in about twenty-four dense clusters, alternately
longer and shorter.

Colour. — Body of hydranth carmine.

Habitat. — Found covering the bottom of a ship in the haibour of Coquimbo.

Localif}/. — Coquimbo, South America.

The above diagnosis has been drawn up from a specimen preserved in spirit, and sent to me
by Dr. J. E. Gray, who received it from the coast of Coquimbo. Though a true Tubularia it is
quite distinct from our European species, as well as from those of the North American coast.
One of its most striking features is the great abundance of its clusters of gonophores, which are
more numerous than in any other species with which I am acquainted.

In size it resembles the Tubularia larynx of our own shores, and like it has its stems
terminated immediately below the hydranths by the peculiar fluted collar, which is found in this
and other species. From Tubularia larynx, however, it is easily distinguished by its simple stems,
and by the profusion of its gonophore clusters.

The common peduncles of the gonophore clusters are destitute of gonophores for a con-
siderable distance from the root, and these naked flexile stems must have given a pendulous
attitude to the clusters during life, though this attitude is not very obvious in the contracted
state of the preserved specimens.

*^* 13. Tubularia spectabilis, Agassiz.

Thamnocnidia SPECTABILIS, — Affossis, Contr. Nat. Hist. U.S., vol. iv, p. 271, pi. xxii, figs.

1 — 20. A. Af/assiz, lUustr. Catal. N. A. Acal.,
p. 195.

" The description already given of the head with its proboscis, the tentacles and the bunches
of niedusoids [sporosacs], the stems and their mode of branching, and the horny sheath of
Parypha \Tubularia\ crocea, apply equally to this hydroid, with the following exceptions : — The
horny sheath is quite uniform and smooth as far as it covers the stem above its base, and is a
very little narrower below than above, but the entangled mass of the base is perhaps more dense
than in Paryjjha."

Tlie, yonophores have "three or four solid, short, unshapely tentacles" [apical processes].
In other respects the structure of the gonophore " is almost identical with that of Parypha
crocea, even to the absence of radiating and circular chymiferous vessels." — Aya^siz.

Development of Gonosome. — Summer and autumn.
Habitat. — On floating timber in brackish water. — Ayassiz.
Locality. — Massachusetts Bay. — Ayassiz.

TUBULARIA TENELLA. 415

Tliongli Aj^assiz raises tlie present species to tlie rank of a separate genus nnder tlie name
of Thnintocnidia, lie gives no precise diagnosis of it either generic or specific, and I liave been
obliged to content myself witli simply quoting his remarks as above. The beautiful figures,
however, which accompany these remarks will greatly aid in identification.

The new genus Thamnocnidia, which Agassiz forms for the present species, appears to be
based on the form of the tentacula-like processes of the gonophore, and the apparent absence of
those parietal canals which in the gono{)hores of Tubidaria indivisa represent the gastrovascular
canals of the more developed medusa — characters which I do not regard as sufficient to justify a
generic separation.

To liis genus Thamnocnidia Agassiz also refers the Tubidaria larynx of the European shores
and the Tubidaria {Thamnocnidia) fenella, Agassiz, a new species of tlie Atlantic shores of North
America.

*

13. TUBULAUIA TENELLA, Agassiz.

Thamnocnidia texella, — Agassiz, Cont. Nat. Hist. U. S., vol. iv, p. 275, pi. xxii, figs.

21 — 30. Alex. Agassiz, Illus. Catal. N. A. Acal.,
p. 195.

" Although this species agrees so closely in nearly all its details with Thamnocnidia {Ttihu-
laria) sjiecfabilis, it has a very different habitat ; it is never found with its congener in brackish
water, but always in the open ocean, among rocky pools. It is a very delicate, graceful animal,
and much the smallest of our Tubularians, having about half the size of Thamnocnidia spcctabiJis
or Pari/pha crocea. It branches very irregularly, loosely, and openly, with a stem of uniform
thickness throughout, about as large as a common sewing needle, or, to be more exact, one
fiftieth of an inch in diameter. The medusoids have been observed in January, -July, August,
and December." — Agassiz.

Habitat . — Kocky pools in the open ocean.
Locality. — Massachusetts Bay, Agassiz.

The above is the only account Agassiz has given us of this species, which he refers to his
genus Thrnnnocnidia. His remarks do not contain sufficient detail to admit of the construction
of a technical diagnosis, and I have, therefore, as in the preceding species, been obliged to content
myself by simply quoting his words, and referring to the beautiful figures by which the external
characters of the species and certain structural details are represented.

416 TUBULARIA CROCEA.

*^* Ik TuBULARIA PACIFIOA, Alhucm.

Under the name oiThamnocnidia tiihularoides,?i species of Tuhiilaria is also recorded by Mr.
A. Agassiz, but without sufficient detail for a technical diagnosis. The following paragraph
contains all he says of it :

" This species grows in clusters, which, at first sight, would readily be mistaken for a species
of true Tubiilaria, on account of the great diameter of the stem and the large size of the head.
The structure of the proboscis, however, shows plainly that it is a genuine Thamnocnidia, which
can at once be distinguished from its Eastern congeners by the stoutness of the stem and size of
the head, surrounded by as many as thirty and even forty tentacles, in large specimens. Found
growing profusely on the bottom of the coal barges which bring coal from Benicia to the Pacific
Mail Steamship Company's steamers at San Francisco." (A. Agassiz's ' lUustr. Catal. N. A.
Acselephse,' p. 196.)

For reasons already mentioned I regard Thamnocnidia as identical with Tuhularia ; and as
the specific name of tubularoides can scarcely be retained with the generic name of Tuhularia, I
have here ventured to substitute for it that ol pacifica.

The species possesses an interest as being a Pacific representative of the Atlantic forms of
Tuhularia. In the description just quoted from Mr. A. Agassiz allusion is made to "the strac-
ture of the proboscis, as proving the species to be " a genuine Thamnocnidia!' I do not know to
what peculiarity this statement refers. In the hydranths of such European species as Prof.
Agassiz would refer to his genus Thamnocnidia, there are certainly no characters which would
justify a separation from Tuhularia.

^^^^^ Sub-genus Parypha, Agassiz.

Sporosacs without evident gastro-vascular canals ; apical processes in female sporosac
laterally compressed.

*^* 15. TuBULAMA CKOCEA, AgassK.

P.iRYPH,\ CROCEA, — A(jassiz, in Contr. Nat. Hist. U. S., vol. iv, p. 249, pi. xxiii. A. Ayassiz,
in Uiustr. Catal. N. A. Acal., p. 195.

TEOPHOSOME. — Hydrocaulus consisting of bunches of stems, which are " at
the base very much contorted, irregularly branched, and densely intertwined," each
stem ascending singly from this entangled mass to a height of from two and a half to
three and a half inches ; pebisajrc " wavy or slightly nodose, or faintly ringed at
irregular distances." Htdrantus with each tentacular circlet consisting of about
twenty-four tentacles, disposed in a single verticil.

TUBULARIA CRISTATA. 417

GONOSOME. — GoxopuoRES in ten or twelve pendulous racemes, which are
disposed in two or three rows one over the other, and which surpass in lenglli tlie
proximal tentacles of the hydranth ; female gonophores with from six to ten laterally
compressed crest-shaped, hollow tentaculiform processes ; male gonophores destitute of
processes ; radiating and circular canals absent.

Colour. — Perisarc light yellow ; generative mass deep yellow, on both male and female
gonophores.

Deveiopiiiciit of Gonosoiiie. — Sunnner and autumn.

Habitat. — In brackish water, attached to floating timber, &c.

Locality. — Boston Harbour, Prof. Agassiz.

The present species is referred by Agassiz to one of those separate genera into which he has
broken up the genus Tuhularia of the European zoologists. He here substitutes for Tuhularia
his new genus Pari/pha, but, in consequence of the want of a definite diagnosis in his description of
the hydroid, it is by no means easy to perceive the exact characters which he would select as
those entitling it to the rank of a separate genus. He appears, however, to find them in the
absence of gastro-vascular canals, by which the sporosacs of the present species are distinguished
from the sporosacs of those species which, along with the European Tuhularia indivisa, he would
still retain in the old genus Tuhularia ; and in a peculiarity of the tentacida-like processes with
which the female sporosac is crowned, and which he believes sufficient to separate the present
species from those which he would place, along with the European Tuhularia, lari/iir, in his new
genus Thamnocnidia .

Tuhularia crocea has not yet been found on the European side of the Atlantic. It is
described by Agassiz as growing in great luxuriance on floating timbers in Boston Harbour,
where the sea water, even at high tides, contains a large admixture of fresh water from the river.
It affords the subject of one of the beautiful plates in the ' Contributions to the Natural History
of the United States.'

*^* 16. TuBULAKTA CRISTATA, JP Cradi/.

TuBULARiA CRISTATA, — M'Cradij, in Proc. Elliott Soc. of Nat. Hist., vol. i, p. 156.
Parypha CRISTATA, — Agussiz, Contr. Nat. Hist. U. S., vol. iv, p. 342.

TROPHOSOME. — Htdrocadlus attaining a height of from two inches to nearly
three inclies.^ Hydranths slender, the basal portion " not much exceeding in
diameter the width of the portion above it ;" twenty or more tentacles in the proximal

M'Ciady does not state wliether iiis Tubuluria crislata is a simple or a branched form.

418 TUBULARIA MESEMBRYANTHEMUM.

verticil, the distal circlet composed of at least two series, and containing eighteen or
more tentacles.

GONOSOME. — GoNOPHORES -with "eight rows of thread-cells, which run up the
outer surface like meridian lines," and with from four to eight laterally compressed
tentaculiform processes which surround a very distinct orifice at the distal end.^

Colour. — Ilydraiiths rose colour, cceiiosarc varying from yellow through reddish-yellow into
rose colour.

Development of Gonosome. — From March to September.
Habitat. — On rocks exposed to the ocean, near low-water mark.
Hathymetrical distribution — Laminarian zone.
Localiti/. — Sulivan's Island, South Carolina, M'Crady.

Though no figure of Tiibnlaria cnstata is given, and though some points of value in a specific
diagnosis — such as the simple or branched condition of the hydrocaulus, the erect or pendulous
condition of the gonophore clusters, and the presence or absence of gastro-vascular canals — have
been left unnoticed in M'Crady's description, there is yet sufficient detail to show that the species
is a well-marked one.

The peculiar bands of thread-cells, which extend in meridional lines along the outer surface
of the gonophore, present a striking character, and call to mind what would seem to be a similar
feature in the gonophore of Van Beneden's Tabularia Duiiioriieri, for which Agassiz constitutes,
as I believe wisely, a new genus, under the name of Ectojjleura.

The orifice in the summit of the gonophore of Tabularia cristata is occasionally so large as to
allow of the extremity of the manubrium being protruded through it, a fact also noticed by Agassiz
in his Tabularia (Paryphd) crocea.

The Tabularia cristata is also one of those species for which Agassiz constitutes his
genus Paryplta.

*^* 17. TUBULAUIA MESEMBRYANTHEMUM, Allinan.

TROPHOSOME. — Hydrocaulus consisting of a cluster of simple slender stems,
destitute of annulation, springing from a plexus of filiform tubes, and attaining a
height of about four inches ; ccenosarc forming a collar below the hydranth ; stem not
longitudinally striated. Hydraktus with from twenty to twenty-four tentacles in the

^ It is possible that these processes are coufiued to one sex, though no statement to that effect is
made in M'Crady's description of the species.

TUBULARIA MESEMBRYANTHEMUM.

419

proximal series, and with aljout twenty-four in tlie distal, where tliey arc disposed in
two closely approximate alternating verticils, forming a single circlet.

GONOSOME. — Gonophores in short, erect, dense clusters; destitute of gastro-
vascular canals ; female with eight compressed, crest-like, apical processes, which are
replaced in the male by four small rounded tubercles. Actinula destitute of oral
tentacles at the period of its liberation.

Fis. 84.

Fig. S3.

colour

Tuhularia Mesemhryanihemum.
Fig. 83. — A portion of a colony, natural size.
Fig. 84. — Magnified details. A. Hydrauth with gonophore-clusters j a, collar.

B. Female sporosac; the spadix projects through the apical orifice, and detached masses of the generative plasma are seen
lying free in the cavity of the sporosac.

C. Female sporosac, showing two tubular processes, a, a, sent off from the base of tlie spadix.

D. Male sporosac, the spadix projecting through its apical orifice.

E. Transverse section of the stem, a, perisarc ; i, ectoderm ; c, endoderm, sending off two opposite vertical laminx,
which unite in the axis and divide the cavity of the stem into two similar longitudinal chambers.

Cohitr. — Body of hyJranth and spadix of gonophore pale vermilion; perisarc stra'

Bcvehpmenf of Gonosome. — starch.
Habitat. — Attached to submerged rocks.
Bathpnetrical Dhtribidlon. — Zone of the Cystomra.
Locality. — Gulf of Spezia, G. I. A.

54

420 PROVISIONAL SPECIES.

Tuhularia mesembrijantliemum has the simple habit of T. indivisa, but with more slender
stems. It is a very beautiful species, with large delicately coloured hydranths. The processes
which crown the female sporosacs (Woodcut, fig. 84, b, c) arc laterally compressed, and possess
the form which Agassiz regards as characteristic of his genus Faryplia. They surround a
well-marked orifice in the summit of the sporosac, and the spadix may usually be seen protruding
through this in the mature sporosac, a condition which it appears is also common in the
American species which Agassiz would refer to Parypha.

