[From the Proceedings of the Boston Society of Natural History,
Vol. XV., March 19th, 1873.
ON THE
SYSTEMATIC POSITION
OF THE
BRACHIOPODA.
BY EI)WARD S. MORSE,
PlOF. COMP. ANA TOMY AND ZO,(LO(GY, lBO\WDOIN COLLEGE.
LINGULA PYRAMIDATA STIMI'SON. FROA LIFE.
BOSTON:
PRESS OF A. A. KINGMAN.
1873.








[From the Proceedings of the Boston Society of Natural History,
Vol. XV., March 19th, 1873.
ON THE
SYSTEMATIC POSITION
OF TI-E
BRACHIOPODA.
BY EDWARD S. MORSE,
rlPOF. COMIP. ANATOMY AND ZOOLOGY, BOWDOIN COLLEGE.
LINGULA PYRAIIDATA STIMPSON. F1ROM LIFE.
BOSTON:
PRESS OF A. A. KINGMAN.
1873.








THE SYSTEMATIC POSITION OF THE BRACHIOPODA.
BY EDWARD S. MORSE.
To Japetus Steenstrup,
Who first recognized the Annelidan Affinities of the Brachiopoda, this Contribution
is respectfully dedicated.
PREFACE.
To the systematic zoologist it is instructive to mark the changes
that have taken place in the classification of animals within the last
thirty years, changes not only resulting from  further knowledge of
the internal structure of animals, and of their embryology and early
stages, but changes resulting from a new interpretation of data previously well known.
Up to comparatively recent times, such distinguished authorities
as Agassiz, and Vogt, have suggested the association of the Vorticellidcla and Foraminifera2 and the Ctenophora3 with the Mollusca.
Not only were these additions rejected, but already have been re-.moved the Cirripedia, and the shell-bearing Serpula, Spirorbis, and
other forms originally grouped with the Mollusca.  Are we now to
believe that this work of elimination has ended?  If further dismemberment of this perplexing branch is to take place, one would naturally look for it in that association of classes called the MolluscoiI Louis Agassiz, Essay on Classification. 8vo. ed., p. 108.
2 Ibid, p. 113.
s Carl Vogt, Zoologische Briefe.
(315)




4
dea, first separated by Milne Edwards, and afterwards adopted by
Dana, with pregnant suggestions as to its value as a group by itself.
Recently Kowalewsky, Kupffer, Schultze and  others, have assailed the Tunicata, and demonstrated their kinship with the lower
Vertebrata through Amphioxus.  Leuckart has long maintained
that the Polyzoa have no sort of relation to the Mollusca, but belong
to the Vermes, and recently the distinguished Gegenbaur, in the second edition of his Outlines of Comparative Anatomy, not only assigns
the Polyzoa to the Vermes, but places there also the Tunicata. And
now  in this paper I wish to show that in every point of their
structure, the Brachiopoda are true worms, with possibly some affinities to the Crustacea, and that they have no relations to the MIollusca, save what many other worms may possess in common with them.
In nearly every case the unnatural association of certain groups
with the Mollusca has been due entirely to superficial resemblances,
to' formal analogy," as Forbes would say.
The same reason that first led conchologists and zoologists to include Spirorbis and Serpula and the Cirripedia, as well as the Foraminifera, with the Mollusca, namely, the presence of a calcareous
shell, also brought the Brachiopoda into the same category.  But
while there was some resemblance between the cases of certain tubicolons Annelids and the shell of Vermetus, or the flattened form
and lateral hsells of Anatifa, and the lamellibranchia-te shell, or the
chambered shell of certain Foraminifera and the Nautili, there was
but little to suggest an affinity with the lateral lamellibranchiate
shells, in the dorsal and ventral plates of the Brachiopoda.
The mere possession, however, of a calcareous shield of some sort,
whether in one piece, or several pieces, whether a tubular or a chambered shell, furnished sufficient reasons for most zoologists to include
creatures bearing such shelly coverings with the Mollusca.  Hence
we find Lamarck, at one time placing Anomia and Discina together.
And Cuvier, allowing the accepted views of the time to lead him
astray, forsook his principles based upon internal structure, and regarded the relations of the Cirripedia as molluscan.
It is amusing now to look back and see with what quiet resignation
the conchologists (for such they were rightly called) permitted the
removal of those forms which possessed no shelly covering, with what
stolid indifference they allowed other unprotected forms being forced
upon them, and with what obstinate pertinacity they withstood the
removal of such groups as possessed a limy shell.
(316)




5
Says Mr. G. B. Sowerby, the great English conchologist, after
Thompson had so clearly shown that the Cirripedia were crustaceans
and not molluscans; " Without describing the facts, or entering upon
the arguments, with which he (Thompson) supports this opinion, we
must be permitted to say that we do not think he has fully'demonstrated it; at the same time, considering that, as far as we hitherto
knew, the Cirripeds were all attached, the circumstance of their being
free when very young accounts well to our mind for the fact of
each species being found  attached to peculiar situations, which
would only be compatible with the notion of their being at one time
free agents, and possessed of an intuitive volition, determining their
choice of situation." 11
Every worker knows how blindly one will work, when his mind
is imbued with the accepted views of the subject, when he does not
dream  of questioning what he has always been taught to believe,
particularly when those teachings come from the highest authorities.
Even so distinguished a naturalist as Prof. Huxley, after he had rerepeatedly  observed  the  external openings of the oviducts in
Rhynchonella, confesses that " pre-occupied with the received views
on the subject (namely, that oviducts were hearts), I at once interpreted them as artificial."2  In the same way Prof. Owen thought he
saw a minute perforation at the extremity of the intestine of Terebratula, where no such opening exists.  As Lingula, and Discina had
an anal opening, it was quite natural to believe that the other
Brachiopoda formed no exception to the rule.
Many elaborate investigations of the Brachiopoda had been made
by such eminent naturalists as Cuvier, Vogt, Owen, Hancock, Huxley, Davidson, Lacaze-Duthiers, Gratiolet and Carpenter, and in all
their memoirs no doubts had been expressed as to their molluscan
nature; therefore, on commencing the study of the Brachiopoda, thirteen years ago, I had no more doubt of their molluscan character,
than  of the vertebrate character of birds, and  attempted  only
to show more closely the homologies which I believed existed between
the Brachiopoda and Mollusca.  When at last they had been forced
into the place where I believed they rightly belonged, the result of
that work was published in the Proceedings of the Essex Institute,3
i Sowerby, Genera of Shells.'Huxley, Proc. Royal Soc., London, Vol. vii., p. 113.
s Classification of the Mollusca based upon the Principle of Cephalization.
Proc. Essex Inst. Vol. iv. 1865. And Silliman's Journal, Vol. XLII, July, 1866.
(317)




6
and afterwards republished in Silliman's Journal.  The fact that in
that paper the Brachiopoda were turned up side down, and end for
end, shows the violent methods resulting from faith in accepted views.
It is a simple matter of justice to myself that I make this confession,
and I may also say that my studies of the Brachiopoda hlave been
made, not for the purpose of describing new species or genera, to
show their geographical distribution, or to tabulate the number of
species known, but simply and solely, to determine their affinities;
and that some weight may attach to the radical views here advanced,
I may, with satisfaction, state that my investigations on the subject
embrace a series of observations on the anatomy and early stages,
of Discina, from  an immense mass of material in alcohol furnished me by Prof. Verrill.  I have also carefully studied living Lingula, Rynchonella, and Terebratulina, and the early stages and embryology, of the latter.
Some of these investigations have already been published, and I had
hoped to present them all before publishing this paper, but as some
time will be required to prepare the results, and the necessary plates
on Lingula and Discina, I am reluctantly compelled to present this
first.
INTRODUCTORY CONSIDERATIONS.
The changes here proposed in the removal of the Brachiopoda
from  the Mollusca, and their association with the Vermes, make
necessary a comparison between the Mollusca, as now restricted, and
the Vermes.
Many naturalists now  hold the opinion that the Mollusca are
descended from the Vermes. Indeed, it would seem from the rapidly
accumulating data that the Vermes underlie the whole animal kinodom, with the exception of Protozoa.  Only on this hypothesis, that
the Mollusca are derived from the Vermes, can we understand the
otherwise strange assemblage of characters displayed by such Mollusks as Chiton, Dentalium, Pneudermodon.
In our comparisons we are justified in selecting as typical Molluscal
those groups which have remained unchanged the longest.
We leave out of consideration the Polyzoa, since they are not only related to the
Brachiopoda, but because they are regarded as worms by Leuckart, Gegenbanr, and
many others, and also the Tunicates, regarded by many naturalists as forming the
base of the vertebrate series; others placing them with the Vermes, and by all
separated from the Mollusca proper.
(318)