The male sporosac (d), instead of possessing the eight flattened processes of the female, has
four small round tubercles, bearing a close resemblance to the apical processes of Tubidaria larynx,
&c. There is here, also, a well-marked orifice, through which, as in the female, the summit of
the spadix may often be seen to be protruded.

A remarkable fact noticed in the present species is the tendency of the spadix in the female
gonopbore to throw out from its base lobes which contain a continuation of its cavity, and which
not unfrequently become so elongated as to assume the appearance of radiating canals (c, a, a).
They were observed to vary in number from one to three, but were not invariably present. I
never met with them in the male.

When a transverse section (f) of the stem is carefully made, its cavity is found to consist of
two great longitudinal chambers separated from one another by a partition formed by two plate-like
processes of the endoderm, which project from two diametrically opposite longitudinal lines until
they meet in the axis. The walls of these chambers are clothed with vibratile cilia, so minute as
to be with difficulty detected, and in this respect contrasting strongly with the long conspicuous
cilia which clothe the canals in the stem of Tubidaria indivisa.

In making sections of the living stem, distinct evidence was afforded of the irritability of the
endoderm, which might often be seen immediately after the act of section to encroach upon the
cavity of the stem at the inner edge of the surface of section to such an extent as nearly to shut it
in. After a time, however, this encroachment of the endoderm recedes and fully exposes the
double cavity of the stem.

Tubidaria mesembryanthemuni was obtained in consideraljle abundance in the more sheltered
parts of the Gulf of Spezia, where it occurred growing upon rocks at a little distance below the
lowest tide level.

PROVISIONAL SPECIES.

TuBULARiA ASPEUA, Allman.

Under the provisional name of Tubidaria aspera, I would indicate a hydroid whose dried
stems are among the collection in the Jardin des Plantes. The soft parts have entirely dis-
appeared, so that nothing can be seen either of the hydranths or of the gonosome, and a satis-
factory determination is accordingly impossible.

The stems form dense tufts, contorted and entangled below, and then, becoming free, attain
a height of about three inches, and a thickness of about -^th of an inch. They are mostly

HYBOCODONID.E. 421

simple, but occasionally give off a branch. The perisarc is of a papyraceous consistence, but its
most characteristic feature is the presence upon the larger stems of slightly elevated annular
ridges which follow one another at short and rather irregular intervals. They are very distinct
under a low magnifying power, and give rise to a sensation of roughness when the dry stem is
drawn through the fingers. They appear to indicate successive periods of growth, similar to
what takes place in Tahularia indivisa, after each casting of the hydranth, where, however, the
indications of periodic growth are neither so numerous within the same space nor so distinct as
in the present species.

A note appended to the specimen informs us that it was brought from Coijiiiuibo by
M. Gaudechaud.

EYBOCOBONIDM.

TROPHOSOME. — Htdrocaulus developed, invested by a chitinous perisaec.
HTDK.iNTHS witli a proxlmal and a distal set of filiform tentacles.

GONOSOME. — GoNOPHORES medusiform planoblasts.

HYBOCODON, Agassis.

Name. — ^From iljSoc, hump-backed, and kwcwv, a bell, in allusion to the unsymmetrical form
of the planoblast.

TROPHOSOME. — HYDROPnTTON consisting of a simple (or branched ?) htdro-
CAULTJs, rooted by a filiform htdeorhiza. Hxdranths flask-sbaped, abruptly distinct
from tbe supporting liydrocaulus ; the proximal set of tentacles long, and forming a
single verticil, tbe distal set short, and arranged in two distinctly separated verticils.

GONOSOME. — Planoblasts springing from the body of the hydranth, between
the proximal and distal sets of tentacles. MsDrsA at the time of liberation with a
deep-belled umbrella, simple-mouthed manubrium, four radiating canals, and with
only one marginal tentacle, which is prolonged from the distal extremity of one of
the canals, and is furnished with a bulbous base destitute of distinct ocellus.

The genus Ihjhocodon Avas established by Agassiz for a large and beautiful tubularian from

422 HYBOCODON PROLIFER.

Massachusetts Bay. It is a strongly marked genus, while its well-developed medusae, each with
its single tentacle and unsymmetrical bell-margin, indicate a decided approach to Conjmorpha.

Hybocodox proltter, Agassiz.

Hybocodon PROLIFER, — Agassiz, Contr. Nat. Hist. U. S., vol. iv, p. 24.3, pi. xxiiia, figs.
10, 11, and pi. XXV. A. Agassiz, Illustr. Catal. N. A.
Acal., p. 193.

TROPHOSOME.— Hydrocaulus consisting of solitary or sparingly aggregated
stems, wliich attain a height of about two inches, the stems gradually enlarging from
the base, until just below the hydranth they attain a thickness of one sixteenth of an
inch; PERISAEC destitute of annulations, except towards the summit of the stem,
where it becomes dilated and furnished with annular constrictions ; ccenosarc with
longitudinal orange-red striae. Hydranth with its two distal verticils composed each
of about sixteen tentacles, the more distal of the two consisting of tentacles which
are about half the length of those forming the other.

GONOSOME. — Umbrella of medusa with five orange-red granular bands, which
extend upon its outer surface from the codonostome to within a short distance of the
apex, two of these bands lying one on each side of that radiating canal which
corresponds to the solitary tentacle, the others lying one over each of the three
remaining radiating canals; bulbous base of the marginal tentacle large and
proliferous ; tentacle smooth for some distance from the base, and then to its extremity
covered with annular groups of thread-cells.

Colour. — Deep orange-red.

Development of gonosome. — January.

Habitat. — In pools of pm-est sea water at low-water mark.

Batliymetrical didribution. — Laminarian zone.

Locality. — Massachusetts Bay, Agassiz.

Hobocodon jirolifer, the only representative yet discovered of its genus, forms the subject of
one of the beautiful plates in Agassiz's ' Contributions to the Natural History of the United
States.' One of its most striking features is found in the constant tendency of its medusae to
multiply themselves by the formation of buds which are developed from the marginal termina-
tion of one of the radiating canals, that, namely, which is continued into the solitary tentacle.
The buds are produced in clusters from this point, and when each attains a certain stage of
maturity, it gives rise in the same way, and from the corresponding point of its radiating canal
to a similar brood of medusa-buds.

ECTOPLEURA. 423

Agassiz has shown that tlie gcnciMl orange colour of the stem is produced by longitudinal
bands of pigment cells upon the inner wall of the coenosarcal cavity. He has further shown
that this wall forms ridges which project into the cavity of the stem, but that the cavity itself,
unlike that of Tuhularia indivisa, is single and continuous, while its walls, except on the bands
of pigment cells, are clothed with vibratile cilia.

ECTOPLEURA, Agassis.

Name. — From kruc, on tlie outer side, and irXivpa, a rib, in allusion to the prominent
longitudinal ribs of the planoblast.

TuBULAUiA, — Van Beneden.

TROPHOSOME.— Hydrocaulus filiform, rooted. Htdeanths flask-shaped,
abruptly marked off from the supporting stalk ; tentacles of the proximal set longer
than those of the distal.

GONOSOME. — Planoblasts on branched peduncles, which are borne on the body
of the hydrantli between the proximal and distal verticils of tentacles. Medusa at
the time of liberation with a nearly spherical umbrella and simple-mouthed manu-
brium ; four radiating canals and four marginal tentacles ; no distinct ocelli ; umbrella
furnished with eight prominent longitudinal ribs, formed of linear series of thread-
cells.

To his genus Edopleitra, Agassiz, as has been already said, refers the Tuhularia Dumortieri
of Van Beneden. In thus separating Van Beneden's Tuhdarian from the true Tulularm,
Agassiz seems to me to be fidly justified, the phanerocodonic condition of the gonophore
affording in itself an- important generic character. I cannot, however, so easily assent to the
correctness of associating with it in the same genus the Sarsia pulcheUa of Forbes, the Sarsia
turricula of M'Crady, and the Sarsia nodosa of Busch.^ These hydroid medusae are very
different from the medusa of Van Beneden's Tuhularia Dumortieri, while one of them, Sarsia
turricula, has been traced by M'Crady, if not with absolute certainty, at least with high proba-
bility, to a coryniform trophosome."

It is possible that the medusa named Ectopleura ocracea by Mr. A. Agassiz,' has been
rio-htly referred to this genus ; but as we know nothing of its trophosome, its generic determi-
nation cannot be regarded as otherwise than provisional.

' See Agassiz in ' Cont. Nat. Hist. U. S.,' vol. iv, p. 343.
- ^I'Crady, in ' Gyiiophth. of Charleston Harbour.'
» ' Illustr. Catal./ p. 191, fig. 320.

424 HYDROLARID^.

EcTOPLEUKA DuMOKTiERi, Van Bened.

TuiiULARiA DuMORTiERi, — Van Beneden, Reclierches sur I'Embryogenie des Tubulaires, p. 50,

pi. ii ; Rech. sur Fauue litt. de Belgique, p. 111.
EcTOPLEURA DuMORTiERi, — Aycissiz, Coiit. Nat. Hist. U. S., vol. iv, p. 343. Allman, in

Ann. Nat. Hist, for May, 1864. Hincks, Brit. Hydr.

Zooph., p. 124, pi. xxi, fig. 4.

TROPHOSOME. — Htdrocaulus attaining a Leiglit of about one inch, isolated,
slender, simple or sliglitly ramified ; perisarc with annular constrictions. Htdranths
comparatively large.

GONOSOME. — GoNOPHORES on short, erect, slightly branched peduncles.

Hahitaf. — Attnched to Flustras, the carapaces of crabs, &c.
LocaJ'Uy. — Coast of Ostend, Van Beneden ; Isle of IMan, Mr. Hincks.

I have never met with an example of the present genus, and have obtained the generic and
specific characters here given from M. van Beneden's description of his 'luhdaiia Dunwrtieri,
the only species of the genus yet discovered, at least with its trophosome. The Belgian zoologist
informs us that this hydroid is very abundant on the coast of Ostend, where it occurs ou Flustras
and Halodactylas, and on the backs of crabs, in the form of isolated stems, never forming tufted
masses such as we meet with in many other Tubularians.

HYDROLARIDJ^:.

TROPHOSOME. — Htdrocaultjs undcA^eloped. Htdranths with but two ten-
tacles, which are fiUform and spring from one side of the base of the hypostome ;
mouth with tw^o lip-like lobes.

GONOSOME. — GoxoPHORES raedusiform planoblasts, with six simple radiating
canals and simple marginal tentacles.

In the enumeration of genera on page 241 the family to which Lar has been provisionally
referred is named "Laridae." This name, however, cannot be retained, being already in use for

LA.R SABELLARUM. 425

a well-known family of l)ir(ls. It is difficult to form aiiotlicr in accordance witli the principles
which have rcgidated the nomenclature of the present monograph. The nearest approach to
uniformity will, perhaps, be attained by the use of the name Hydrolarid*, which I have
accordingly adopted.

LAR, Gosse.

Name- — Yxoxa Lar, a household god, in allusion to the mode in which the only known
species is associated with Sabella.

TROPHOSOME. — Htdroehiza a creeping filiform stolon covered with a peeisarc.
Hydranths fusiform ; hypostome separated by a constriction from the body ; oral
lobes in the form of two opposable plates.

GONOSOME. — Planoblasts borne on blastosttles which spring from the
hydrorhiza and terminate distally in a globular cluster of thread-cells. Medusa at
time of liberation with a sub-globular umbrella; manubrium moderately large,
destitute of oral tentacles ; marginal tentacles six, with bulbous bases destitute of
ocelli.

The trophosome of this remarkable genus was described some years ago l)y Mr. Gosse.
Nothing, however, of its gonosome was known until Mr. Ilincks had the good fortune to obtain
it on the northern coast of Devonshire. To him, therefore, we are indebted for rendering us
acquainted with the missing element, which was needed for a complete diagnosis of the genus,
and which can alone remove it from the domain of provisional genera, to which it would
otherwise have to be relegated.

Lab, SABELLARUM, Gosse.

Lar SABELLARUM (trophosoiiie), — Gosse, in Trans. Linn. Soc, vol. xsii, 1857, p. 113, pi. xx.

(Trophosome and gonosome), Hincks, in litteris,
July 5th, 1872.

TROPHOSOME. — Htdraxtos attaining a height of about Toth of an inch,
scattered at short intervals on the creeping hydrorhiza ; hypostome witli a patch of

426 L.\R SABELLARUM.

imbedded thread- cells near its summit ; tentacles smooth, and when extended about
as long- as the body of the hydranth.

GONOSOME. — Blastostyles cylindrical, slender. Planoblasts springing from
a point a little above the middle of the blastostyle in a sub-verticillate cluster of three
or four; manubrium about half as long as the vertical diameter of the umbrella-
cavity, with a constriction just above its oral extremity ; margin of umbrella with a
minute tubercle at the middle point between every two tentacles.

Development of Gonosome. — July.

Habitat. — Attached to the tubes of Sahellce round tlie orifice.

Batliymetrical distribution. — Coralline zone.

Locality. — North coast of Devonshire, Mr. Gosse and Mr. Ilincks.

Under the name of Lar sahellannn Mr. Gosse described a certain enigmatical organism which
made its appearance in his aquarium, growing round the orifice of the tube of one of the
sea-worms [Sabella). He tells us that from a creeping network of filaments which extended
round the mouth of the Sabella tube there sprang numerous irregularly fusiform bodies, each
terminating distally in a head-like lobe, immediately below which were two long tentacles. The
terminal lobe is described as capable of opening itself out by the separation of two broad
flattened lips, which then diverge from one another in the manner of " the leaves of a half-
opened book."

The zooids of the colony are described as singularly energetic in their motions, and
Mr. Gosse gives a graphic account of the various forms and attitudes assumed by them. He
informs us that " about twenty bodies having a most ludicrously close resemblance to the human
figure, and as closely imitating certain human motions, were standing erect around the mouth of
the tube, when the Sabella had retired into its interior, and were incessantly tossing about their

arms in the most energetic manner The head-lobe moved to and fro freely on the

neck; the body swayed from side to side, but still more vigorously backward and forward,
frequently bending into an arch in either direction, while the long arms were widely expanded,
tossed wildly upward and then waved downward as if to imitate the actions of the most tumul-
tuous human passion."

A characteristic figure, in which these various attitudes are represented, accompanies the
description.

This description by Mr. Gosse contained, up to the present time, all that we knew of the
singular organism which formed the subject of it ; and though we had no information regarding
its gonosome, it was yet evident that Lar sabellaniiii was the trophosome of a hydroid which,
however anomalous, had (as Mr. Gosse himself recognised) its nearest immediate relations with
the GymnoUastea.

Just, however, as the last sheet of the present work was on the point of going to press, I
received a letter from Mr. Hincks informing me that he had just dredged up a colony of Lar
sabellarum with its gonosome on the coast of North Devon. Mr. Hincks did not fail to profit
by the opportunity thus afforded of making a careful study of the animal, and it is from his

LAR SABELLARUM. 427

description, accompanied by a bcautifnl drawing, that I have liere been enabled to sni)plcmcnt
in important details Mr. Gosse's account of one of the most singular and distinct of the
Hydroida.

The fact of the radiating canals in the planoblast being six in number constitutes an unusual
though by no means unprecedented condition of the gastro-vascular system. The small tubercle
which is interposed between every two tentacles on the umbrella-margin, and which Mr. Hincks
informs me contains minute bodies like thread-cells, is probably the origin of what in the adult
medusa would become an interradial marginal tentacle. Mr. Hincks has further observed that
the gonophore is naked, no ectotheca being present at any period of its development, a condition
in which it resembles Clavatella and some other hydroids with undeveloped or naked hydrocaulus.

It is, however, in the trophosome that the most striking characters are to be met with.
The singularly unsymmctrical form of the hydranth, with its tentacles reduced to two, and thrown
altogether to one side, and the two lip-like lobes with which the mouth is provided, are characters
so unique as to necessitate the separation of Lar from all other known hydroid families. So far
as I can judge from the description and figure, the mouth is also situated laterally on the
hypostome, being, along with its two lips, directed towards the same side as that which carries
the tentacles.

55

BIBLIOGRAPHY OF THE HYDROIDA.

In the following Catalogue the general arrangement is chronological. When, however, the
same author has more than one work assigned to him, his publications are placed in a sino-lc
series under his name, without reference to their exact chronological sequence in the general
ari'angement : —

Leeuwenhoek, Ant. von. — Part of a Letter from Mr. Antony von Leeuwenhoek, F.R.S., concerning

green Weeds growing in water, and some Animalcula [Hydra, &c.] found about them. ' Phil.

Trans.,' 1703.
Reaumur, Rene-Antoine Forchaud de. — Histoire des Insectes, Paris, 1742.
JussiEU, Bernard de. — Examen de quelque Productions Marines qui ont ete mises au nombre des plantes,

et qui sont I'ouvrage d'une sorte d'Insectes de Mer. 'Mem. de I'Acad. Roy. des Sciences.'

Paris, 1742.
Trembley, Abraham. — Meinoircs pour servir a I'Histoire d'un genre de Polypes d'eau douce a bras en

forme de cornes. Lcyden, 1744.
Ellis, John. — An Essay towards a Natural History of the Corallines and other Marine Productions of

the like kind commonly found on the Coasts of Great Britain and Ireland. To which is added

the Description of a large Marine Polype, taken near the North Pole by the Whale-fishers in the

Summer of 1753. London, 1755.
Bastek, Jobe. — Opuscula subseciva. Harlenii, 1762.
Pallas, Petr. Sim. Elenchus Zoophytorum sistens Generum Adumbrationes generaliores et specierum

cognitarum succinctas descriptiones cum selectis Auctorum Synonymis. Hagaj-Comit., 1766.
Spicilegia Zoologica quibus nova; imprimis obscurae Animalium species iconibus, descript. atque

commentariis illustrantur. Berolini, 1767 — 1780.
FoRSKAL, Petrus. — Descriptiones Animalium quEe in Itinere Orientali observavit. Havniee, 1775.

Icones Rerum Naturalium quas in Itinere Orientali depingi curavit Petrus Forskal. Havnia^,

1776.
MuLLER, Otho Fredericus. Zoologia Danica sen Animalium Dania; et Norvegia; variorum et minus

notorum Icones. Hafnia?, 1777 — 1780.
Cavolini, Filipo. — Memorie per servir Alia Storia dei Polypi Marini. Napoli, 1785. Translated in

1813 into German by Sprengel.
SoLANDER, Daniel. — The Natural History of many Curious and Uncommon Zoophytes collected from

various parts of the Globe by the late John Ellis, F.R.S. ; systematically arranged and described

by the late Daniel Solander, M.D., F.R.S. London, 1786.
EsPER, Eugenius Johann Christoph. — Die Pflanzenthiere in Abbildungen nach der Natur mit Farben

erleuchten nebst Beschrcibungen. Nuremberg, 1791 — 1797.
Peron, F., et C. A. Lesueur. — Notions preliminaires sur les Meduses. ' Ann. du Museum,' 1809.

Tableau des Caracteres gen^riques et specifiques de toutes les Especes de Meduses counues jusqu' a

ce jour. ' Ann. du Museum,' 1809.
L.amarck, J. Babt. P. Ant. de Monet. — Histoire Naturelle des Animaux sans Vertebres. Paris, 1815 —

1822. 2e Edit, par MM. Deshayes et Milne-Edwards. Paris, 1835—1843.
Lamouroux, J. V. F. — Histoire des Polypiers Coralligenes flexible vulgairement nommes Zoophytes.

Caen, 1816.
Exposition Methodique des genres de TOrdre des Polypiers. Paris, 1821.
CuviER, Geo. Leop. Chr. Fred. Dajob. — Le Regne Animal dibtribuc d'apres son Organization, pour

430 BIBLIOGRAPHY.

servir de base a I'llistoire Naluielle des Animaux et d'lntroduction a I'Aiiatomie Comparee.
Paris, 1817.
AuDOUiN, J. Victor, et H. Milne-Edwards. — Resume des Recherches sur les Animaux sans Vertebres

faites aux Isles Chausey. 'Ann. des Sciences Nat.,' 1828.
EscHSCHOLTZ, Job. Friedr. — System der Acalepben. Berlin, 1829.
Rapp, Wilhelm Ludwig v. — Ueber die Polypen in Allgemeinen und die Actinien insbesondcre. Weimar,

1829.
Ehtienbekg, Chr. G. — Beitrag zur pbysiologisthen Kcniitiiiss der Corallentbiere im Allgemeinen und

besonderes des Rothen Meeres. ' Abhandlungen der Berlin. Akad.,' 1832.
AVagner, Rudolf. — Neue im Adriatisclien Meere gefundene Art von nacktem Armpolypen. Isis,

1833, iii.
Blainville, Henri Marie Uucrotay de. — Manuel d'Actinologie ou de Zoopbytes. Paris, 1834.
Lesson, Rene Primivere — Mem. sur la famille des Beroides (Beroidese, Less.) ' Ann. S. nat.' (2e Ser.),
V, 1836, p. 235.
Hist. Nat. des Zoopbytes Acalepbes. Suites a Buffon. Paris, 1843.
LovEN, S. L. — Beitrag zur Kenntniss der Guttungen Campanularia und Syncoryne. ' Midler's Archiv,'

18.37.
Saks, Martin. — Bidrag til Sordyrenes Naturhistorie. Bergen. 1829. (Transl. in Isis, 1833.)

Beskrivelser og Jagttagelser over nogle moerkelige eller nye i Havet ved den Bergenske Kystlevende
Dyr af Polypernes, Acalepbernes, Radiaternes, Annelidernes og Molluskernes Classcr. Bergen,
1835.
Fauna littoralis Norvegia?. Heft i. Cbristiania, 1846.

Bidrag til Kundskaben om Middelbavets Littoral-Fauna ' Nyt. Mag. Naturvid,' ix, 1857.
Om Ammestegten Corymorpba og dens arter samt de af disse opammede Meduser. ' J'orbandl. i
Vid. Selsk. i Cbristiania,' 1859. Transl. in 'Ann. Nat. Hist.,' Nov., 1861.
Brandt, J. F. — Prodromus descriptionis Animalium ab H. Mertensio in Orbis Terrarum circumnaviga-
lione observatorum. ' Recueil des Actes de la Seance publique de I'Academie Imperiale des
Sciences de St. Petersbourg,' 1833-34.
Ausfiibrlicbe Beschrebung der von C. H. Mertens auf seiner Weltumsegelung beobacteten Schirm-
quallen. ' Mem. de I'Acad. Iniper. des Sciences de St. Petersbourg,' vi ser. Sci. Nat., tom. ii.
Petersbourg, 1838.
Krauss, C. F. F. — Beitrage zur Kenntniss der Corallineen u. Zoopbyten der Siidsee. Stuttgart, 1837.
Johnston, George. — A History of Britisli Zoopbytes. Edinburgb, 1838.

Id. Second Edition. 1847.
Couch, J. — Fauna of Cornwall. London, 1838.

Hassall, Artb. H. — Catalogue of Irisb Zoopbytes. 'Ann. Nat. Hist.,' vol. vi, 1841.
Supplement to a Catalogue of Irisb Zoophytes. ' Ann. Nat. Hist.,' vol. vii, 1841.
On the Phosphorescence of Zoophytes. ' Ann. Nat. Hist.,' 1842.
Remarks on Three Species of Marine Zoophytes. 'Ann. Nat. Hist.,' 1843.
Hassall, A. H., and J. Coppin. — Descriptions of Three Species of Marine Zoophytes. ' Trans. Micr.

Soc.,' vol. iii, 1852.
Steenstrup, Job. Japetus. Ueber den Generationswechsel oder die Fortpflanzung und Entwicklung
durch abwechselnde Generationen, eine eigenthiimlicbe Form der Brutflege in den niedern
Thierclassen. Uebers. vonLorenz. Kopenh., 1843. Also translated for the Ray Society by Busk.
London, 1845.
Untersuchungen iiber das Vorkommen des Ilermapliroditismus in der Natur. Transl. in ' Froriep

Notizen,' i, 1847.
En ny og tropisk art af smaagoplernes Ammesltegt : Corymorpba. (C. Januarii, Steenst.) ' Kio-
benhavn, Vidensk. Meddel.,' 1854.
Dujardin, Fee. — Observations sur un nouveau genre de Medusaires (Cladonema) provenant de la Meta-
morphose des Syncorynes. 'Ann. des Sci. Nat.,' 2e ser., 1843.
Memoire sur le Developpemcnt des Medusaires et des Polypes Hydraires. ' Ann. des Sci. Nat.,'
1845.

BIBLIOGRAPHY. 431

QuATREFAGES, A. lie. — IMomoire sur rEleutherie dichotome, nouvcau genre de Rayonnes voisin des
Hydres. ' Ann. des Sci. Nat.,' 2e ser. 1843, vol. xviii.
Memoire sur la Synhydre parasite. 'Ann. des Sci. Nat.,' 2e s^r. 1843, vol. xx.
Gray, John Ed. — Fauna of New Zealand (Sertulariada;), in ' Dieffenbach's Travels in New Zealand.'

London, 1843.
Meneghini, Giuseppe. — Osservazzioni sull' ordine delle sertulariee della classe de' Polipi. ' Mem. del'

J. R. Istituto Veneto,' vol. 2, 1845.
13ENEDEN, P. J. van. — Mem sur les Campanulaires de la cote d'Ostende consideres sous le rapport
physiologique, embryogenique et Zoologique. ' Nonv. Mem. de 1' Acad, de Brux.,' torn. 17, 1844.
Recherches sur I'Embryogenie des Tubulaires, et rhistoire naturelle des differents genres de cette

famille qui habitcnt la cote d'Ostende. ' Nouv. Mem. de I'Acad. do Brux.,' toni. 17, 1844.
Sur la reproduction des Campanulaires. ' Bull, de I'Acad. de Belgique,' 1847.
Observations relatives a la reproduction de divers Zoophytes, &c. ' Compte Kend.,' torn, xlix,

1859.
Recherches sur la Faunc littorale de Belgique (Polypes). ' Mem. de I'Acad. Roy. de Belgique,'
1867.
Forbes, Edward. — Contributions to British Actinology. ' Ann. Nat. Hist.,' vol. vii, 1841.

On the Morphology of the Reproductive system in the Sertularian Zoophytes. 'Ann. Nat. Hist.,'

1844, vol. xiv.
On the Pulmograde Medusae of the British Seas. ' Rep. Brit. Assoc.,' 184G.
A Monograph of the British Naked-eyed Medusae. ' Ray Society.' London, 1848.
Frey, Henr., und Leuckaut, Rud. — Beitriige zur Kenntniss wirbelloser Thiere. Braunschweig,

1847.
Dalyell, Sir John Graham. — Rare and Remarkable Animals of Scotland. London, 1847.
Desor, E. — Hydroids from Nantucket. ' Proc. Bost. Soc. Nat. Hist.,' vol. iii, 1848.