7
A typical Lamellibranch, and a typical Gasteropod, will be admitted by all, as best representing this branch, for while other groups
have widely changed since their first appearance in past ages, we
find the Lamellibranch and Gasteropod of the lower Silurian as typical as present existing forms, e. g., Modiolopsis, Avicula, Murchisonia, Pleurotomaria, and from the tracks and tubes, and still later
setae, we are safe to assume that the Annelids were as characteristic
of their classes in past geological times, as at present.
We cannot compare the Trematodes and Turbellarians, with the
Nudibranchiate Mollusks, for however much resemblance some may
see in their adult condition, as among the Planarians,' certain characters of external symmetry in common, their respective embryos are
identical with their respective divisions, the one being annulated, the
other developing a foot, and a nautiloid shell.
Leaving these out of consideration, then, and taking the dominant
characters displayed by the Vermes on the one hand, and the Mollusca
as cited on the other, we have in the Vermes, a form, whose length is
much greater in proportion to its breadth than in the Mollusks; the
Fig. 1.                       Fig. 2.
se *if
e   i     
Transverse section of Annelid  Transverse section of molluscan
after Carus.                archetype after Carus.
b. bands suspending intestine in perivisceral cavity. i. intestine. s. segmental
organ. se, se. setie.
worms being drawn out as it were, the Mollusk being concentrated.
The worm  is perfectly bilaterally symmetrical, depressed, flattened
or circular, the dorsal and ventral regions so near alike in many cases,
as to be distinguished with difficulty, and the body never flattened
laterally.  The Mollusk is also bilateral, but often asymmetrical,
the dorsal and ventral regions are very unlike, and the body almost
always flattened laterally.  This latter character is so marked, more
especially among the Lamellibranchiata, as to have led Prof. Agassiz
I Girard placed the Planarians with the Mollusks.
(319)




8
to apply the term  laterality as distinguishing the Mollusks, while
the term  tergality was applied to the whole Cuvierian branch of
Articulates.  Agassiz has also called attention to the fact that while
the display of structure is upon the sides in Mollusca, it is upon the
back in the Articulata, though numerous and important exceptions
occur in both groups.
In the worm, the locomotor muscles are intimately connected
through the entire length of the body with the integumentary system,
especially on its dorsal and lateral walls (Rolleston). In the Mollusk, on the contrary, the locomotor muscles are connected ventrally
with a specialized creeping disk, the foot.
In the Mollusk, with few exceptions, the viscera are carried, above
the foot in a protruding chamber.
Fig. 3.
Molluscan archetype from Carus.
In the worm, the symmetry of the body is never disturbed by
the viscera.  The tegumentary envelope, when separate from the
body proper, forms at most a projecting, or an everted collar about
the head, and, in a few instances, a continuous free membrane along
the sides of the body. In the worm this envelope, and adjoining
parts possess chitinous outgrowths in the shape of scales, rarely a
shell, but commonly setae, the latter being a marked character of
the Vermes.
In the Mollusk the tegumentary envelope is prolonged, and oftentimes continuous, forming a sack or mantle, inclosing a conspicuous
cavity, and protecting the gills. Hence the name Saccata, proposed
by Prof. Hyatt. Sets, or scales, are not present; while the possession
of a calcareous shell, composed of one or more pieces, furnishes the
only material to nine-tenths of those who study them. In the worm
the plates, when present, and the thickened integument, are perforated with minute tubules, a character not possessed by the Mollusk.
In the worm, the digestive canal is straight, rarely convoluted,
and suspended freely in the perivisceral cavity. (See Fig. 1, b.)
(320)




9
In the Mollusk, the intestine is always convoluted, not suspended
freely in the perivisceral cavity, but intimately blended, or united
with other organs.
In Vermes there is a peculiar depuratory apparatus characteristic
of all. In the Annulata this apparatus takes the shape of bilaterally
symmetrical tubes, in pairs, opening externally and communicating
with the perivisceral cavity by distinct independent infundibilform
orifices. (See Fig. 1, s.)
In the Mollusca, with exception of certain Cephalopoda, nothing
of the kind is found, and where such communication does exist
between the organs and the surrounding medium, it is by means of
simple orifices in the walls of the cavity.
In the Vermes, especially in the Annulata, a nerve collar is found,
from  which start two parallel chains of ganglia, oftentimes widely
separated.
In Mollusks there is also a nerve collar surrounding the cesophagus,
and no double chain, but nerves are thrown out to the sensory, motor,
and parieto-splanchnic regions.  Hence the names Homogangliata,
and Heterogangliata. In the Annulata, with the exception of the
Discophora, the generative products are set free in the perivisceral
cavity, receiving from the fluid therein contained, certain nourishment. In Mollusks this never occurs, though in certain  Cephalopods the products of the generative organs are set free into a compartment of the perivisceral cavity, and from there find egress through
the oviduct.
With the exception of the Octopoda, the             Fig. 4.
oviduct is single. In the latter group there                  s
are two oviducts bilaterally symmetrical,
all of these features being vermian.
In Vermes the embryo never possesses
a single or double shell, and with few         n  
exceptions is distinctly annulated,l while) 
among the Mollusks, even when devoid          b" 
of a shell in the adult, the embryo early                      -— f
develops a shell composed of one or two
pieces.
^.~pieces~.    ^   ft''Embryo of LamellibranOther differences of minor importance              chiate.
might be mentioned as separating distinct-    s, shell; m, mouth; f,
ly the typical worm, from the typical Mol-   footi b, byssal gland;
lusk, but the leading characters here pre1 See "NOTE," p. 10.             (321)




10
sented sufficiently indicate the wide divergence of the two great
Divisions.
NOTE. The annulated embryo of the worms is characteristic of most of them,
from the Rotifer to the highest Chetopod. In all, the body is generally divided
into a few transverse segments. In the Lamellibranchiate and Gasteropod the
embryo early develop the velum, or foot, projecting from a bivalve, or a nautiloid
shell. In Chiton the larva is annulated, according to Loven. Pneumodermon,
among the Pteropods, has the body banded by transverse circles of cilia. In
Dentalium the larva resembles that of a true worm.
The affinities of Dentalium are not clearly understood. It was placed among
the Annelids by Cuvier and Lamarck, and then among the Mollusks by Deshayes
and De Blainville, as Gasteropods. Since then they have been bandied about
from one end of the series to the other. Lacaze-Duthiers,l who has made the
most thorough investigation of them, makes a separate class, Solenoconchia,
with their affinities mostly among the Lamellibranchiates. Huxley places them
with the Pteropods, on account of the rudimentary head, neural flexure of intestine, presence of epipodial lobes, and the character of the larva. With all
these diverse relations, I would suggest that they certainly bear some relations
to the Tetrabranchiate Cephalopods, in the numerous and retractile tentacles,
the dorsal turn of the shell, and the strict identity between a peculiar bilateral
cartilaginous body which occurs in the head of Dentalium, as well as in the
head of Nautiluspompilius.
Having thus connoted the leading features characteristic of each
Division, our next object is to inquire whether all the characters
of major and minor importance possessed by the Brachiopods are not
held in common by the worms, and are in no wise possessed by the
Mollusks.
General Proportions of the Body.
In Mollusks, while we may have the body divided into a creeping disk, and a visceral portion, the visceral portion  usually  carried above in a protruding chamber (See Fig. 3), or the mantle
prolonged  behind, to form  the tubes, we do  not have the body
constricted transversely, forming a thoracic, and an abdominal portion.   We do not find  such a feature as  a caudal appendage,
nor are the Mollusks ever attached, save by the adhesion of the
calcareous shell, or by the byssus.  Among the lower worms, as, for
example, some Rotifera, certain forms are fixed by their posterior
portion.  In  the Rotifera, as well as in the tubicolous Chmatopods,
1Annales des Sciences Naturelles, 1866-57.
(322)




11
the body is in most cases distinctly divided into a thoracic portion,
and an abdominal, or caudal portion.  In Pectinaria the caudal portion is separated from  the body by a deep constriction, and is apodous. In Serpula, Protula, and Ly-          Fig 
silla, the separation of the body into two regions is
strongly marked.  In Sabellaria alveolata, the caudal
portion is very long, cylindrical, and apodous.
In the Brachiopods, the body is also distinctly di- 
vided  into a thoracic and caudal portion.  The cau-.
dal portion varying greatly in function and charac-  Young UotiIfr.
ter, in different groups. In some, this portion is very
short, and firmly fixed to some point of support.  In Rhynchonella
and Terebratulina it is capable of sustaining the body, and of twirlino it round in various directions, or more correctly the peduncle
appears to be firm  and elastic, and the body is capable, by certain
muscles, of twirling round upon it.  In other Brachiopods, as in
Discina, the peduncle, or caudal portion, has its special set of muscles, attached to the outside of the ventral plate, and its cavity is
in direct communication with the perivisceral cavity by an azygos
opening.
In Lingula pyramidata (and I presume the other forms of Lingula
will present no essential difference) the peduncle is nine times as
long as the thorax, free, active in its
vermian contortions, and possessing the              Fig. 6.
power of fabricating a sand tube. The
thorax also possesses this power.  (See
figures 1 to 7 in Plate I.)  Not only      
is the body often enveloped in a sand
case, but this species of Lingula has                  i  - >
the power of covering the bottom of    <'"" 
any vessel in which it may be confined                         @
with a sinuous sand tube, precisely similar to the tubes made by Terebella,                              
and allied forms under like circumstan-   grla pyranidata showing ances.  (See Fig. 7, Plate I.)  Though   nulations and circulation of
fluid within.
the peduncle of Lingula is capable of
varied and rapid movements, is partially annulated, shows a constant
circulation of the perivisceral fluid within, is possessed of mucous
pores, yet no trace of setm are seen upon its walls, and this is in
(323)




12
accordance with what we see in worms, that those segments without
appendages or setm are the caudal ones.
A prominent character of the higher worms is the annulations, or
rings marking the body. In the Gephyrea, however, this feature is
not so obvious as in the peduncle of Lingula, while in many of the
lower worms, e. g., Chvetognatha, Nematoidea, Acanthocephala,
there are no segments, and in the Rotifera the segmentation is
external.  Therefore the absence of this character in the Brachiopoda is unimportant.
The presence of more than one segment in the Brachiopoda is
possibly  marked in Rhynchonella, where two pair of segmental
organs, or oviducts, occur.  In Lingula, also, a deep constriction
occurs just back of the posterior occlusor muscles, a membraneous
partition at this point tends to separate the perivisceral cavity, and
the stomach has a corresponding ridge upon its walls.  All these features certainly suggest segmentation.   The arrangement of the
muscles in Lingula into distinct sets, first the anterior occlusors,
then the posterior occlusors, next the external, central and posterior
adjustors, and finally the divaricator muscles, suggests segmentation
of the body, as seen in the Arthropods.  This feature is less marked
in Discina, though still apparent, and with their external peduncular
muscles one might, with propriety, theoretically form  a number of
rincs.
The dorsal and ventral symmetry is a distinctive character in worms.
This symmetry is often so complete as to render the determination of
above and below a
Fig. 7.                       matter of great dif6,                      ficulty, and, as in
Sternaspis, a source
—;se      of confusion.  (See
transverse section of
-:se          Annelicl, Figure 1.)
~" ~~0 ~    ~~The same  feature
is likewise characTransverse section of Lingula.                     
teristic  of  many
b. bands suspending intestine in perivisceral cavity.   Brachiopoca,  pari. intestine. s. segmental organ. o. ovaries. 1. liver. g. gills. se, se. setxe.                     ticularly with Lingula, where  these
regions externally are almost precisely alike, and where single valves
of Lingula are found fossil, or their impressions are seen upon the
(324)




13
rocks, it is almost impossible to determine whether they are dorsal
or ventral.
No one, however, would mistake these regions in the Lamelllbranchiate, or Gasteropod.
In the Brachiopoda, with the exception of Lingula, there is a
a great concentration of the body, quite unlike anything seen in the
Vermes. Lancaster, however, describes, in the Annals and Magazine of Natural History, a worm, Clacetogaster verrmicularis, one
of its chief points of interest being the exceedingly small number
of segments, four or five only.
~Many Rotifera are also highly concentrated, or cephalized, with
dorsal and ventral flattening, and with a chitinized integument.
As to this concentration of parts in the Brachiopoda, it would be
strange indeed if the worms alone should not show this concentration of structure in some of their forms.  This same diversity occurs
in all the other groups, as in the Crustaceans, the highly cephalized
Brachyurans, and the elongated Macrourans, and among the Cirripeds, the concentrated and flattened Coronula, and the long pedunculated Anatifa.  Or among the Echinoderms, the flattened Mellita
or Scutella, and the worm-like Holothurian.   Or, again, in the
Lamellibranchiates, the concentrated Isocardia and the attenuated
worm-like Teredo.  Other examples might be given in the Polyps,
Ctenophors, Gasteropods, Insects, Fishes and Reptiles.
Concentration, or cephalization of a structure, while modifying
the character and functions of parts, and even  obscuring their
ready interpretation, can in no wise affect the relation of the animal
sustaining such features, though it may account for certain peculiarities attending such conditions, in the same way that parasitism
may account for the absence of certain organs, characteristic of
related forms.
Integumentary Organs.
In Mollusks the tegumentary envelope is almost always extended
to form a sac, or mantle, which is open below, sometimes resting like
a cap upon the back, or better, extending itself in a wide membrane
about the viscera, or it may hang upon the two sides of the body,
split below in a median line, but not on the sides.
This envelope, or mantle, usually secretes a shell composed of carbonate of lime, and is attached to it by special muscles in limited
areas, so that when these are separated, the envelope is found to
(325)




14
have no sort of connection with the inner walls of the shell secreted
by it, from  which the shell readily drops.  This feature renders
possible the formation of pearls, by irritating substances or parasites,
finding their way between the mantle and shell.
The molluscan shell is never perforated with tubules passing perpendicularly through, from one surface to the other, nor are there
any minute ramifications of the mantle, or, other portions of the
soft parts, entering the substance of the shell, and consequently no
adhesions of the body, save by the special muscles above alluded to.1
In the Annulata the integument is rarely ever extended beyond
the limits of the body.  When this is the case, it forms a broad
crmembrane bordering the thorax, as in Protula, Serpula crater (See
Plate I., fig. 10), and others, or it surrrounds the head in a collar,
often everted, split upon the sides and notched in the median dorsal
region, and separated in a median line below.  In Protula Dysteri,
a broad membrane borders the lateral aspect of the thorax, from
which the seta spring. In Serpula crater, the membrane borders the
thorax diagonally, being free at the posterior dorsal region of the thorax.
In Brachiopods we have an extension of the tegumentary envelope
from above and below, enclosing the arms.  This membrane is also
split upon the sides, and is directly to be comFig. 8.        pared with the cephalic collar in certain tubiculous annelids, as Sabella, for instance,
\if[      where it differs only in degree.  In Sabella
the cephalic collar does not cover the bilobed arms, but it is a split upon the sides,
and notched in the median dorsal line.  In
K/~-~/'~%~    -.  many Brachiopods there is also a notch in
/     the median line, and the genus Pygope of
Link, the notch divides the collar into two
Showing cephalic collar
of Sabella.      lobes, which afterwards unite, leaving a hole,
m. median dorsal notch.  or space in the shell.  In other words, this
1. lateral notch-      membranous collar covers simply the base
of the arms in those worms possessing it. While in the Brachio1 In his work on the Microscope, Dr. Carpenter has described, as peculiar to
Anomia, an irregular net work of minute tubules, running parallel to the surface
of the shell, scantily distributed in the inner layers, but very abundant in the
outer layers. In his last edition of this valuable work, Dr. Carpenter explains
the character of these minute tubules, and refers them to the action of a parasitic
fungus. Mr. Mark Stirrup, at a meeting of the Manchester Philosophical Society,
England, exhibited sections of various shells, and showed that in Anomia, the
ramifying tubules were produced by a fungoid growth.
(326)




15
pods the collar covers and  protects the  arms, which, however,
may partially project, as in Terebratulina and Lingula, and wholly
so in Rynchonella psittacea, as I observed last year.1
Furthermore this pallial membrane, or cephalic  collar of the
Brachiopods, is not to be compared to the mantle in the Mollusca, as pointed out by Dr. Carpenter in  1854.  In  a paper on
the  peculiar arrangements of the  sanguiferous system  in  Terebratula, and  other  Brachiopods, 2  he  says: "The  membrane
which is commonly spoken of as the mantle, and which may be
stripped from the shell by the use of sufficient force to overcome its
adhesions, must, I maintain, be considered as really its inner layer
only; for I find that an outer layer exists, so intimately incorporated
with the shell as not to be separable from it without the removal of
its calcareous component by maceration in dilute acid.  When thus
detached, this outer layer is found to be continuous with the membrane lining the perforations in the shell." I have observed that
when the test 8 is removed in Lingula pyramidata, the perivisceral
cavity is often exposed, of such extreme tenuity is the inner lining
membrane.
From  a figure and description given byW. Baird,4of a peculiar worm case
Terebellafiabellum, the cephalic collar might have had the proportion of many
Brachiopods, in being broader than long. He describes the orifice of the tube
as being circular, and says, "the most characteristic feature, however, in the
structure of this tube, is the fan-shaped expansion of filaments at its upper
orifice. This orifice is circular, and has on its dorsal surface a projecting lip, or
kind of hood, which extends beyond the mouth for a short distance, whilst from
the ventral side springs another lip or hood."
Dr. Dawson5 represents the worm case of Vermilia serrula from the Gulf
of St. Lawrence, in which there is a marked thoracic enlargement.
In a few  worms only, do we find dorsal scales, as in Polynoe, or
posterior and ventral scales, as in Sternaspis, and when these occur,
they are chitinous, as in Discina. In Brachiopods the dorsal, and
ventral shells, or plates, are unlike anything we know of in worms.
Their composition and structure, however, their dorsal and ventral
1 American Journal Science and Art, Vol. iv., Oct., 1872.
2 I.roc. Royal Soc., London, Vol. vii., p. 32.
3 The test of this species, when dried, wrinkles and folds together like the scute
of Lepidonotus.
4 Journal Linnaan Soc., Vol. viii., p. 157, pl. 5, figs. 1, 2.
a Canadian Naturalist, Vol. v., p. 24.
(827)