Sur la generation medusipare des Polypes Hydraires. 'Ann. des Sci. Nat.,' 3 ser., tom. 12, 1849.
On the Development of Campauularia. ' Proc. Bost. Soc. Nat. Hist.,' vol. iii, 1849.
On a Syncoryne from Boston. ' Proc. Bost. Soc. Nat. Hist.,' vol. iii, 1849.
Carpenter, W. B. — Review of various works on the Development and Metamorphoses of Zoophytes.
' British and Foreign Medico-Chirur. Review,' Oct., 1848.
Principles of General and Comparative Physiology. London, 1851.
VoGT, Carl. — Zoologische Briefe. Fraukfurt-am-M., 1851.

Huxley, Th. H. — An account of Researches into the Anatomy of the Hydrostatic Acalepha?. ' Brit.
Assoc. Reports for 1851.'
On the Anatomy and Affinities of the Family of the Medusae. ' Phil. Trans.,' 1849.
Upon Animal Individuality. ' Proc. Roy. Institut.,' 1851 — 54.
Lectures on General Natural History. ' Medical Times,' 1856—57.

The Oceanic Hydrozoa. A description of the Calycophoridoe and Physophoridie observed during
the voyage of H.M.S. "Rattlesnake" in the years 1846 — 50, with a general introduction.
' Ray Society.' London, 1859.
Notes on Medusae and Polypes. 'Ann. Nat. Hist.,' 1850.
KoLLiKER, Albert. — Die Schwimrapolypen oder Siphouophoren von Messina. Leipzig, 1853.

Icones HistiologieEe. Leipzig, 1864 — 65.
Leuckart, Rud. — Zoologische Untersuchungen. Giessen, 1853.

Nachtrage und Berichtigungen zu dem ersteu Bande von J. van der Hciven's Handbuch der

Zoologie. Leipzig, 1856.
Bericht iiber die Leistungen in der Naturgeschichte der niederen Thiere. ' Wiegmann's Archiv,'

1854, &c.
Ueber den Polymorphismus der Individuen, oder die Erscheinungen der Arbeitstheilung in der

Natur. Giessen, 1851.
Beitrage zur Kenntniss der Medusa-Fauna von Nizza. ' Wiegmann's Archiv,' xxii, 1856.
MuMMEEY, J. B. — On the Development of Tubularia indivisa. ' Quarterly Journ. of Micros. Science,'
1853.

432 BIBLIOGRAPHY.

Gegenbaub, Carl. — Zur Lelire vom Generationswechsel unci der Fortpflanzung bei Medusen und

Polypen. Wiiizburg, 1854.
Versuch eines Systemes der Medusen. ' Zeilschr. fiir wissen. Zool.,' 1857.
Grundziige der Vergleichenden Anatomie. Leipsig, 1859.
HiNCKs, Thomas. Notes on the Reproduction of the Campanulariadrp. ' Ann. Nat. Hist.,' 1852.
Further notes on British Zoophytes, with descriptions of new species. 'Ann. Nat. Hist.,' 1853.
On Clavatella, a new genus of Corynoid Polypes and its Reproduction. ' Ann. Nat. Hist.,' 1861.
A Catalogue of the Zoophytes of South Devon and South Cornwall. ' Ann. Nat. Hist./ 1861

—62.
On the Production of similar Gonozooids by Hydroid Polypes belonging to different Genera.

'Ann. Nat. Hist.,' 1863.
On some new British Hydroids. 'Ann. Nat. Hist.,' 1863.
On new British Hydroida. 'Ann. Nat. Hist.,' 1866.
On the Development of the Hydroid Polypes, Clavatella, and Stauridia; with remarks on the

relation between the Polype and the Medusoid and between the Polype and the Medusa.

' Brit. Assoc. Rep.,' 1861.
A History of the British Hydroid Zoophytes. London, 1868.
Supplement to a Catalogue of the Zoophytes of South Devon and South Cornwall, with descriptions

of new species. ' Ann. Nat. Hist.,' 1871.
GossE, P. H. — A Naturalist's Rambles on the Devonshire Coast. Loudon, 1853.
Tenby : a Sea-side Holiday. London, 1856.

On a new form of Corynoid Polypes (Lar Sabellarum). ' Trans. Lin. Soc.,' 1859.
Allman, Geo. James. — On the Anatomy and Physiology of Cordylophora. ' Phil. Trans.,' 1853.

On the Structure of the Reproductive Organs in certain Hydroid Polypes. ' Proc. Roy. Soc.

Edinb.,' 1857.
Additional Observations on the Morphology of the Reproductive Organs in the Hydroid Polypes.

' Proc. Roy. Soc. Edinb.,' 1858.
Notes on the Hydroid Zoophytes. ' Ann. Nat. Hist.,' 1859.
On Laomedea Tenuis. ' Brit. Assoc. Rep.,' 1859.
Note on the Structure and Terminology of tlie Reproductive System in the Corynidje and Sertularidae.

' Ann. Nat. Hist.,' 1860.
Notes on the Hydroid Zoophytes : — 1. Onthe Sexual Zooid of Dicoryne conferta. 2. On Corymbogo-

nium capillare. ' Ann. Nat. Hist.,' 1861.
On the Generative Zooid of Clavatella. ' Brit. Assoc. Rep.,' 1862.
Notes on the Hydroida : — 1. On the Structure of Corymorpha nutans. 2. Diagnosis of new species

of Tubularida;, obtained during the Autumn of 1862 on the Coasts of Shetland and Devonshire.

' Ann. Nat. Hist.,' 1863.
On the Reproductive System in the Hydroida. ' Brit. Assoc. Rep.,' 1863.
On the Occurrence of Amoebiform Protoplasm and the Emission of Pscudopodia among the Hydroida.

'Ann. Nat. Hist.,' 1864.
On the Construction and Limitation of Genera among the Hydroida. 'Ann. Nat. Hist.,' 1864.
Note on the Structure of certain Hydroid Medusse. ' Brit. Assoc. Rep.,' 1867.
On a mode of Reproduction by Spontaneous Fission in the Hydroida (Schizocladium). ' Quart.

Journ. Mic. Science,' 1871.
Remarks on Prof. Schulze's Memoir on Cordylophora lacustris. ' Quart. Journ. Mic. Sci.,' Jan. 1872.
Thomson, Wyville. — Notes on some British Zoophytes. ' Ann. Nat. Hist.,' 1853.
Wright, T. Strethill. — On Hydractinia Echinata. ' Edinb. New Phil. Journal,' vol. v, 1857.

Observations on British Zoophytes (Clava, Eudendrium). 'Edinb. New Phil. Journ.,' vol. vi,

1857.
Observations on British Zoophytes (Laomedea acuminata, Trichydra pudica, Tubularia indivisa,

Coryne gravata, Stauridia producta ; reproduction by ova of Medusoid of Campanularia Johnstonij.

' Edin. New Phil. Journ.,' vol. vii, 1858.
On Monoecious Reproduction in Tubularia larynx. ' Edin. New Phil. Journ.,' vol. viii, 1858.

BIBLIOGRAPHY. 433

Observations on Britisli Zoophytes (Atractylis ; roprodiiction of Eudciulrium ranunim ; development
of Bourgainvillia; fixed medusoids of Lamoedea dichotoma; Sertularia fallax, Laomedca lacerata).
' Edin. New Phil. Journal,' vol. ix, 1859.
Observations on British Zoophytes (reproduction of Turris neglecta ; Coryne implexa, Bimeria
vestita, Garveia nutans, Goodsirea mirabilis, Eudendrium arbuscula). ' Edin. New Phil. Journ.,'
vol. x, 18.59.
Observations on British Zoophytes and Protozoa (Atractilis coccinea). ' Edin. New Phil. Journ.,'

vol. xiv, 1861.
Observations on British Zoophytes and Protozoa (Clava nodosa). ' Edin. New Phil. Journ.,'
vol. xvi, 1862.
Leidy, Jos. — Contributions towards a Knowledge of the Marine Invertebrate Fauna of the Coasts of Rhode

Island and New Jersey. ' Journ. Acad. Nat. Sc' Philad., 1855.
jNI'Crady, J. — Gymnophthalmata of Charleston Harbour. ' Proc. Elliot Soc. Nat. Hist.' Charles-
ton, 18.59.
BoECK, Chr. — Beskrivelse over en Tubularie fra Belsund paa Spitzbergen, Tubularia regalis. ' Forhaiidi.

Vid. Selsk.,' 1859.
Bkonn. H. G. — Die Classen und Ordnungen des Thier-Reichs (Actinozoa). Leipzig, 1860.
Lewes, G. H. — On a New British Species of Hydra. ' Ann. Nat. Hist.,' 1860.

Sea Side Studies at Ilfracombe, Tenby, the Scilly Isles, and Jersey. Edinburgh and
London, 1858.
Peach, Ch. W. — On the Zoophytes of Caithness. ' Brit. Assoc. Rep.,' 1859.

Murray, Andrew. — Descriptions of New Sertularida; from the Californian Coast. ' Ann. Nat.
Hist.; 1860.
On Sertularia Tricuspidata. 'Ann. Nat. Hist.,' 1860.
Houghton, W. — On the Hydra rubra of Mr. Lewes. ' Ann. Nat. Hist.,' 1860.
Keferstein, Wilh. — Und Ernst Ehlers. ' Zoologische Beitrage.' Leipzig, 1861.
Jaeger, G. — Ueber das spontane Zerfallen der Siisswasserpolypen nebst einigen Beraerkungen iiber

Generationswechsel. ' Sitzb. Akad. AV'ien.,' 1860.
Claparede, a. Rene Ed. — Beitrage zur Fauna der Schottischen Kliste. ' Zeitsch. f. wiss. zool.,' 1860.
Eeobactungen iiber Anatomic und Entwicklungsgeschichte wirbelloser Thiere an der Kiiste der
Normandie angestellt. Leipzig, 186.3.
Busk, George. — A list of Sertularian Zoophytes and Polyzoa from Port Natal, Algoa Bay, and Table
Bay, in South Africa ; with remarks on their geographical distribution and observations on the
genera Plumularia and Catenicella. ' Brit. Assoc. Rep.,' 1850,
Observations on some parts of the Anatomy of two species of Naked-eyed Medusa; [1848]. ' Trans.

Micr. Soc.,' iii, 1852.
Observations on certain points in the Anatomy of a species of Thaumantias [1849]. ' Trans.

Micr. Soc.,' iii, 1852.
An account of the Polyzoa and Sertularian Zoophytes collected in the voyage of the " Rattlesnake'
on the coasts of Australia and the Louisiade Archipelago. Voyage of H.M.S. " Rattlesnake."
Appendix, London, 1852.
On a new Hydroid Polype belonging to the genus Coi'dylophora, AUin., discovered by Senator
Kirchenpauer. ' Journ. Microsc. Soc.,' i, 1861.
Hodge, Geo. — -Contributions to the Marine Zoology of Seaham Harbour. ' Trans. Tyneside Nat. Field
Club,' 1861.
On a new Hydroid Zoophite (Podocoryne Alderi). 'Trans. Tyneside Naturalists' Field Club,' 1861.
Krohn, a. — Beobachtungen iiber den Bau und die Fortpflanzung der Eleutheria, Quatrcfaijes. ' Wieg-

mann's Archiv,' 1861.
MuLLER, Fr. — Cunina Kollikeri, Beitrag zur Naturgeschichte der QZginden, in Wiegmann's ' Archiv,'
1861.
Polypen und Quallen von Santa Catharina. Wiegmann's ' Archiv,' 1861.

Ueber die Ursache der Stromungen in der Leibeshole der Sertularien. Wiegmann's ' Archiv,' 1863.
Polypen und Quallen von santa Catharina die Formsvandlungen der Lyriope Catharinensis.
Wiegmann's ' Archiv,' 1859.

434 BIBLIOGRAPHY.

Agassiz, Louis. — On the Structure and Homologies of Radiated Animals, with reference to the systematic
position of the hydroid Polypi. * American Association Proc.,' 1849.
Contributions to the Natural History of the Acalephse of North America. Boston, ' Mem. Amer.

Acad.,' iv, 1850.
On the Morphology of the Medusae. ' Amer. Assoc. Proc.,' 18.50.
Classification of Polyps. ' Amer. Acad. Proc.,' iii, 1852 — 57.

Contributions to the Natural History of the United States of America, vols, iii and i v. Boston,
1860, 1802.
Greene, Joseph Reay. — On the Acalephse of the Dublin Coast, with descriptions of seven new naked-
eyed forms. ' Nat. Ilist. Review,' iv, 1857.
On the Morphology of the Hydrozoa with reference to the constitution of the sub-kingdom Coelen-

terata. ' Proc. Roy. Irish Acad.,' vii, 1858.
A Manual of the sub-kingdom Cculentcrata. London, 1861.
Agassiz, Alexander. — On the Acalephan Fauna of the southern coast of Massachusetts (Buzzard's Bay).
Boston, ' Proc. Nat. Hist. Soc.,' viii, 1861—62.
On the mode of development of the marginal tentacles of the free niedusffi of some hydroids.

Boston, ' Proc. Nat. Hist. Soc.,' ix, 1863.
Illustrated Catalogue of the Museum of Comparative Zoology at Harvard College. No. II ' North
American Acalepha,' Cambridge, U.S., 1865.
Agassiz, Mrs., and Alexander. — Sea-side Studies ; Marine Animals of Massachusetts Bay. Boston,

U.S., 1865.
Carus, Jules Victor. — C'oclenterata in ' Handbuch der Zoologie,' von Peters, Carus und Gerstaecker.