16
arrangement, the fact that in Lingula and Discina and allied forms
the shells have their borders free all round, while in those that
interlock at their posterior margins, there is no ligament to act upon
them, as in the lateral shells of the bivalve Mollusk, all these features
together preclude the possibility of any comparison between them
and the molluscan shell.  It is, therefore, more natural to regard
the Brachiopod shell as a dense and thickened integument, to be
compared to similar regions in the Arthropods and in the worms,
simply as dorsal and ventral plates, and from certain considerations
to follow, we believe this relationship will be admitted.  The presence of nearly fifty per cent. of phosphate of lime in the test of
Lingula, both recent and fossil, is a feature peculiar to the hardened
integuments of the higher Arthropods, and entirely unlike anything
found in the molluscan shell. In the hardened integument of Crustacea, tubular pores exist, which according to Dr. C. De Morganl are
organs of general, or special sensibility, as he finds in many cases
the tubules surmounted by hairs.
He says, " The shell canals are comparatively fine, more resembling coarse dentinal tubes, but they are lined by a sheath, and have
contents prolonged from  the vascular layer.  The relations of the
contents of the tubes to the internal integuments, may be shown by
tearing away the latter from the shell, when the contents will often
be drawn out of the canals; and it may be seen that they are prolongations of the outer layer of the integument, enclosing the elements of the vascular layer within their cavities."
In the Annelida, also, there appears a system of minute pores,
and Kolliker (we copy from  Claparede) asks whether they are not
homologous to the tubular pores (porenlandle of the Germans) of the
Arthropods, or whether they may not be compared to the apertures
of the cutaneous glands, such as those discovered by Mr. Leydig in
the Piscicolae.  To this, M. Claparede says, positively, that the two
categories of pores exists in the Annelida.  Those which serve for
the discharge of certain secretions seem to exist in all species. He
says, furthermore, that the canalicular pores are much smaller, and
much closer together, and,do not correspond with glands, and that
they occur only in the species with a thick cuticle, and not even in all
these.
In the test of most Brachiopods, similar minute tubular canals
1 On the structure and functions of the hairs of the Crustacea. Phil. Trans., London, Vol. 148, p. 897.
(328)




17
exist. They do not open outwardly, at least, not so far as known at
present.  In the test of Lingula pyramidata, they  are exceedingly
minute, and  closely  crowded  together.   r.. Gratiolet  has also
observed them  in the test of L. hians.   In  Discina  I have failed
to  find  them.   Dr. Carpenter,1 who  is one of the  highest  authorities on the subject, states  that thesetubules  are  intimately
connected  with the vascular  layer, which  sends ccecal prolongations into each one of them.  Hancock questions this view, and
states that from  his observations, the coecal processes spring fiom
the reticulated layer of the pallial membrane, though he admits the
constant presence of corpuscles in the cceca, which strongly resemble
the blood corpuscles.
Hancock, in speaking of the tubules, says: "The best mode of investigating
these organs is to dissolve the shell, and t thethey are exposed in various stages
of growth, adhering to the margins of the mantle. They are arranged in rows,
and are cylindrical, with the distal extremity obtusely rounded, and are pedunculated from the first; the peduncle is long and narrow; the cceca at the extreme
edge are small, but rapidly increasing in size backward; the terminal, or enlarged portion, is almost constantly stuffed full with the so-called blood corpuscles. When observed in this way, these organs have very much the character of secerning follicles, but what function they really subserve is difficult to
determine; it may be that they have something to do with the growth and
reparation of the shell, though it is not easy to understand how. They are
-probably, as suggested by Prof. Huxley, the homological representations of the
vascular processes that penetrate the test of the Ascidian; and if so, it would
seem  likely that they have lost much of their functional importance; and, in
fact, their entire absence in forms closely allied to those in which they are
highly developed, augurs that they are not of high functional importance."
Albany Hancock, on the organization of the Brachiopoda. Trans. Phil. Soc.,
London, Vol. 148, p. 837.
See also Prof.. King, on the Histology of the test of class Palliobranchiata.
Trans. Royal Irish Acad., Vol. xxiv, part xi, 1869. In some observations on
the early stlages of Terebratulina, I found the tubules in the very youngest
stages of the shell.2
Claparede states that only worms with a thickened integument have
those peculiar canals, and not even all these, and accordingly we
find in Brachiopods, as worms with  a  thickened  and  indurated
integument, that while  many  possess these  canalicular tubes, in
others they are quite absent.
1 Proc. Royal Soc. London, Vol. VIT., p. 32.
2 Memoirs, Boston Soc. Nat. lHist., Vol. II.
2                                        (329)




18
It seems then reasonable to believe that some sort of relation exists
between this marked  feature of the Brachiopod test, and  similar
features in the worms and Crustacea.  In the Brachiopods these
tubules may have undergone some functional change, but until otherwise proved, they must be regarded as a distinct vermian character.
In a large sipunculoid worm  from  Panama, kindly loaned me by
Prof. Verrill, the inner lining membrane sends minute cccal processes into  the integument, which  is much thickened.  There is
possibly herb some relationship.
We also  find  in the  Annelida mucous pores.   This character
must be  common to all those  animals secreting mucus from  the
surface, whether Mollusk, or Annelid; and as the peduncle of Lingula
is glairy with  mucus, we should  naturally  expect to  find them
present.  This I easily succeeded in doing in living Lingulae, using a
1 inch objective of Powell and Leland.  They were very minute,
and closely crowded together.  Their presence has never before been
observed.  The test of Lingula was also very glairy, but whether the
mucus, which appeared to cover it, was exuded from  the test, I
could not determine.
Setc.
Fig. 9.
In  Mollusks, locomotion  is  effected
mainly by a special organ, the creeping'
u -liX  |disk or' foot."  There is never secreted
c iK',  \'\'\\ll   hairs, spines or setae.,\ \\ 1 I                 In the embryo of some Lamellibranchiates,
ifi!}'ll/l~il'I^ llt~ll  there is said to occur two or three little spines,
which are arranged along the ventral median
line.
a  11...     1:'"il      Ina some species of Chitons, tufts of stiff
b-'-'g'      spines issuing from the girdle have long been
A'10;   known.
>.~~      /?""X!      Whether these spines are homologically reTA      Bi'" "    \lated to the setm of worms, remains to be
seen. From a cursory examination of them,
Ciiton spiculosus Gray.    they appear to be modifications of the minute
A. side view of Chiton, mag-  pavement-like granules that occur in the girnified. B. side view of one   dle of mann  Chitons.
tuft of bristles in the girdle.
C. a tuft of bristles largely    In Chiton spiculosus Gray, (alcoholic) fiom
magnfied. a. line of girdle.  Gaudaloupe, kindly  loaned me for examinab. base of tuft.(3
(330)




19
tion by Mr. Bland,the tufts of spines readily separated from the girdle; the
whole tuft is closely united together, and seems to be almost entirely superficial. The following figures represent an outline of the
species, with a tuft separated and enlarged as well as a       Fig. 10.
single tuft, as it appears upon the girdle, showing as
well the minute spines which project from the exterior  \
border of the girdle. The bases of the spines in the tuft
are abruptly truncated. 
A  section of the girdle of Anzicula Emersoni shows         \ 
the shorter granules, which are but slightly embedded, 
as well as the longer spines, which are more deeply                  i
seated, and one spine intermediate between the long and 
short ones, which is again only partially embedded.                
These all arise from  the homogeneous cartilagineous    i
portion of the girdle, and do not reach to the muscular
layer beneath, from which they are entirely separate.
They are therefore probably immovable, save what mobility attends the folding of the girdle.:Amicula Emersonii
In  worms, there  is found, as characteristic
a. cartilaginous
of the higher  as well as many  of the lower               layer. b. muscular layer.
groups,  the  production  of  sets, or  bristles,
which perform  important service in locomotion.  Claparede says, in
regard to the sefte of worms, that " some authors regard them as enclosed in a sac, which is only an invagination of the integument;
others think they are formed in an internal follicle, and only secondarily arise to the surface.  This second opinion only is correct.  In
certain cases (in Hesione and others, for example), the whole bundle
issues in  a compact form  through a single pedal aperture, but in
others, the seta has its own7 orifice; this is the case especially with
the flabelliform  bundles.  The pore from which each seta issues is
not previously formed, but is perforated by the seta itself." 1  Mr.
Lancaster regards the seta  in the earth worm  as secretions of the
so-called setigerous gland.
The arrangement of these setm in worms is usually upon the sides,
in two series, above and below.2  They are moved by muscles, and
not only have the power of protrusion, but freely  swing back and
forth.
A marked feature of the Brachiopods is the peculiar setm fringing
1In Pectinaria, a transverse row of long ones project forward from the head. In
Sternaspis, nine bunches project from the borders of each scutum behind.
2 See translation of introduction to Claparede's paper, Ann. Mag. Nat. Ilist., 3d
Series, Vol. xx., p. 314.
(331)




20
the borders of the extended membranes. (See Plate I.) As in worms,
they are arranged upon the sides and front, in two series, above
and below.  (See Fig. 7, and compare with  Fig. 1) and are also
secreted by regular follicles, each seta protruding from  the follicle
singly, or, in some cases, two or three setm issuing from  the same
orifice. In the fixed Brachiopoda, they have only a limited power
of motion. In these forms they are very short in the adult, but
very long in the young, even much larger than the animal, as I
have repeatedly observed in young Terebratulina, Rhynchonella, and
Discina; in  the latter genus, even  exceeding the length of the
animal three or four times. In the errantian Brachiopoda, as in Lin.
gula pyramidata, the setigerous follicles are entangled in a mesh of
muscular fibres, and locomotion is effected chiefly by them;  the
setae swinging freely back and forth, the dorsal plate oscillating from
side to side, as first noticed long ago in another species of Lingula,
by Carl Semper.
In  Discina, the  setm  are very long, and  crowded together.
According to Fritz MUller, in the early stages of a species of Discina,
from  the east coast of Brazil, the animal not only has the power
of swimming, but uses the larger pair of setm  thrust out behind, to
push itself along.  IIe says these bristles have great freedom  of
motion, sometimes thrust out horizontally, and again crossed to opposite sides.l
In the Annelid Arenicola, the first few anterior segments are setigerous only.  They bury themselves in the sand, forming a sand
tube, loose and not adhering, a tube which leaves room for the ramified branchia to display, and the setme, by arching over the branchia, protect them, and prevent the sand in which they are buried,
from falling in upon the gills.
Lingula pyramidata also protects the gills in the same way, as
I have  repeatedly observed  in  specimens  kept in  confinement.
When buried in the sand, the dorsal and ventral shells considerably
separated, the sete are brought together in such a p6sition that their
extremities meet, and the sand is thus kept back while the water
freely enters.
In worms the seta are often of various kinds in the same individual.  A  similar diversity is seen in the bristles of Discina.  The
sete of the embryo worm  are peculiar in being very long, strongly
Reichert ancd Du Iois-Reymond's Archiv, 1860, p. 72.
(332)




21
barbed, and deciduous.  In the young of all Brachiopods thus far
observed, the setae are also very long, as above remarked.  And Fritz
Miiller has shown that in the
Fig. 11.    embryo of Discina there are         Fig. 12.
also remarkably barbed setm
of great length, which  are
afterward disccrded.  Alex.
A(rassiz has also called atten-'      tion to the fact, that in Pal-          ii
eozoic worms the seta were 
barbed.,N 8           Tube Building.
The fabricating  of sand         Discina.
Nerine cirractulus.
tubes for the protection of  Deciduous seta of
Deciduous seta of
larval worm, lerine the body is not a charac- larvalDiscina,from
cirratulus, from Cla-                           Frtz Mller.
parade.           teristic feature of Mollusks.
Lima builds a peculiar "nest"
in attaching pebbles and fragments of shells together by byssal
threads, and imprisoning itself in that way. Gastrochena also forms
a flasked-shape cavity, in which it lives, and from which it has no
means of withdrawing. These features in Mollusks may be said to
bear only a remote resemblance to the tube building of worms.
In worms, the building of tubes is a prominent feature of several
groups, from the lowest to the highest.
Certain Rotifers, after attachment by their caudal portion, fabricate
a sand tube into which they retract.  (See Fig. 5.)
Many sipunculoid worms occupy the dead shells of Dentalium,
Littorina and other shells, and partially close the aperture, and even
extend it by a mud tube of considerable density.
Of great importance, however, in these comparisons, is the fact
that those worms, which  are  edentulous, which have  the body
divided into two regions-the thoracic and caudal, and which have
a bi-lobed lophophore, the two arms often appearing spirally twisted,
surrounding  the mouth, and supporting  ciliated  cirri, are  all
famous tube builders. Sometimes the case is gelatinous or chitinous,
often the tube is calcareous, deposited in successive lines of growth,
and resembling the shell of the Gasteropod; but more frequently
the tube is made of fine sand, mud, bits of shell, and coarser debris
that the builder meets with.  When Terebella is kept in confine(333)




22
ment in a bowl or dish, it covers the bottom of the vessel with an
irregular sand tube. When I first found Lingula pyramidata buried
in the sand shoals of Beaufort Harbor, North Carolina, I was surprised to find them living free in the sand, and not attached by their
peduncle as I had supposed. My astonishment was greater to find
that the peduncle was sheathed in a sand tube.  When this tube was
broken or removed from the peduncle, they promptly formed another
one. The shell, nearly to the anterior margin, would often be enclosed
in this sand case. When the peduncle was broken off, a bulb of sand
would soon be agglutinated to protect the broken end, and not only
sand was used, but bits of seaweed; and in one case a little stick was
incorporated in this structure.
I brought home with me to Salem, Mass., a number of living specimens, and these were kept alive in large bowls, from June till October, by imitating as far as possible their natural surroundings. They
would often protrude above the surface of the sand, and instantly
jerk back when alarmed.
On emptying the sand from the bowl one day, great was my surprise to find that all of the Lingulas had covered the bottom of the
bowl with large irregular sand tubes, cemented to the sides and bottom of the dish, the tubes running over each other, and presenting
precisely the appearance as that produced by Terebella and allied
forms when kept in dishes in this way.  (See Plate I., Fig. 7.)
In this place it is proper to state that the peduncle of Lingula is
highly mobile. When removed from the sand it twists and turns in
all sorts of worm-like contortions, and in Plate I., accompanying this
paper, some actual sketches are given of different individuals, showing the various contortions of the peduncle, as well as the character
of the sand tubes. Fig. 6 shows a portion of the peduncle of Lingula
pyramidata, drawn from life, showing its annulated character.  The
direction of the corpusculated fluid circulating through the central
cavity is indicated by the arrows.
Muscular System.
In respect to the muscles of the integument, the Brachiopods bear
the closest resemblance to the worms.
In worms, the muscles of the integument are arranged in two
layers, transverse and longitudinal, producing a reticulated appearance. The same arrangement is distinctly seen in the perivisceral
walls of Discina and Lingula, as well as in the peduncle of the
(334)




23
latter.  In the early stages of Discina, the reticulated appearance
produced by the two layers of muscles is particularly noticeable.
The presence of a ponderous dorsal and ventral plate, so peculiar
to many Brachiopods, accounts for the  extraordinary  muscular
apparatus to control their movements, as well as to move the body
upon the peduncle, in those forms that are attached.  This muscular
apparatus is unlike anything we find in the worms, though the powerful retractors of many sipunculoid worms, with their broad expanded
bases, recall similar features in certain  muscles of the Brachiopods.  The massive character of the muscles is more like the
Lamellibranchiates, save that in the latter the muscles are transverse,
and their only function seem to be to close the shell, their relaxation
allowing the elastic ligament within or without the shell to force or
pull the shells open, as the case may be. In the Brachiopods no such
ligaments are seen, the dorsal and ventral plates being opened, as
well as closed, by special muscles.  In one group of Brachiopods the
plates interlock at their posterior margins, and are restricted to
opening and closing in a vertical line. In Discina and Lingula the
plates do not interlock, and their posterior margins are free, the dorsal one lapping some way over the peduncle; it can therefore swing
freely back and forth, or oscillate from  side to side, as observed by
Carl Semper in L. anatina, and by myself in L. pyramidata, in its
acts of crawling, or burying itself in the sand.  (See Plate I., with
references.)
Perivisceral Cavity and Circulatory System.
In the higher worms, the perivisceral cavity is lined by a delicate
membrane noticed by Rathke, Quatrefages, Clapardce, and others.
Delle Chiaje designated it the tunica peritooeale.
In some worms having a rudimentary vascular system, according
to Claparbde, this membrane is ciliated, and prompts the circulation
of the perivisceral fluid.  This fluid in worms appears to be corpusculated and nutritive blood.  Most worms appear to possess an extensive vascular system which contains a colored, but not a corpusculated
fluid; this is the pseudo-halmal system of authors.
In Brachiopods, also, the perivisceral cavity is lined by a delicate
membrane, which in Terebratulina and Rhynchonella is strongly
ciliated as I have plainly observed in living individuals.
(335)




24
In  Lingula this membrane appears to  extend  into the pallial
sinuses, as is probably the case with other Brachiopods.   At all
events, the circulation in Lingula is induced by ciliary action, as can
be plainly seen through the transparent shell of Lingula pyramidata,
and this fluid is that of the perivisceral cavity, and is corpusculated.
Carl Semper, in his studies of Lingula anatina, says, that in that
species there is no heart proper, and that the blood is propelled
through the vessels by vibratile cilia.  As early as the year 1862, he
gave particulars, and has repeatedly insisted upon this anomalous
state of things.
To John B. Macdonald, however, belongs the credit of first calling
attention to these peculiarities in Lingula.
In the year 1861, Mr. Macdonald2 announced the discovery " of a
determinate circulation of spherical and violet-tinted corpuscles in all
the ramifications of the pallial sinuses, not dependent on the contractions of a pallial cavity, but upon the undulations of a ciliary lining."
The vascular system described by Hancock, with a vessel upon the
dorsal surface of the intestine, I have never succeeded in studying
satisfactorily.   In  Lingula pyramidata I have not been able to find
the vescicle upon the back of the intestine, but the vessel I have
clearly made out. In Discina I have made out the vescicle. This difficulty of finding a heart has been shared by others.  Carl Semper
could not find it, and Dr. Lancaster in the February number of the
Annals and Magazine of Natural History for 1873, p. 93, says in regard
to Terebratulca vitrea: "I entirely failed to convince myself that the
organ regarded  by Mr. Hancock as a heart really has the function
of one in  T. vitrea.  I repeatedly opened fresh  specimens with
rapidity, in order to witness its contractions, if any, but never saw
such contractions; nor could  I find vessels in connection with it,
nor evidence that it had muscular walls.  Dr. Krohn, of Bonn, had
equally been unable to obtain evidence that this curious little dilatation, has the function of a heart."3 From injected specimens of
Lingula, and from observations on living Terebratulina and Rhyncho1 Zeitschr. fiir Wissensch. Zool., xiv., p. 424.
2 On the Physiology of the Pallial Sinuses of the Brachiopoda. Trans. Linnean
Soc. xxIII.,p. 373, plate xxxv.
3 in a late study of living Terebratulina I observed a distinct circulation going
on in the sinuses of the pallial membrane, but whether these currents were induced
by ciliary action I failed to make out. The fact, however, that the delicate membranes in the perivisceral cavity are clothed with cilia, I clearly established.
(336)