Leipzig, 1863.
Norman, A. Merle. — On undescribed British ^Hydroida, Actinozoa, and Polyzoa. ' Ann. Nat. Hist.,'

1864.
Haeckel, Ernst. — Beitriige zur Naturgeschichte der Hydro-medusen. Leipzig, 1865.

Generelle Morphologie der Organismen. Allgemeine Grundziige der organisch. Formen-Wissen-
schaft, Mechanisch begriindet durch die von Charles Darwin reformirte Descendenz-Theorie.
Berlin, 1866.
Ueber die fossilen Medusen der Jura-zeit. ' Zeit. fur Wiss. Zool.,' Bd. xix, 1869.
KiRCHENPAurAUER, G. II. — Die Seetounen der Elbmiindung : ein Beitrag zur Thier- und Pflanzen-
Topographie. Abhandl. Naturwissensch. Verein in Plamburg. Hamburg, 1862.
Nen Sertulariden aus verschiedenen Ilamburgischen Samnilungen, nebst Allgemeinen Bemerkungen

iiber Lamouroux's Gattung Dynamena. ' Acad. Ctcs. Leop. Nova Acta,' xxxi, 1864.
Ueber die Hydroidenfamilie, Plumularidfe, einzelne Gruppen derselben, und ihre Fruchtbehaltcr.
' Abhand. Naturwissenschaftl. Verein in Hamburg,' V Band, 1872.
Clark, H. J.— Tubularia not Parthenogenous. 'American Journal of Arts and Sciences,' 1864.

Mind in Nature, or the Origin of Life and the Mode of Development in Animals. New York,

1865.

MoBlus, Carl. — Ueber den Bau den Mechanismus und die Entwicklung der Nessclcapseln einiger

Polypen und Quallen. Abhandl. aus dem Gebiet. der Naturwissens. hrsg. von dem Naturwissens.

Verein in Hamburg, 1866.

BoECK, A. — Om to tylsynelactende bilateral-symmetriske Ilydromeduser, Dipleurosoma typica og

Stuvitzii. Vidensk. Meddel. fra den Naturh. Forennlng, 1866.
Reichert, Carl B. — Ueber die contractile substanz (sarcode, protoplasma) und ihre bewegungs-Erschei-

nungen bei Polythalamien und einigen anderen niederen Thieren. Berlin, 1867.
Semper, Carl. — Reisen ini Archipel der Philippinen. Leipzig, 1867 — 68.
Heixer, Cam. — Die Zoophyten und Echinodermen des Adriatischen Meeres. Wien., 1868.
MECZiSiKOFF, Prof. — Ueber die Entwicklung einiger Ccelenterata. ' Bull, de FAcad. de St. Pelers-

bourg,' XV, 1870.
ScHULZE, F. E. — Ueber den Bau und die Enwicklung von Condylophora lacustris,^l//w. Leipzig, 1871.
Kleinenberg, N. — Hydra, eine anatomisch-entwicklungsgeschichtliche Untersuohung. Leipzig,
1872.

INDEX

FAMILIES GENERA AND SPECIES OF GYMNOBLASTIC HYDROIDS DESCRIBED

IN THIS MONOGRAPH.

The Names printed in Italica are Synonyms. Those between quotation commas are the names
of provisional or doubtful species.

PAGE

PAGE

ACAULIS

. 377

ClONISTES .

. 309

Acaulis primarius .

. 378

Cionistes reticulata

. 309

ACIIARADBIA

. 375

CLADOCORYNID^ .

. 379

Acharadria larynx

. 376

Cladocoryne

. 379

ACTINOGONIUM

. 272

Cladocoryne tioccosa

. 380

Actinogoiiium pusillum

. 273

CLADONEMID/E

. 356

Amalthea .

. 392

Cladonema

. 356

Amalthca Januarii

. 394

Cladonema radiatum

. 357

Sarsii

. 393

CLAVIDiE .

. 242

uvifera .

. 393

Clava

. 242

Arum Cocksii

. 382

Clava cornea

. 243

Atracttjlis arenosa

. 300

discreta

. 246

" bitentaculata"

. 329

diffusa

. 247

linearis

. 328

leptostyla .

. 248

margarica

. 301

membranaeea

. 243

palttala

. 325

multicorais

. 246

" quadritentacula

a" . 329

nodosa

. 249

rainosa

. 311

repent

. 246

repens .

. 323

squamata .

. 243

sessilis

. 325

CLAVATELLIDiE .

. 383

BIMERID^

. 294

Clavatella

. 383

BlMERIA

. 297

Clavatella prolifera

. 384

Bimeria vestita

. 297

Clavultt Gossii

. 259

BOUGAINVILLlDJi: .

. 310

CORDYLOPHORA

. 251

BOUGAINVJLLIA

. 310

CordylopLora lacustris

. 252

Bougainvillia Carolinensia

. 316

albicola

. 254

fruticosa

. 314

CORTDENDRIIM

. 262

" Mertensii "

. 318

Corydendrium parasiticum

. 262

musciis

. 317

Corymbogoninm capillare

. 335

ramosa

. 311

CORYMORPHIDJi .

. 386

superciliaris

. 315

COBYMOHPHA

. 387

Campaniclava

. 260

Corymorptia annnlicornis

. 375

Campanielava cleodorse

. 261

glacialis

. 396

Candelabrum phrygium

. 382

Januarii

. 394

56

436

INDEX TO GENERA AND SPECIES.

PAGE

PAGE

Corymorpha nana

. 391

Eudendrium insigne

. 337

nutans

. 38S

" pusillum "

. 319

pendula

. 397

racemosum

. 341

Sarsii

. 393

rameum

. 334

uvifera

. 393

ramosum

. 332

CORYNIDiE

. 264

ramosum

, 311

CORYNE

. 264

" cingulatum "

. 342

" Coryne amphora "

. 272

teuue

. 340

Briareus .

. 290

vaginatum

. 339

caespes

. 270

Garveia

. 294

" filifera "

. 272

Garveia nutans

. 295

fritiUaria .

. 320

Globiceps tiarella

. 369

fruticosa .

. 269

Gemmaria .

. 289

ghindulosa

. 266

Geramaria implexa

. 290

implexa

. 290

Gymnocoryne

. 288

Listerii

. 282

Gymnocoryne coronata

. 288

mirabilis .

. 278

HA-LATRACTUS

. 390

" nutans "

. 271

Halatractus nanus

. 391

gravata

. 277

Halocharis spiralis

. 287

" prolifica "

. 272

Halocordyle

. 368

pusilla

. 266

Halocordyle tiarella

. 369

ramosa

. 269

Herittia glandulosa

. 266

"ramosa "

. 272

Heteractis annidicornis

. 375

" rosaria"

. 285

Heterocordyle

. 307

" sessilis "

. 271

Heterocordyle Conybearei

. 307

stauridium

. 357

Heterostephanus

. 374

vaginata .

. 268

Heterostepbanus annulicorui

s . 375

Fan Benedenii

. 273

Hqtpocrene SKpci ciliaris

. 315

vermicularis

. 267

HYBOCODONID^ .

. 421

COEYNITIS .

. 286

Hybocodon

. 421

Corynitis Agassizii

. 287

Hybocodon prolifer

. 422

COEYNOPSIS

. 354

Hydra aculeata .

. 353

Corynopsis Alderii

. 355

HYDRACTINIDiE .

. 342

Corythamnion bacciferum

. 293

Hydractinia

. 343

DICORYNID^

. 292

Hydractinia areolata

. 343

DiCOEYNE .

. 292

echinata

. 345

Dicoryne conferta

. 293

polyclina

. 347

stricta .

. 293

Hydeanthea

. 301

DiPLURA

. 319

Hydranthea margarica

. 301

Diplura FritiUaria

. 320

HYDROLARIDiE

. 424

ECTOPLEURA

. 423

Lar

. 425

Ectopleura Dumortierii

. 424

Lar sabellarum

. 425

Evcoryne elegans

. 369

Larida

. 424

Echinocormm clavigerum

. 345

Manicella fusca .

. 297

EUDENDRIDiE

. 330

Margelis Caroliiiensis

. 316

EUDENDRTUM

. 330

ramosa .

. 311

Eudendrium annulatum

. 338

Meeona

. 257

arbuscula

. 336

Meroua coraucopite

. 258

bacciferum

. 295

MONOCAULIDiE

. 395

capillare

. 335

Monocaulus

. 395

dispar

. 338

Monocaulus glacialis

. 396

humile

. 337

pendulus

. 397

INDEX OF GENERA AND SPECIES.

437

PAGE

PACK

MYRIOTHELID.E .

. 381

Syncoryne eximia

. 282

Myriothkla

. 381

ferox .

. 283

Mijriothela arctica

. 382

frutescens

. 281

phrygia

. 382

gravata

. 277

NEMOPSID^ .

. 360

" Johnston! "

. 285

Nemopsis .

. 360

" Listerii "

. 284

Nemopsis Gibbesii

. 362

Loveni

. 276

PENNARID^

. 363

" Lovenii "

. 285

Pennaria .

. 363

niirabiiis

. 278

Pennnria Cavolinii

. 364

pulchella

. 279

distijcha

. 364

racemosa

. 276

gibbosa

. 366

reticulata

. 2S3

tiarella

. 369

Sarsii .

. 275

Perigonimus .

. 321

Syndictyon reticxdatum

. 283

Perigonimus linearis

. 328

Syndydra parasites

. 345

minutus

. 324

TUBICLAVA .

. 255

muscoides

. 322

Tubiclava cormicopice

. 258

muscus

. 317

fruticosa

. 257

repens

. 323

lucerna

. 256

palliatus

. 325

TUBULARID^

. 398

serpens

. 327

TUBULARIA

. 398

sessilis

. 325

" Tubularia aspera "

. 420

vestitus

. 326

attenuata

. 410

PODOCORYNID^ .

. 348

Bellis .

. 409

PODOCOKYNE

. 348

calamaris

. 412

Podocoryiie aculeata

. 352

coronata

. 406

Alderii

. 355

Couthouii

. 403

areolata

. 353

cristata

. 417

carnea

. 349

crocea .

. 416

Crimea

. 303

Dumortieri

. 424

fuckola

. 304

gracilis

. 406

proboscidea

. 351

bumilis

. 411

tuhvUiria

. 349

indivisa

. 400

RhizocUiie areolata

. 343

insignis

. 405

Rhizogeton

. 250

larynx

. 406

Rhizogeton fiisiformis

. 250

mesembryanthemu

111 . 418

Sertularia parasitica

. 262

pacifica

. 416

pennaria

. 364

polycarpa

. 413

Spadix purpurea .

. 382

' regalis

. 404

Stauridium

. 370

simplex

. 411

Stauridium productum

. 371

spectabilis

. 414

Stipula ramosa

. 269

tenella

. 415

Stylactis .

. 302

TURRIDiE .

. 259

Stylactis fucicola .

. 304

TURRIS

. 259

inermis .

. 305

Turris neglecta .

. 259

Sarsii

. 303

Wrigiitea .

. 298

SYNCORYNID^

. 274

Wrigbtea arenosa

. 300

Syncoryne .

. 274

Vorticlava

. 372

"Syncoryne cbaniissonis "

. 272

Vorticlava bumilis

. 372

cleoilorce

. 261

proteus

. 373

decipiens

. 280

Zanclea implexa .

. 290

LIST OF WOODCUTS.

PASB

I. — Generalised diagram of a Gymnoblastic Hydroid to illustrate terminology ... ... xiii

II. — „ „ Calyptoblastic Hydroid to illustrate terminology ... ... ib.

Fig.

1. Diagraniatic section of a hydroid ... ... ... ... ... 27

2. Hydrosoma of Campamdaria Johnstoni ... ... ... ... ... 23

3. Portion of the hydrosoma of Calycella syringa, showing the operculate hydrotheca? ... 2.5

4. Group of zooids from a colony of -H(/f/r«c<i';ii'a ecAi«a^a ... ... ... 30

5. Ambulatory medusa of Clavatella prolifera ... ... ... ... 31

6. Free locomotive sporosac of i)/cor(/Me ... ... ... ... ... ih.

7. Adelocodonic gonophore of //yf/;«c//«j'« ec/i/wa^a ... ... ... ... 32

8. Oceania coronaia ... ... ... ... ... ... ... 33

9. Medusa, oi Campa)mhi)ia Jo/iiisloni ... ... ... ... ... 35

10. Sexual zooid (sporo.sac) budding from a radiating canal in the blastocheme of Obelia

geniculata ... ... ... ... ... ... ... ib.

11. Medusa (r/m« iJrt»Y/(V), a blastocheme of unknown trophosome ... ... ... 37

12. Portion of the ribbon-shaped sporosac, much enlarged, in a female specimen of Tima Bairdii ib.

13. Diagrammatic sections of phanerocodonic and adelocodonic gonophores ... ... 39

14. Longitudinal section through the hydranth and hydrotheca oi Laomedea Jlexnosa ... 41

15. Types of gonophores ... ... ... ... ... ... 44

16. Male gonophore of ii;/f/c?«fZ(7'MW... ... ... ... ... ... 46

17. Sarsia slranyulata ... ... ... ... ... ... ... ib.

18. Gonangium, with its contents, from ZaoOTefZe«^ca;?<o«a!... ... ... ... 48

19. „ „ Obelia geniculata ... ... ... ... ib.

20. „ „ Laomedea repens ... ... ... ... 49

21. Female gonangium, with acrocyst of &»Yw/an'o j5?^??i«7o ... ... ... ... 50

22. ,, „ Calycella lacerata ... ... ... ... ib.