25
nella, the various membranes, called by Huxley and Hancock the
gastro-parietal, ileo-parietal, and lateral parietal bands, are found to be
vascular, and the circulation taking place within these membranes
may be looked upon as representing the pseudo-hmmal system  of
authors.
These membranes intimately invest the oviducts, and in Rhynchonella the circulation of this system  can be seen following the spaces
between this membrane and the outer walls of the oviduct.
The red corpuscles in Lingula occur in the perivisceral fluid.
Other bodies of a fusiform shape, some elongated and others nearly
round, are also met in the perivisceral fluid. These are amoeboid in
their appearance, and may be seen bending and turning as they
course through the more delicate ramifications in the pallial membrane.
All these features I hope to figure in my memoir, now in preparation, on Lingula.  The colored corpuscles are similar to those which
occur in the perivisceral fluid of the Sipunculoid worms, as well as in
Glycera, and other worms, noticed by Claparede.
According to Lacaze-Duthiers, the two systems of circulation are
remarkably distinct in Bonellia and Sipunculoid worms, and he
queries whether the Brachiopods do not possess the same two systems.
Claparede says: " L'immense majorit6 des Ann6lides ne presente pas
de mouvement ciliare dans la cavit6 p6rivisc6rale, sauf k l'entree des
organes segmentaires.  Je ne connais, pour ma part, le vetement
ciliare perivisceral que dans les groupes suivants: chez tons les Aproditiens, chez tous les Glyceriens, chez tous les Polycirrides, chez les
Tomoptdridiens, et enfin chez une petite Terebelle assez anormale
(Terebella vestita).  Chose frappante, toutes ces Ann6lides, h 1' exception de la petite Terebelle et de l'Aplihodita aculeata, sont completement depourvues de vaisseaux.   Or, de ces deux exceptions,
l'une, l'Aprodite, est un animal a systeme vasculaire dans tous les
cas rudimentaire, appartenant a une famille d' ailleurs toute anangienne, l'autre, la Terebelle, appartient a une famille en g6neral vasculaire, mais dont une tribu cependant, celle des Polycirrides, est
anangienne.  Je dois, en pr6sence de ces faits, regarder le mouvement ciliare p6rivisceral comme une fonction vicariante de la circulation, chez les Anndlides depourvues de systeme circulatoire proprement dit."l
i Mm. de la Soc. de Phy. et Hist. Nat. de Geneve. Tome xix., 2d part, p. 329.
(337)




26
In the Brachiopods, therefore, particularly in Lingula, where the
vascular system  appears to be quite rudimentary, the presence of a
ciliated perivisceral coat is to be expected.
It must be confessed that much work has yet to be done in clearing
up the obscurity which still exists in regard to the circulatory system
of the Brachiopods.  What little is known about it, however, points
to vermian affinities.
Digestive System.
In worms, the digestive tract usually takes a direct antero-posterior
course without convolutions.  There are, however, marked exceptions
to this rule.
In the Sipunculoid worms, the intestine is not only convoluted, but
in many of them the anus terminates in front.  In a curious worm,
described and figured by Philippi,l under the name of ILHmenteria,
an anterior vent is described.
In those remarkable worms, Phoronis and Crepina, whose external
outlines in every particular so closely resemble the Hippocrepian
Polyzoa,2 the anus terminates in close proximity to the mouth.
In the Acanthocephali, the digestive tract is said to open into the
general cavity of the body in some, while in others it ends in a cecal
sac.
In  Temnocephala,3  a  Trematode  worm, the ccecal processes
from the stomach are much like those in young Brachiopods.
Among the Rotifera, in some groups the female has the cesophagus
terminating in a coeal stomach.  The anus, when present, terminates
anterior to the caudal portion.  The Turbellarians are also devoid
of an anus.
The anomalous features here presented by some worms, in the
absence of an anus, or the possession of a coecal stomach, and the
anterior termination of the anus, are fully repeated in the Brachiopoda. In one entire division of the Brachiopoda, represented by
Terebratula, the stomach terminates in a coecal sac.  In Terebratulina the alimentary tract is closed posteriorly.  Nor has the slightest
1Acad. delle Sci. di Torino, series ii, tom. x.
2 In Phoronis the oviducts with bilateral openings also terminate in front. Its
bilobed lophophore, reddish circulating fluid and embryonic stages all resemble the
Brachiopoda. In fact, Phoronis represents an important connecting link between
the Polyzoa and the worms.
3 Zeitschrift fur Wiss. Zool., Vol. xx, p. 307.
(338)




27
trace of an anus has been detected in Thecidium, Waldheimia, Rhynchonella, and several other genera that have been examined. In the
very early stages of Terebratulina, I have seen the rejectamenta escape from the mouth, and in no case has the appearance of an anal
perforation been discovered. In Terebratulina, the alimentary tract
pursues a direct antero-posterior course without convolutions, while in
Linguta and Discina the anus terminates anteriorly on the right side.
In Lingula, the intestine makes a few turns, while in Discina it makes
a single turn to the right.
In many worms, diverticular channels often spring from the lateral
walls of the intestine. In certain worms the liver appears as protrusions of the alimentary canal.
In Brachiopods the liver is composed of masses of ccecal ramifications which in youno Terebratulina, Rhynchonella and Discina, commence as simple civerticular channels of the stomach.
In Lingula pyramidata, an examination of these cceca to their
extremities revealed the presence of diatoms and other food, showing
that the contents of the stomach enters these diverticular processes,
and that the process of digestion is carried on in these parts, as in
the lower worms.  The fact that in young Rhynchonella, a distinct
peristaltic action is seen going on in the l;epatic cceca strengthens
this supposition.  While the brown hepatic lines are arranged parallel to the longer axis in the coeca in Terebratulina, in Rhynchonella
they are curiously arranged in a spiral manner.
In the higher worms, the intestine is freely suspended in the perivisceral cavity, and held there by delicate membranes which spring
from the parietes of the body.  (See Fig. 1.)
In Brachiopods, the intestine is likewise suspended freely in the
perivisceral cavity by delicate membranes which. spring from  the
parietes of the body.  These membranes were called by Huxley, the
gastro-parietal, ileo-parietal, and lateral parietal bands. (See Fig. 7.)
Cephalic Region.
In many of the tubicolous Chcetopods, as well as in Phoronis, the
head is furnished with a tuft of ciliated cirri.  Sometimes these
appear to surround the mouth in a single circle, as in the marine
Polyzoa, and in certain fluviatile forms. In others, they spring from
arms, spreading like two fans in some, while in others, each arm
is developed into a closely wound spiral of several turns. These
spring from what has been called a cartilaginous base. In Protula
media Stimpson, each arm  makes a single graceful turn.  The cirri
(339)




28
springing from these arms, are often highly and beautifully colored,
sometimes each cirrus is banded with brown.
In those with a closely wound spiral arm, as in Amp7hitrite ventrilabrum, the outer margin of the arm  at the base of the cirri, is
bordered by a delicate membranous frill, possibly corresponding with
the calyx in Polyzoa, but precisely identical with the brachial fold in
the arms of the Brachiopoda, as will be seen by reference to the sections to be presently given.
In the Brachiopods, the two arms springing from  the head are to
be directly compared to similar parts just described in the worms.
They  also  spring from  a cartilaginous base, and sustain ciliated
cirri, and in Lingula, Discina and Rhynchonella, they are developed
in a closely wound spiral, as in Amphitrite.  Furthermore in Rhynchonella, they can be unwound and protruded from the pallial cavity
as I had the good fortune to observe in living Rhynchonella from the
St. Lawrence.' (By instantly dashing the strongest alcohol upon the
specimen, I was enabled to preserve it with the arms extended.)
In Lingula the arms can be partially unwound, and what is very
significant, the cirri in Lingula pyramidata are banded with light
brown, as in certain species of Sabella.
The cirri in Brachiopods
Fig. 13.            Fig. 14.        show  a  rapid  circulation
through  their  transparent
walls.  They are employed
\\c( /\  \ f g       to secure particles of food,
ci \     which  they convey to the
mouth, and in every respect
they  are  strictly  identical
with  similar organs in the
Annelids. A transverse secIr    tion of the right arm of AmE[^-l     phiitrite ventrilcbrulm, Fig. 13,
s                  s              and of the right arm of LinTransverse section    Transverse section  
of arm of Amphii-   of arm of Lingulca    ula pyramidata, Fig. 14, is
trite ventrilabrum.   pyramnidata.       here presented. These secci. cirri. bf. brachial fold. s. sinus.    tions are, in each case, taken
midway between the base and
extremity of the arm. They much more closely resemble each other,
than corresponding sections of two Brachiopods resemble each other.
1 Am. Journ. Science and Arts, Vol. iv, Oct. 1S72. Otto Frederic M3iller, according to Von Buch, also saw Rhynchonella gracefully uncoil its arms.
(340)




29
In Lingula the brachial sinus is quite large, while in Amphitrite,
it is smaller.  The brachial fold in the former is wider.  The differences between the two, however, are differences of porportion of parts
simply. In certain Polyzoa, just as in that curious worm, Phoronis,
the cirri spring from a horse-shoe-shaped lophophore, hence they are
called hippocrepian Polyzoa.  In the early stages of Terebratulina,
T have shown that at first the lophophore is circular, as in the lower
Polyzoa, and afterwards the hippocrepian character reveals itself.1
It was important to learn whether the mouth opened between the
Fig.15.           a            ci
st       ne  cb                n a   bf
Longitudinal section of anterior portion of Lingula.
Fig. 16.
c                a      ci
ait
st   ~       C   cb       ma  bf
Longitudinal section of anterior portion of Amphitrite ventrilabrum.
tn. mouth. oc. oesophagus. st. stomach. a. arm. ci. cirri, bf. brachial fold.
cb. cartilaginous base of arm. s. sinus leading to arm. c. c. cephalic collar or
pallial membrane.
1 Mlorse. Early Stages of Terebratulina. Mem. B. S. N. 1I., Vol. 1. These observations have since been confirmed by Dr. E. Ray Lankester in Annals Mag. Nat.
Ilist., February, 1873, on the young of Terebratula vitrea.
(341)




30
outer and inner folds of the arms, in those worms possessing these
appendages, as in the Polyzoa, Phoronis, and the Brachiopods.  I
sought in vain the works of Claparede, Quatrefages and others, for
a figure of a longitudinal section through the mouth of some worm of
this character. Mr. Alex. Agassiz kindly gave me a large specimen
of the Amp]hitrite ventrilabrum,n  from the Bay of Naples, and I made
a careful longitudinal section through the head (Fig. 16). The right
arm was developed in a closely wound spiral of several turns, and
greatly exceeded the length of the left one. I have shlown the latter,
however, as it presents a clearer view of the twist and character of
the arm, as well as the crenulated membrane, or brachial fold which
borders the base of the cirri. In this section, as well as in the
accompanying one of Lingula pyranmidata, the cirri are partially
removed, as they form a confused tangle, and thus obscure the parts
that are desired to be shown.
The mouth is bordered by two membranous lips, which in the
Brachiopods are highly sensitive and movable in all those thus far
examined.  The mouth in Amphitrite, as well as in other worms of
that nature, is placed in the same relation with the head, and points
downward, as in the Brachiopoda.  The almost membranous shells
have been removed, and the dorsal pallial membrane is turned back,
to correspond in position with the dorsal cephalic collar in Amphitrite, which is normally
Fig. 17.            Fig. 18.      turned back, as in Sabella and Serpula.  In
other respects the sections are correct draw-.S^  f^\l            <>L>,> A ^ings of their leading
K ^\   \ 7     features, and are in no
way modified to resemHead of Discina.    Head of Sabella.       e       r     
ble each other. I have
lettered the parts alike, and their almost absolute identity may be
readily seen without further comments.  The relations insisted upon
as existing between the cephalic collar of the Annelids possessing
it, and the pallial membranes of the Brachiopods, are well shown in
these sections.  Even the relative position of the base of the dorsal
cephalic collar in the worm, corresponds to the same parts in Lingula
and Discina, the dorsal one springing from the head in advance of
the ventral one. See also Fig. 8, p. 14.
Figures are here given of a young Discina, Fig. 17, and of a
(342)




31
Sabella, Fig. 18, to show the similarity of the cephalic arms in the
two.  In one the portions are concentrated, while in the other they
are drawn out.
Dr. E. Ray Lankester,' in an exceedingly interesting sketch of Terebratula
vitrea, compares its arms and cirri, to the gills of a Lamellibranchiate. In
this comparison he is certainly wrong, for the gills of the Lamellibranch develop upon the sides, and in Sphnerium and other genera, towards the posterior
aed; while the cirri, with the arms sustaining them in the Brachiopods, are
strictly cephalic. Had he compared the arms of the Brachiopod with the membranous palpi of the Lamellibranch, he would have come nearer the true affinities, for these form folds above and below the mouth, are united partially on
their inner margins as I have observed in Unio and other genera, and in some
species are very long.
Renal Organs.
Claparede has noticed that in the segmental organs of many worms
a portion of the tube is glandular, and he has reason to suppose that
the glandular portion represents the renal organ.  Whether Claparede is right in his conjectures or not, it is interesting to recall the
fact that years previous to this statement, Huxley suggested that the
glandular portion so very conspicuous in the oviducts of Brachiopods
was of a renal nature.2
The slightest examination of the oviducts of Brachiopoda shows
the tubular portion not only glandular, but colored, as in the oviducts
of worms.  This portion is also intimately connected with the vascular system, and whether renal in its nature or not, the closest similarity exists between these portions and similar parts in the segmental
organs of Annelids.
Nervous System.
The general plan of the nerve system in the Vermes appears to be
that of a nerve collai surrounding the oesophagus, sending off a ventral cord in a median line of the body.  In the lowest worms there
seems to be simple cephalic ganglia without the ventral cord.  In
Sipunculoid worms a single cord running along the ventral portion of
the body sends off delicate threads at right angles to it.  In the
higher worms the ventral cord is longitudinally divided into two
symmetrical halves. In some worms there appears only a slight space
1 Annals and Mag. of Nat. Hist. Vol. xi, Fourth Series, No. 62, p. 93.
2 Hancock, ibid., p. 822.
(343)




32
separating the two halves; in others the ventral cord is distinctly
separated, but united at each segment by transverse threads, as in
Sabella, or its two halves may be united by ganglionic enlargements,
which sometimes correspond to the number of segments.  In the
Hirudinae the ventral ganglia are fewer in number than the segments. In Aphrodite and Polynoe, according to Grube, there are
more ventral ganglia than segments.
It was at one time believed to be typical of the Articulata that
each segment was characterized by a ganglionic enlargement of
the ventral cord, and it was supposed to be particularly so with the
higher Annelids.  Claparede calls attention to the fact that this is
not so in all cases, and especially in relation to the cephalic and thoracic
ganglia.
In the Nemerteans the ventral cord is widely separated, and runs
along each side of the body, and without ganglionic enlargements.
There appears to be, therefore, various conditions of the nervous system, in which there is in some a simple cosophageal collar; in others,
a ventral nerve cord, which may be single, or divided into two lateral
halves, sometimes widely separated, sometimes nearly approximating,
with, or without, ganglionic enlargements or threads connecting them.
In some worms accessory pedal ganglia are found.
In the Brachiopoda we have two lateral ventral cords, widely
separated, and connected at the oesophagus by ganglionic enlargements, which send off threads to the pallial membranes, and to the
various muscles.
In Lingula these lateral threads seem  to be double, connected by
commissures. In Discina, whose nervous system I have more especally studied, the nerve cords are bilaterally symmetrical, and widely
divaricating.  There are no ganglionic enlargements during their
course to the posterior end of the body, but in their track sending
off delicate threads, which in Lingula blend with their muscular
fibres, or pass round the muscles blending with their exterior fibres.
In Discina these lateral nerve cords terminate by ganglionic enlargements in the last two posterior muscles. These nerve cords were
correctly interpreted by Cuvier and Owen though mistaken for arteries by Hancock.   This error was corrected by Dr. Gratiolet, in his
study of Lingula hians;   and while studying Discina, before becoming aware of Gratiolet's researches, I found  these supposed
i Jour. de Conchyliogie. 2 Serle. Vol. Iv, p. 162.
(344)




33
arteries to be nerve cords, and traced them  to their posterior ganglionic enlargements.
Respiratory Apparatus.
The pallial membrane in the Brachiopoda sustains the principal
respiratory apparatus. It is in this membrane that the larger vessels
occur, and as I shall show in my memoir on Lingula, the pallial
menmbrane is divided into oblique transverse sinuses, which run parallel to each other.  From these arise numerous flattened ampullae,
which are highly contractile.
The circulating fluid courses in regular order up and down these
sinuses, entering each of the ampulle in turn.
Vogt and Owen were quite right in their determinations of these
organs.  They represent the branchiae of Lingula; but from the contractile nature of the ampulle, and their almost certain contraction
in alcohol, they have escaped the notice of others who have studied
specimens in which the ampullae were inconspicuous.
Thus Hancock was unable to find them  in the two species of Lingula, studied by him, and says, " The bladder-shaped enlargements
of the lateral pallial sinuses, alluded to by Dr. Vogt, are nothing more
than swellings occasioned by the contraction of the pallial, or marginal fold, which, pressing upon the extremities of the sinuses, throw
their walls into wrinkles, and hence their peculiar appearance." 1
These ampullae are very conspicuous in living Lingulae.  On one
side of the dorsal pallial membrane in an ordinary specimen there are
twelve sinuses, having in the aggregate eighty-five ampulle, numbering from five to eleven in each sinus. In life they form very interesting objects.  They project, or hang from  the walls of the pallial
membrane, like teats.  Their walls are perfectly transparent, and
the circulating fluid can be seen rapidly coursing into, and out of each
one in turn.
The following figure (Fig. 19) shows            Fig. 19.
a row of five amlpulle drawn from life,
within which the blood corpuscles can  
he seen circulating.                            -.-..\
Claparede says that in the normal 
branehia of an Annelid there cannot
be any mixture of arterial and venous blood.  The artery travelling a; far as the end of the branchia, where it returns as a
Hancock, Trans. Royal Soc. o. V   cLVIII, p. 852.
3