23. Mature female gonangium with marsupial chamber in (Sfer^w/arm rosare« ... ... 51

24. Young female gonangium of iSfrtufo/'w rosacea ... ... ... ... ib.

25. Very young female gonangium of (Sfe/^i^/arj'a rosacea ... ... ... ... 52

26. Male and female gonangia of iSfer^e^Zarea <aOTan«ca ... ... ... ... 53

27. A portion of the hydrosoma of Co/)/)?';w'a arc^a ... ... ... ... 54

28. Gonangium with meconidia of Go«o</(yr(»a ioi'e??j (female) ... ... ... 57

29. „ gonangial hydranths in i/a/ectMTO /ia/ec/wMOT ... ... ... 58

30. Development of the corbula in vi(//aojo^e??/aj3/?<»ia ... ... ... ... 60

31. ,, spermatozoa in Laomedea flexuosa ... ... ... ... 65

32. Shedding and renewal of the hydranth in r^Jw/ar/a j«(Z/c!'4a ... ... ... 69

33. Development of hydranth and hydrotheca in iao»8e(/ea^ca:Mosa ... ... ... 72

34. „ the planoblast in Carymorpha nutans ... ... ... ... 77

35. Segment of a yonng .^quorea ... ... ... ... ... ... 79

36. Medusa, probably young Z)'?;/a ... ... ... ... ... 82

37. >&r«/rt, with medusa-buds borne by the manubrium ... ... ... ... 83

38. Medusa, bearing clusters of medusa-buds on bases of marginal tentacles ... ... ib.

440 LIST OF WOODCUTS.

Fig. pagb

39. Development of the ovum in icro»!f'f/c'fl'^C'a:?/o«a ... ... ... ... 86

40. Diagram of Corymorpha ... ... ... ... ... ... 103

41. „ Dicoryne ... ... ... ... ... ... ib.

42. „ Campamilaria ... ... ... ... ... ... ib.

43. „ Laomedca ... ... ... ... ... ... 104

44. „ Tuhularia indivisa ... ... ... ... ... ... 109

45. J, „ larynx ({enmle) ... ... ... ... ... ib.

46. Blastostyle of hydractinia with gonophores ... ... ... ... ... ib.

47. Hydranth of Clata, surrounded by globular clusters of gonophores ... ... ... ib.

48. Viitt of s. tentncle of Syncoryne jmlchella ... ... ... ... ... 112

49. Structure of medusa of »S'y?zcori/MejB2(/c/ie//a ... ... ... ... 113

50. Portion of a. Tumulas of Anten?iularia ante)inma ... ... ... ... 116

51. Hydrothecfe of ^5'/rto^;A£'«/a /;/(/»2(7, with hydranths and nematophores ... ... 117

52. Diagramatic views of the thread-cell ... ... ... ... ... 118

53. Development of thread-cell in i^yJra ... ... ... ... ... 122

54. Canalliculation of endoderm in vl«fe?mzJana a??<ewiraa ... ... ... 126

55. Growing extremity of adventitious branch of OieZ«a f/icAotomn! ... ... ... 133

56. Part of umbrella-margin in (Si/Heoj'y^je eaja'mio! ... ... ... ... 139

57. „ „ in Tyarojjsis scotica ... ... ... ... 140

58. „ „ in medusa of Campanularia Johnstoni ... ... ... 141

59. Meinsa. of Ohelia ffeniculafa ... ... ... ... ... ... 142

60. Young sporosac of Coryne p>imlla ... ... ... ... ... 149

61. ScliKocladium ramosum, showing reproduction by spontaneous fission ... ... 152

62. Diagramatic longitudinal section of ^ciiwj'a ... ... ... ... 192

63. ,, transverse section of *4ciwia ... ... ... ... ... ib.

64. „ longitudinal section of fi^f/ra ... ... ... ... ... 193

65. „ transverse section of iy(/(7ra ... ... ... ... ... ib.

66. „ longitudinal section of a Hydroid Medusa ... ... ... ... 194

67. „ transverse section of a Hydroid Medusa ... ... ... ... ib.

68. Diagram of a siphonophore ... ... ... ... ... ... 195

69. Diagramatic longitudinal section of a discophorous medusa ... ... ... 196

70. „ transveise section of a discophorous medusa ... ... ... ib.

71. „ longitudinal section of ie<c«r««;'i"a ... ... ... ... 197

72. „ transverse section of i?<cp?7?«r/a ... ... ... ... ib.

73. „ longitudinal section of Utroe ... ... ... ... ... 198

74. „ transverse section of Beroe ... ... ... ... . . . ib.

75. A portion of a colony of Syncoryne eximia infested by Phoxichilidium coccineum ... 201

76. Conjoined twin meduscc develoved from a specimen of >S'y?«corywe/)MZc/«e//a ... ... 202

77. A portion of a colony of CurdyJuphora lacusti'is, showing the transformation of a spadix into

a hydranth ... ... ... .. ... ... ... 204

78. Synthecium elegans, showing the origin of the goiiangium within the hydrotheca ... 229

79. Stylactis inermis ... ... ... ... ... ... ... 306

80. Fragment of a colony of Pe««(7?'/o C'«to/i«i« ... ... ... ... 365

81. Acharadn'a larynx ... ... ... ... ... ... ... 377

82. Cladoconjnejloccosa... ... ... ... ... ... ... 380

83. Tubulai'ia mesemi/ryunihemum, ■nnturnl she ... ... ... ... ... 420

84. ,, ,, magnified details ... ... ... ... ib.

GENERAL INDEX.

Abyssal zone, its hydroid fauna, 167.

Acalephse, North American, illustrated catalogue of, by Alex. Agassiz, 19.

Acalepha, instituted by Cuvier as a class, 10.

Acaulis, its supposed free trophosome probably only a detached hydranlh, 68, 378.

Acaulis primarius, two very distinct forms referred by Stimpson to this species; reasons for rejecting

this view, 378.
Acraspeda and Craspedota, Gegenbaur's division of medusa; into, 16.
Acrocyst, 48.

Actinia compared with Hydra, 194 ; compared with a hydroid medusa, 195.
Actinogonium, its embryonal development, 95.
Actinula, development through, 90.

Adelocodonic gonophores, 30 ; gastrovascular canals in, 45 ; their condition in Eudendrium, 45.
yEgenidffl, their metamorphosis, 80.

^ginidan medusa^ produced by budding from a geryonidan, Haeckel's account of, 84.
^gineta cjenimifera, buds occurring in the cavity of its stomach, 84.
Africa, South, large proportion of its hydroids identical with British species, 158.
Agamic reproduction, 150.

Agassiz, Alex., his illustrated catalogue of North American Acalephse, 19; observes the development of
the egg in Timaformosa and in Mclicerhim campanula, 97 ; his researches on the development
of the marginal tentacles in the hydroid meduste, 80, /lote.
Agassiz, Prof. L., his contributions to the natural history of the United States, 17 ; he maintains the
hydroid nature of the rugose and tabulate corals, 3, note, and 18 ; his comparison of the
interseptal chambers of Actinia with the radiating canals of a medusa, 192 ; his dismember-
ment of the genus Tubularia, 400.
Aglaophenia, a genus specially characteristic of intertropical and warmer temperate seas, 156.
Alcyonaria, diagnostic characters of the order, 188.
AUiogenesis of Haeckel considered, 106.
Alternation of generations, discovered by Chamisso, 10; its law enunciated by Steenstrup, 13 and 101 ;

accurately interpreted by Carpenter, 101.
Amalthaa Januarii, its hydranth the largest and its stem the thickest yet discovered among the Hydroida,

394.
Ancestral type of the Hydroida, 230.
Antennvlaria antennina, canalliculation of endoderm in, 125 ; peculiar free bodies occurring in its

gonangium, 59.
Archydra, 230.

Ascidians, compound, Savigni's researches on, 9.
Atractjlis, validity of this name considered, 299.
Atrium, 33.
Australian province, 162.

442 GENERAL INDEX.

Bathymetrical distribution of the Hydroida, 162; its relation to marine vegetation, 164.

Bathymetrical zones of Forbes, necessity of their modification, 163; bathymetrical zones of the present
monograph, 166; variations in the physical condition of the various parts of the same bathyme-
trical zone, 164; great extent of the bathymetrical range of hydroid life, 164.

Beneden, van. Prof., his memoirs on the canipanularian and tubularian hydroids, 14; his institution
of the genus Hydractinia, 14; his litoral fauna of the coast of Belgium, 19; his rectification
of Ehrenberg's genus Eudendrium, .331 ; his observations on the medusa of Cladoiwma
radiatum, 3-59.

Bifurcation of zooids in Hydractinia, 204.

Bimcria vcstita, great extent of its chitinous excretion, 298.

Blainville, M. de, his ' Manuel d'Actinologie,' 12.

Blastocheme, 29; how distinguished from gonocheme, 35; various forms of its generative bud, 36.

Blastostyle, 29 ; its form, 33; true and false blastostyles, 34; its buds, 47; compound blastostyle, 47 ;
development of blastostyle, 73.

Boreo-Celtic province, 162.

Bougainvillia-medusa traced by Wright and Agassiz to a Eudendrium-like trophosome, 311.

BougaimiUia Britannica of Forbes probably the adult planoblast of the JEudendrtum ramosum of van
Beneden, 313.

Bougaiuvilia, species provisionally referred to this genus, 318.

Brandt, his account of the medusoe observed by Mertens, 12.

Buds formed by medusa;, 82.

Busk, his observation of transverse strife in the fibrillated tissue of the umbrella, 114 ; his ob-
servations on the structure of the lithocyst, 141, note; his observations on the lithocyst under
polarized light, 143, note ; on the phosphorescence of medusic, 145 ; his determination of Australian
hydroids, 156.

Calyptoblastea, a sub-order of hydroids ; characters of this sub-order, 189.

Campanularia, supposed nervous cord in, 138.

Canalliculation of endoderm in Tubularia, 124 ; in Conjmorpha nutans, 125 ; in Antetmularia

antennina, 125.
Carmarina hastata, Haeckel's account of its metamorphosis, 81.
Carpenter, Dr. W. B., his views of the zoological individual, 22, note ; gives the true interpretation of

the law of alternation of generations, 101.
Carus, J. v., his treatise on the Coelenterata, 18.
Cartilage-tissue described by Haeckel in the Geryonidse, 115.
Cavolini, his researches on the Hydroida of Naples; he describes the production of a medusiform

gonophore ; he maintains the vegetality of the Nullipores, 8 ; he was the first to observe somatic

currents in the Hydroida, 130, note: notices the difference between the male and female gonophores

in Eudendrium racemosum, 341.
Ceratostera of Leuckart, 16.

Chamisso, his discovery of alternation of generations, 10.
Chitine, perisarc of the Hydroida composed of, 26.
Chlorophylle, green matter of Hydra viridis compared with, 135.
Cilia, vibratile, their presence on the walls of the somatic cavity, 123 ; their action in causing the

somatic currents, 132.
Circulation in the Hydroida, 130.

Cladocoryne, the only hydroid known with the tentacles of its hydranth branched, 379.
Cladoncma radiatum, development of eggs in, 96; its anatomy, 216; structure of its gonophore, 217;

Dujardin's observations on, 357.
Claparede, his views of the development of Tubularia, 93, note ; his account of the direct development

of Lizzia from the egg, 100.
Clark, Prof. J., his views of the development of Tubularia, 92, note.
Classification of the Hydroida, 187 ; tabular view of hydroid classification, 191.

GENERAL INDEX. 443

Clava, restoration of Gmelin's genus by Johnston, 243.

Clava sqiiamosa, its peculiar liydrorhiza, 245; development of ovum in, 245.

Clavatdla jirolifcra, ambulatory gonophore of, 30; anatomy of, 212; structure of its jjlanoblasts, their

development and gemmation, 213 ; its planoblasts compared with an ordinary hydroid medusa,

215; distinct from Eleutheria, 384; its discovery by Hincks, 385.
Climacograptus, a genus of Gra])tolites, 176.
Codonostome, 33.
Coeleuterata, establishment of this group by Leuckart, 15 ; manual of, by Greene, 18 ; homologies of,

192; Kowalewsky's and Semper's views of the body-cavity of, 193, note.
Ccenosarc, 26.
Cohn, his comparison of the green matter of Hydra with that of Infusoria and with the chlorophyllc of

plants, 135.
Coloured matter in the Hydroida, a secretion, 134.

Contractility in the Hydroida dependent on fibrillated tissue and on simple sarcode, 136.
Coppinia arcta, its structure, 55.
Corals, animality of, demonstrated by Peysonelle, 5 ; tabulate and rugose corals, the hydroid nature of

these maintained by Agassiz, 3, note, and 17.
Coralline zone, its hydroid fauna and associated \egetation, 168.
Corbula, its structure and development, 60 ; consists of a special and excessive development of the

nematophoral system, 181, note.
Cordyhphora lacustris, conversion of its spadix into a hydranth, 204 ; account of its structure by Schultze,

228 ; its fresh-water habitats ; probably an introduced species ; a light-shunning animal, 253.
Corydendrhan paras/tiotm, egg-like bodies described by Cavolini in its hydrocaulus, 263.
Corymorpha, development of the planoblast in, 77 ; canalliculation of endoderm in, 125; probable spon-
taneous fission in, 153 ; its anatomy, 208; structure of its stem, 209 ; its hydrorhizal filaments,

209; structure of its hydranth, 209; its planoblasts, 210 ; its free frustules, 211 ; its medusa

generically identical with the Steenstrupia of Forbes, 390.
Coryne, revision of the genus by Sars, 242; provisional and indeterminable species of, 271.
Corynoides, description of, by Nicholson ; no evidence of its hydroid affinity, 172.
Corynopsis, its planoblasts identical in form with those of Bougainvillia, 354.
Cosmopolitan species of Hydroida, 160.
Craspedota of Gegenbaur, 16.
Ctenophora, diagnostic characters of, 188 ; compared with hydroid medusa, 197 ; an order of Hydrozoa,

and not of Actinozoa, 199.
Cunina octonaria, M'Crady's account of its metamorphosis, 81.

Cunina KolUkeri, Fritz Muller's account of budding from the internal surface of the stomach, 83.
Ctmina prolifera, Gegenbaur's account of buds in cavity of stomach, 84.
Cunina, its direct development from the egg, 100.
Cuvier, institutes his primary group lladiata ; defects of this group, 9 ; his publication of the ' Regne

Animal' and establishment of the class Acalepha, 10; recognises the difference of structure

between the Actinozoal and the Hydrozoal types, 10.
Cystoseeia, zone of, 304, note.

Dalyell, Sir J. Graham, his observations on the development of medusfc from hydroid trophosomes, 15;

he describes the decapitation and re-formation of hydranths in Tubularia, 69.
Decapitation, spontaneous, and re-formation of hydranths in Tubularia, 69; the re-forniation of the

hydranths a process of metamorphosis, 70.
Deep-sea zone, its hydroid fauna, 169.
Dendritic form, a result of gemmation, 67.
Development, a department of morphology, 66 ; development of the bud and of the cuibryo compared,

66 ; development of the hydranth in the Gymnoblastea, 68 ; in the Eleutheroblastea, 60 ; in the

Calyptoblastea, 71 ; embryonal development, 85.
Dicorijne conferta, locomotive sporosac of, 31 and 226; anatomy of Dicoryne conferia, 226; its piano-

444 GENERAL INDEX.

blast compared with an ordinary liydroid medusa, 277 ; its tropliosome indistinguishable from
that of Heterocordyle, 294.

Dictyonema not a graptolite, 184, note.

Digestion in the Hydroida, 128.

Dinema of van Beneden not distinct from Perigonimus, 322.

Dioecious condition of liydroid colony almost universal, 148.

Diplograptus, a genus of graptolites, 176.

Discophora, determined by Eschscholfz as a distinct group of the niedusoc, 10 ; fossil impressions of
Discophora, 176 ; diagnostic characters of the order, 188.

Distal end of hydrozoma, 24.

Distribution of the Hydroida in space; horizontal distribution, 155; bathymetrical distribution, 162;
wide distribution of genera, 156; limited distribution of species, 157 ; wide distribution not dependent
on production of planoblasts ; distribution of specific forms of hydroid medusae limited,159; distri-
bution of the Hydroida in time, 170 ; tabular view of the distribution of the Hydroida in time, 184.

Dujardin, his observations on the production of medusa" by hydroid trophosomes, 14 ; his account of
the development of the eggs in Cladonema, 96 ; his observations on the Cladonema medusa and
its development from its trophosome, 96.

Echer, his views of the structure of Hydra, 123.

Ectoderm, 21 ; its structure, 110.

Ectotlicca, 32.

Edwards, Prof. H. Milne, insists on the distinction between the Hydrozoal and Actinozoal types of

structure, 11 ; he compares the Flustrse with the compound Ascidians, 11.
Ehrenberg, his use of the name Bryozoa as equivalent with the Polyzoa of Thomson, his demonstration

of the sexuality of the Hydroida, 12; his institution of the genus Eudendrium, 331.
Eleutheria, ambulatory gonophore of, 30; distinct from Clavatella, 384.
Eleutheroblastea, a sub-order of the Hydroida; characters of this sub-order, 189.
Ellis, his researches, 6.

Ellis and Solandcr, their admirable iconogra])hy, 7.
Embryonal development, 85 ; embryonal development in the Hydroida compared with that of the

Vertebrata, 88.
Endoderm,21; its structure, 122; ciliated surface of, 123 and 228; its ridges, 126 ; its condition in the

hydranth of Tuhnlaria indicisa, 124; its structure in the tentacles of the hydranth, 126; in the

tentacles of the planoblast, 127.
Endostomata of M'Crady, 80.
Endotheca, 32.

Esper, his ' Pflanzenthiere,' 7.
Eschscholtz, his recognition of the hydroid medusse as forming a group distinct from that of the

Discophora, 10.
Eudendrium, peculiar condition of its male sporosacs, 45; its male sporosacs compared with Sarsia

strangidata, 45; atrophy of the hydranth in various species, 236.
Euclendrinm ramosum liable to be infested by a parasitic Pycnogonidan, 201 ; incorrectly described by

authors as having a non-fascicled stem, 333.
Eadendrhtin 2n<sillum of Sars probably a Bougainvillia, 319.
Evolution of the Hydroida, 230.
Exostomata of M'Crady, 80.

Families of the Hydroida, characters employed in tlicir limitation and diagnosis, 237.

Fascicled stems, their nature, 262, note.

Fibrillated tissue, 112 ; in the umbrella and velum, 114; structure of fibrilla? in the tentacles of Tubu-

laria, 113.
Fission, spontaneous, in the medusa; Kolliker's observations on ; in the trophosome, 151 ; in 8chizo-

cladiiim, 152; probable in Corhnorpha nutans, 153.
Forbes, Edward, his division of medustc into Gymnophthalmia and Steganophthalmia ; his monograph

GENERAL INDEX. 445

of the British iiakcd-cyed iiiediisa>, 14; his views of the morphology of corbuloc and goiiangia,

61, note ; his batliymetrical zones, 163.
Forel, liis researches in the deep fauna of the Lake of Geneva, 170.
Formula; of the genetic succession of zooids, 101 — 105.
Forskal figures a Clava, and gives the first good figures of liydroid medusse, 7.

Gartner institutes the genus Coryne, 264.

Ganeia nutans, its sporosac provided with rudimentary radiating canals, 296.
Gastrovascular canals, 33.

Gegenbaur, his memoir on the medusrc ; his division of the medusw into Acraspeda and Craspedota,
16 ; his memoir on the alternation of generations in the Hydroida ; his outlines of comparative
anatomy, 17; his case of buds occurring in the cavity of the stomach in Cimina jtrolifcra, 84;
> his observations on the developTnent of eggs in Lizzia KoUUceri and in Oceania armata, 97.

Geryonidre, their metamorphoses, 80 ; cartilaginous tissue in, 115, note ; lithocyst in, 142, note.
Glossocodon curyhia, Ilaeckel's account of its metamorphosis, 81.

Gmclin, his edition of the ' Systema Natura;,' state of hydroid zoology at the time of its publication, 8.
Gonangium, 34 ; its nature; always associated with a hydrotheca, 47 ; development of, 74; aggregated
gonangia in Coppinia, 54; gonangium springing from within the hydrotheca in Syntliecium, \li'.K
^^Gonangial hydranths of Halecium lialecinuin, 58.
Gemmaria imjilexa, its anatomy, 223 ; its remarkable planoblasts, 291.

Gemmation, its various sites, 150 ; occasionally coexistent with sexual reproduction in the medusa, 150.
Genera of liydroids, characters employed in their limitation and diagnosis, 237.
Generation, 147.

Generative elements, their origin, 148 ; a product of the endoderm, 149.
! Genetic succession of zooids, fornuilse of, 101 — 105.
;Geryonia, its direct development from the egg, 100.
Gonocheme, 29, 32.

Gonophores, 29; gradations in their complexity, 43.
Gonosome, definition of, 23 ; general view of, 29.
Gonothyrsea, its meconidia, 55.

Gosse, his observation of the development of the egg in Turris, 97 ; his account of Lar sahellarum, 42().
Grant, Prof., his recognition of a polyzoal type in Flustra, 11.

GraptoliteS; their tyi)ical form ; monoprionidian and diprionidian forms ; their probable affinities with
the Hydroida, 176; referred by authors to the calyptoblastic hydroids ; their rod or solid axis,
177 ; comparison of their rod with that of Rhabdopleura, 178 ; their callicles compared with the
nematopliores of the Plumularida;, 179 ; their affinities, both hydroidal and rhizopodal, 180 ; their
supposed gonangia as described by Hall, 181 ; and by Nicholson, 182 ; early stages of their deve-
lopment; various views of their nature, 183.
Green, Prof. J. R., his ' Manual of the Ccclenterata,' 18 ; his discovery of the Diplura medusa, 3.'->0.
Growth of the tissues dependent on cell-formation, 132.
Gubernaculum, 47.

Gulf Stream, exploration of, by the United States Coast Survey, 165, note.

Gymnoblastca, a sub-order of liydroids ; characters of this sub-order, 189 ; genera of Gymnoblastea pro-
bably for the most part limited in their geographical distribution, 160; synopsis of the families
and genera of the Gymnoblastea, 239.
Gymnophthalmia and Steganophthalmia, division of medusje into, by Forbes, 14.

Haeckel, Ernst, his researches on the ^Eginidoe and Geryonidre, 19 and 80; his views as to
' alleogenesis' considered, 106; his account of 'medusa cartilage,' \l^, note ; his researches on
the nervous system of the Geryonidte, 138; his observations on the lithocyst in the Geryonid<c,
142, note ; his view of the aboriginal type of the Hydroida, 230.
Halecium halecinum, its gonangial hydranths, 58.
Hensen, his account of the lithocvst, 141, note.

Hincks, his 'History of British Hydroid Zoophvtcs,' xi, Prcf.; he describes Ophiodes and its peculiar
appendages, 28; his sub-orders of the Hydroida, 190, note; he demonstrates the identity of form

446 GENERAL INDEX.

in the planoblasts of Stauridmm produclum and Sijncoryne eximia, 370 ; his discovery of CIavatella>
385 ; his discovery of the gonosome of Lar sabellarum, 425.

Histology of the Hydroida, 110.

Honiolognes of the Ctt'lenterata, 192.

Homologous parts in the Hydroida, scheme of, 43.

Huxley, Prof, his group of the Nematophora equivalent with that of the Cuilenterata of Leuckart, 15;
his division of the Coelenterata into Hydrozoa and Actinozoa, 16, 187 ; his monograph of the
Siphonophora ; his systematic arrangement of the Hydrozoa, 17 ; his determination of the two con-
stituent membranes of the hydroid body, 22 ; his employment of the term ' zooid ;' his definition of
the Biological Individual, 22, 7tote, and 107.

Hydra and its power of budding discovered by Leeuwenhoek, 4 ; investigated by Trembley, 4 ; development
of ovum in, 94; spermatogenous tubercles in; ovigerous tubercles in, 93 ; its wide geographical
distribution, 160; Hydra compared with Actinia, 194; compared with Lucernaria, 197.

Hydra viridis, its green corpuscles, 123.

Hydranths, 24 ; the two largest as yet known belong to species one of which is tropical and the other
arctic, 394.

Hydrallmaitia falcata, its great batliymetrical range, 165.

Hydractinia cretacea, fossil of the Upper Greensand, 173.

Hydractitiia Michelini, fossil, of the Miocene, 173.

Hydractinia pliocena, fossil of the Coralline Crag, 173.

Hydractinia echinata, liable to be infested by a parasitical pichnogunidan, 202 ; bifurcation of zooids in,
204; its anatomy, 220; structure of its hydrophyton, 220; its hydranths and spiral z(joids, 221 ;
its gonosome, 222; history of its discovery, 341.

Hydrocaulus, 27.

Hydrorhiza, 27.

Hydroid zoology, retrospect of the leading steps in its progress, 20.

Hydroid medusa compared with the discophorous medusa, 197 ; compared with the Ctenophora, 197.

Hydroid, generalised conception of a, 21.

Hydrophyton, 26.

Hydrosoma, 22.

Hydrotheca, 26 ; development of, 72.

Hydrozoa and Actinozoa, classes of the Coelenterata, 187.

Hypostome, 24.

Individuality, biological, 22, note, and 107.

Johnston, his 'History of British Zoophytes;' he proposes the name Hydroida, 13; his restoration of

Gmelin's genus Clava, 243.
Jussieu, Bernard de, he demonstrates the animality of Tubularia, 5.

Keferstein and Ehler, their account of buds occurring in the stomach-cavity of ^Egineta geinmifera, 84.
Kdlliker, his union of the gymnophthalmic medusoD and the hydroid trophosomes into a single group, 15
Kotzebue, his circumnavigatory voyages, 10.

Kowalewsky, his views as to the common cavity of the Cadenterata, 193, note.

Krohn, his account of the development of the egg in Cladoneraa, 96 ; his discovery of spermatozoa in
certain marine hydroids, 148.

Lacustrine fauna, the deep, 170.

Lamarck, his institution of new genera of hydroids, 9.
Laminarian zone, its hydroid fauna and associated vegetation, 168.

Laniouroux, his institution of new hydroid genera; the claims of his nomenclature to acceptance, 9.
Laomcdca Jlexuosa, development of ovum in, 85.

Lar sal/cUarum, its trophosome described by Gosse ; discovery of its gonosome by Hincks, 425.
Leeuwenhoek, Ant. von, he discovers Hydra, notices its budding, and sends an account of it to the
Royal Society, 4.

GENERAL INDEX. 147

Lens-like constituent of ocellus, 139.

Lesson, his ' Ilistoiie Naturelle des Acalephes,' 14.

Leuekart, his establishment of the group Coelenterata, 15; his union of gymnoiihtlialmic mcdusoe and

hydroid trophosoraes into a single group; his group Ceratostera, 16.
Lewes, G. H., his observations on the functions of the thread-cells, 229; he maintains the occurrence of

both adclocodonic and phanerocodonic gonophores in Afflaplicnia myrlophyllum, .304.
Linnajus, the researches of Ellis convinces him of the animality of the liydroids, 6 ; his binomial

nomenclature, 6; his 'Systema Natura:',' state of hydroid zoology at the period of the publication

of Gmelin's edition of it, 8.
Liriojic cathariensis, Fritz Miiller's account of its metamorphosis, 81.
Lithocyst, its structure and situation, 140; its function, 143.
Literal zone, its hydroid fauna and associated vegetation, 168.
Lizzia Kollikeri, development of the eggs in, 97.
Lizzia, alleged direct development from the egg, 100.

Loven, his discovery of the meconidium ; his recognition of differentiated sex in the Hydroida, 147.
Lucernaria, diagnostic characters of the order, 188; comparison of, with Hydra, 197.

M'Crady, his memoir on the Gymnophthalmata of Charleston Harbour, 17.

Man, his agency in determining the distribution of the Hydroida, 159.

Manubrium, 32.

Margelis scarcely distinguishable from Bougainvillia, ■i\\,Hutc.

Marsigli, the polypes of coral regarded by him as flowers, 5.

Marsupial receptacle in Dipliasia (Seriularia) rosacea, 50 ; in Diphasia (Sertularia) fallax ; in Diphasia
[Sertularia) tamarisca, 52.

Meconidia of the genus Gonothyr»a, 55.

MecznikofT, his observations on the direct development of Cunina and Geiyonia from the egg, 100.

Medusa, 32 ; development of, 76 ; homological parallelism between medusa and sporosac, 38 ; between
medusa and hydranth, 40; significance of the medusa in the life-series of the hydroid, 95; deve-
lopment of its eggs, 96 ; fossil hydroid medusae, 174 ; comparison of hydroid medusa with
Actinia, 195.

Mediterranean province, 162.

Melicertum campanula, development of egg in, observed by Alex. Agassiz, 97.

Mertens, his observations and figures of medusae, 12.

Mesotheca, 32.

^letamorphosis of the Geryonidai and jEginidae, 80.

Moebius, his observations on thread-cells, 120, note.

Monocaulus glacialis, remarkable for the high northern latitude of its geographical area, as well as for its
deep bathy metrical zone, 396.

Mona'cious colonies sometimes occur among the Hydroida, 148.

Monopsea, a sub-order of liydroids, characters of this sub-order, 189.

Morphology of the Hydroida, 21.

Miiller, Fritz, his account of budding from the internal surface of the stomach in Cniiina Kulliktri, 83.

Miiller, O. F., his ' Zoologia Danica,' 7.

Myriothela, its development through Actinuloe, 94, 382.

Nematophora, name of a sub-kingdom proposed by Huxley, 15.

Nematophores, term proposed by Busk, their form, 28 ; their sarcode contents and action, 115.

Nemopsis, the sexual lobes of its medusa, 63 and 361 ; its alleged free trophosome probably only a

detached hydranth, 68 and 361.
Nervous system in the Hydroida, 137; Haeckel's researches on the nervous system in the Geryonida?,

138.
New Zealand province, 162.

Nomenclature of the Hydroida, how regulated, 237.
North Pacific province, 102.

448 GENERAL INDEX.

Noj'thern Atlanto-Amevican province, 162.
NulHpores, their vegetality maintained by Cavolini, S.

Obelia, structure of umbrella in, 11.5; supposed nerve-cord in, 1.38.

Oceania armata, development of the eggs in, 97.

Ocellus, -33; its structure and situation, 138; lens-like body occasionally occurs in it, 139; its supposed

function, 139.
Oldhamia, not properly a hydroid, possibly inorganic, its form and geological area, 171.
Ophiodes, peculiar appendages in, 28.
Orientation of the hydroid, 24.
Ovum, hydroid, its structure, 64; development of, 8-5 ; special features in the development of ovum in

Tubularia, 90; in Hydra, 94; ova in the blastostyle of Scrtularia jnimila, 150.

Pallas, his ' Elenchus Zoophytorum ;' he adopts the binomial nomenclature of Linna-us, 6 ; he describes

a Coryne and a CUava, 7.
Pahroeoryne, description of, by MM. Duncan and .Jenkins ; no sufficient grounds for regarding it as a

hydroid, 172; probably a rhizopod, 173.
Palagina. gigantea, fossil hydroid medusa of the Solenhofen slate, 175.
Palpocils, 111.

Parypha, a sub-genus of Tubularia, 399 and 416.
Parasitic pichnogonidan, 200.

Pennaria exceptional among the Gymnoblastea in the distichous disposition of its branches, 305.
Pennaria Cavolinii, its medusa attains to sexual maturity without liberation, 366.
Pennarta gibhosa, the fusiform swellings of its radiating canals not to be confounded with generative

sacs, 367.
Perisarc, 26; a product of the ectoderm, 135.
Perigonium, 32.

Peron and Lesueur, their Australian voyage, 8.
Peyronelle demonstrates the animality of corals, 5.
Phanerocodonic gonophores, 30.
Phosphorescence in the trophosome ; in the planoblast ; manifested under the operation of a stimulus;

action of alcohol on; emission of light by marginal bulbs of medusoe, 145; phosphorescence in

Beroe requires previous seclusion in darkness, 146, note.
Phoxichilklium coccineum, a parasite o( Si/nconjnc ezhnia, 201.
Phyllograptus an anomalous graptolitic form, 177.
Physiognomy of the Hydroida, 1.
Physioiogy of the Hydroida, 128.
Pichnogonidffi parasitic in hydroids, 200.
Planoblasts, 29.

Planula, 88 ; development of, 89.
Plumularians, gigantic, 156.

Podocoryne carnea, sexual maturity of its medusa occasionally noticed before liberation, 350.
Podocorync prohoscidea, sexual maturity of its medusa before liberation, 351.
Polarity of the hydroid, 70.
Polype, use of this term, 24.
Polypi te, use of this term, 24.
Porcupine Expedition, hydroids obtained by it from great depths ; its discovery of the deep cold

North Atlantic area, 165 ; it discovery of abyssal forms of liydroids, 165, 170.
Protohydra, account of, by Greef, 230.
Protoplasm of nematophores, 115. (See Sarcode.)

Provinces of hydroid distribution, 161; Boreo-Celtic province; northern Atlanto-American province;
West Indian province ; North Pacific province ; Australian province ; New Zealand province, 162.
Proximal end of hydrosoma, 24.
Quatrefiges, !M. de, his memoir on Hydractinia (Synhydra), 343 ; his memoir on Eleutheria, 384.

GENERAL INDEX. 449

Kadiating canals, their ilevclopmrnt centiifugal, 80.

Ramification of the trophosome symmetrical and asymmetrical, 67.

Rapp assumes the difference of position in the reproductive organs as affording an essential distinction

between the actinozoal and hydrozoal types, 11.
Rastrites, a genus of graptolites, 176.
" Rattlesnake," voyage of, 17.

Reichert, his determination of a structureless membrane (stiitzlamelle) in the Hydroida, 228.
Respiration in the Hydroida, 134.
Retiolites, an anomalous graptolitic form, 177.

Rhabdophora, name of the extinct group represented by the graptolites, 190.
Rhabdopleura, comparison of, with graptolites, 184.

Rltizogeton fusiformis, supposed conversion of its gonophore into a hydranth, 204.
Richter, his views of the structure and affinities of graptolites, 185.
Rotch, W. D., his discovery of Cladocoryne, 380.

Sarcode-layer on free surface of ectoderm. 111. (See Protoplasm.)

Sars, his revision of the genus Coryne, 242, 265 ; his institution of the genus Perigonimus, 321.

Sarsia strangulata compared with the male sporosac of Eudendrium, 45.

Sarsia regarded by Agassiz and Clark as the ultimate form of the planoblast in Syncoryne, 279.

Savigni, his researches on the compound Ascidians, 9.

Secretion in the Hydroida, 134.

Semper, his views as to the body-cavity of the Coelenterata, 193, 7iotc.

Sensation in the Hydroida, 137.

Sertularia pumila, ccecal processes from its blastostyle, 54; a cosmopolitan species, 161.

Sertularia polijzonias, a cosmopolitan species, 161; its great bathymetrical range, 165; cited as fossil

from the Pleistocene, 173.
Sertularia operculata, a cosmopolitan species, 161.

Sertularina, data for the geographical distribution of the Hydroida hitherto chiefly afforded by them, 159.
Sexes in the Hydroida, comparison of, 147 ; announcement by Ehrenberg of differentiated sex in the

Hydroida, 12, 147 ; recognised by Loven, 147.
Sexual differences among hydroid colonies, 62.
Siphonophora, Huxley's monograph on, 17; diagnostic characters of, 188; comparison of, with

Hydroida, 195.
Solander and Ellis, their admirable iconography, 7.
Somatic fluid in the Hydroida ; currents in it, 130.
Spadix, 32 ; occasionally branched, 45.

Spermatozoa in the Hydroida, their form, 64 ; their development, 65.
Spiral zooids of Hydractinia, 221.
Sporosac or adelocodouic gonophore, its parts, 32 ; locomotive sporosac in Dicoryne, 31 ; development

of the sporosac, 74.
Statocodium, a provisional genus of Syncorynidae, 279, note.
Stauridium, nature of its filiform tentacles, 371.
Staler idmm productum, its planoblasts shown by Hincks to be identical in form with those oi Sijacunjite

eximia, 370.
Steenstrup, his enunciation of the law of alternation of generations, 13, 101.
Steganophthalmia and Gymnophthalmia, division of medusse into, by Forbes, 14.
" Stiitzlamelle" of Reichert, a definite structure, 228.
Stylactis, employment of this generic name justified, 303.
Surface zone, its hydroid fauna, 166.
Succession of zooids in the gonosome, centripetal or centrifugal, 108 ; compared witli the inflorescence of

plants, 108—109.
Syncoryne, provisional and doubtful species of, 284.
Syncoryne pukhclla, congenital union of medusa; in, 202; retrograde changes of form in its mcdus;ic,

203.

450 GENERAL INDEX.

Synthcciiim, origin of gonaiigium within the hydrotheca of, 229.

Teutacula, their variations, 25 ; structure of their endoderm in the liydraiuli, 12.J ; and in the piano-
blast, 127.

Teratology of the Hydroida, 200.

Thamnocnidia, a sub-genus of Tubularia, 399 and 406.

Thompson, J. v., he recognises the essential structure of the Polyzoa ; his employment of the name
I'olyzoa, 11.

Thread-cells, their structure and mode of action, 118 ; their form in Gemmaria implexa, 119; observa-
tions of Moebius on, 120; of Mr. Lewes on, 229; their evolution a physical act, 121; their
development, 122 ; their formations, 128, 229.

Thuiaria articulata, its great bathymetrical range, 165.

Traropsis, development of the egg in, 98; ocellus-like pigment-spot in, 139.

Timaformosa, development of eggs in, observed by Alex. Agassiz, 97.

Touch, sense of, its seat, 144.

Trachynema ciliatum, Gegenbaur's account of its development, 80.

Trachynemites deperdita, a fossil hydroid medusa of the Upper Jurassic strata, 174.

Trophosome, 23.

Tubularia, its animality first demonstrated by Bernard de .Jussieu, 5 ; development of ovum in, 90 ;
endoderm of hydranth in, 124; canalliculation of endoderm in, 124; dismemberment of the
genus by Agassiz, 399 ; provisional species of Tubularia, 420.

Tubularia indieisa, -Aimtomy of; structure of its stem; structure of its hydranth, 205; fibrillated tissue in
its tentacles, 206 ; structure of its gonophores, 207.

Tuhdaria mesembryanthemum, tubular processes of its spadix ; structure of its stem ; irritability of its
hydranth, 420.

Tubularia rcgalis, great size of its hydranth ; its high northern area, 405.

Turris, development of the egg in, 97.

Umbrella of planoblast, 33; its structure, 113; epithelial layers in, 113; fibrillated tissue in, 114; its
structure in Obelia, 115; its ofRce in locomotion, 136.

Vegetation, marine, its relation to the bathymetrical distribution of animals, 164.
Velum, 33 ; fibrillated tissue in, 114; its oflice in locomotion, 136.
Vogt, Carl, his group of the Hydro-medusse, 15.

Wagner, Rud., describes the production of eggs by the medusiform buds of his " Hydra {Podocoryne)
aculeata," 342.

Wehstcria crisioides, fossil of the London Clay, possibly a hydroid, 174, 7iote.

West Indian province, 163.

Wright, Dr. Strethill, observes the development of the eggs in Turris, 97 ; in T/nii/ma/ilias i/icoaspicua,
97; observes the development of the planula in Zygodactyla, 97; describes palpocils. 111 ; spiral
zooids of Hydractinia first described by him, 223 ; describes certain filiform appendages in
Hydractinia; describes the canalliculation of the stem in Tuhtdariaindivisa, 125; distinguishes a
scattered from a clustered form of Clava, 247 ; his revison of the genus Eudendrium and institu-
tion of the genus Atractylis, 299.

Yolk, cleavage of, in the Hydroida, 87 ; its differentiation round the germinal vesicle, 149.

Zoantharia, diagnostic characters of the, 188.

Zoography, hydroid, general principles of, 235.

Zooid, definition of tliis term, 22, tiote ; genetic succession of zooids, 101.

Zooidal and embryonal multiplication in the Hydroida, relation between these, 101.

Zones of depth, 163.

Zygodactyla, development of planula in, 97.

Date Due

^

'>m oW