34
vein.  He says, "Veine et artere sont exactement paralleles l'une h
l'autre.  Dans toute la longueur de la branchie, ces deux vaisseaux
sont mis en communicateur par une double serie d'anses vasculaires
qui passent dans la couche sous-cuticulaire et qui subissent avec la
plus grande facilit  l'action de l'eau chargee d'oxygcne h travers la
cuticule tres-amincie.' 1
He denies, however, the independent contraction of the ampulla,
but says there is a rhythmical contraction of the whole branchia,
Quatrefages, to the contrary, notwithstanding.
In the family Serpulae, Claparede finds features remotely resembling
the description of Quatrefages, where in these Annelids "l'artere se
continue directement dans la veine h la base des branchies, et de leur
point de r6union part un vaisseau unique qui pelntre dans chaque
rameau branchie."  In Discina I have not been able to discover the
slightest trace of these ampulls, though the pallial sinuses are very
prominent; and the central partial partition of ciliated epithelium
which induces the flow of the circulating fluid in these parts, are as
distinctly marked in Discina as in Lingula.  In the other Brachiopods the prominence of the pallial sinuses, with their diaphanous
walls, must be regarded as a respiratory organ.  In all the Brachiopoda the cirri of the arms must also share with the pallial membranes
in this function.
Genital Organs.
Under this head we study the ovaries, oviducts, or segmental
organs, and spermaries.  In the Annelida, according to  Clapardce,
the sexual elements, in course of growth, form  ruffs all around the
vascular axes.
In all cases the ova, when arrived at maturity, detach themselves
from the ovary to float freely in the perivisceral cavity, where they
are afterwards gathered up by the ciliated mouths of the segmental
organs, and discharged by them.
This is precisely the case in the Brachiopoda. In Discina the borders of the delicate vascular membranes are thrown into conspicuous
ruffs by the development of the ova.  In Lingula the ovaries are
intimately bound to the same membranes.  In Terebratulina and
Rhynchonella they not only gather about the large vascular sinuses in
the pallial membranes, but hang in clusters from  the genital bands
1 Soc. de Phy. et l'Hist. Nat. de Geneve. Tome xIx, 2d part, p. 331.




35
in the perivisceral cavity.  In all these cases the eggs are discharged
freely into this cavity, and there float in the perivisceral fluid until
they are discharged from the body.
If we now  consider the ducts by which  these products in the
Brachiopoda find egress from the body, we shall find a startling identity of structure with similar parts in the Annelida.
In all worms, with few exceptions, these ducts assume the shape of
tubes, bilaterally disposed, suspended in the perivisceral cavity by
delicate membranes, and communicating with this cavity by flaring
orifices, which are strongly ciliated, as well as the tubes themselves, to their external orifices; the ciliary action always directing
the currents out of the body.  These are the segmental organs, or
oviducts.
According to Claparede, the segmental organs in the Annelida
present only very simple modifications of a very constant type.  In a
large number of worms, these segmental organs are repeated many
times, a few of them only modified as oviducts.  In other worms they
are reduced in number; in Branchiobdella, according to Dorner, to
two pair; in Terebella parvula to three pair, as Dr. Williams states.
In Protula,' Spirorbis, Sabella, and allied forms, a single pair of segmental organs, modified as tubiparous glands are found in the anterior
part of the thorax.
In Phoronis the ovarian openings are reduced to a single pair, and
these open at the extreme anterior surface of the body between the
arms, from which the eggs escape after having been discharged from
the ovary into the perivisceral cavity.  In the Brachiopoda the ducts,
by which the generative products find egress from  the body must be
described in precisely the same terms as those used in describing the
segmental organs of the Annelids.  The ducts assuming the shape of
tubes, bilaterally disposed, suspended freely in the perivisceral cavity
by delicate membranes, and communicating with this cavity by flarI The tubiparous glands, according to Claparede, represent modified segmental
organs. In Protttla (Salmaci7a) cedificatrix, Claparade represents the tubiparous
glands as opening by a common pore, as in Spirorbis. A careful study made by me
of a species of Protula (probably P. Dysteri Huxley) at Eastport, Mle., showed the
wide separation of these glands, and the fact that they opened by two distinct
pores, thus bringing them nearer the Sabellarians.
The minute structure of these glands, revealed a sinuous line rapidly undulating,
following the inner outline of the gland. This appearance appeared due to ciliary
action. Other features were presented by these curious organs, which led me to
believe that other functions were performed by them beside that of secreting the
tube. Their relations to the segmental organs were unquestionable, however.
(347)




36
ing orifices, which are strongly ciliated, as well as the tubes themselves, even to their external orifices, the ciliary action always directing the currents out of the body.
The concentrated character of the Brachiopods, and the limited
perivisceral cavity, reduce the segmental organs, or oviducts, to the
lowest number, and consequently we find in most of them but a single pair, as in Lingula, Discina, and Terebratulina, while in Rhynchonella two pair of oviducts are present.  It is significant to note
that in the last named genus both pair of oviducts have their inner
mouths turned toward the back.
In a special memoir on the oviducts of Brachiopoda, now in preparation, I shall demonstrate the unquestionable character of these
oroans.
Having studied them in living Lingula, Rlhynchonella and Terebratulina, and in alcoholic Discina, I find them  presenting only simple
modifications of a constant type.'
In Terebratulina the eggs were watched through the transparent
anterior walls, after their separation from the pallial sinuses.  While
floating in the perivisceral cavity, they were seen gathered up by
the flaring ciliated miouths of the oviducts, and were followed, as
they slowly passed through the tubes, and caught as they escaped
from the external orifices.
In Lingula, Discina and Rhynclonella, the external orifices of the
oviducts are simple slits, while in Terebratulina they project from the
anterior parietal walls, like tubercles, as figured by Claparede in the
Annelid Lepidonotus (Hermadion) fragile.  The glandular nature of
the oviducts, and their striking resemblance in this respect, to similar
parts in the worms has been alluded to under Renal Organs.
The following figures of the oviducts of Brachiopods, from  my
own studies, and the oviducts of certain worms, as figured by Claparede and Lancaster, are here given for comparison.
Having considered that portion of the genital system  referring to
the ovaries and oviducts, and shown their entire vermian character,
we come to study the male organs of generation, and in this line of
investigation we have to push into an almost untrodden field.
The Brachiopoda have been regarded by some authors as dioecious,
the vascular sinuses presenting ovaries or testes, according to the sex
1 Morse on the Embryology and Oviducts of Terebratulina. Am. Jour. Sci. and
Arts. Vol. iv, p. 262.
(348)




37
Fig. 20.                        Fig. 21.                       Fig, 22
Fig. 23.                               Fig. 24.
g,se
/    /
O0
Segmental organs of Worms
Fig. 20. Lumbricus. Fig. 21. Pectinaria.  Fig. 22. Eunice.   Fig. 23. Stylodrilus.  Fig. 24. Nereis.  se. segmental organ.  g. genitalia.  v. vascular channel.
i. intestine. Fig. 20 is from Lankester, the rest from Claparede.
Fig. 25.             Fig. 26.          Fig. 27.               Fig. 28.
Segmental organs of  irachiopods.
Fig. 25. Discina. Fig. 26. Lingula.  Fig. 27. Rhynchonella.  Fig. 28. Terebratulina. These firures are from my own studies.
(349)




38
of the individual. Hancock 1 however was inclined to believe the
sexes united in all the forms coming under his observation.  Oscar
Schmidt2 also says that in Terebratula the testicles and egg stocks
are united in the same individual.
After patient study of these parts, I believe that in all the Brachiop-oda the sexes will be found to be separate.  In Lingula the spermaries occur in the perivisceral cavity, in masses like the ovaries.
Having studied them  alive it was found that while in some individuals the ovarian masses nearly filled the perivisceral cavity, in others
spermaries occupied similar positions.
As Discina presents precisely the same characters in the ovaries
springing from  the vascular membranes, and filling the perivisceral
cavity, it is reasonable to suppose that the spermaries correspond in
position with those of Lingula.
A careful study of Terebratulina, lately made during its breeding
season, shows also that in this form  the sexes are distinct.  While
some specimens revealed the vascular sinuses filled with eggs, and
even where the eggs had escaped by dehiscence the scars could be
seen, in others the sinuses showed no traces of eggs, but on the contrary were filled with a creamy mass, slightly granulated, the borders
of these masses being highly ciliated, and when crushed or separated
under the compressor, bunches of spermotozoa and single ones were
revealed.  This probably represents the ovigerous mass of Hancock.
In several females examined, the eggs were attached in clusters to
the genital band, even to the very edge of the mouths of the segmental organs.  And in several males the spermaries were likewise
attached in clusters to the genital band, and in such masses and so
close to the segmental organ that the accessory vescicle of Huxley
was obscured by them.
The masses of spermaries adhering to the genital band, and floating free in the pervisceral cavity, presented some curious features.
They assumed the shape of long filiform  appendages, attached by
common centres to the genital band, and surrounded by an almost
imperceptible cellular mass. The threads widened gradually to their
distal extremities where they ended bluntly, and were capped with a
few large brownish cells.  The spermatozoa were thickly clustered in
blunt fusiform masses at the extremities of the threads, forming a
sort of brush.  The same brownish granules appeared in the sinuses,
1 Hancock, ibid., p. 824.
2 Zeitschrift fiir ges.Naturwissenschaften, 1854, p. 325.
(350)




39
and likewise tipped the clusters therein contained, only these clusters
were not supported on  long  threads, as in those which  sprang from
the genital band in the perivisceral cavity.  The glandular portion
of the segmental organ in  the male appeared. much darker than  in
the female.  As Rhynchonella presents similar features in the ovaries
contained in the pallial sinuses, we believe that the spermaries will be
found in like positions.
In this connection we must also consider the accessory vescicles of Huxley
(accessory hearts of Hancock). After careful study of these minute organs
fiom a large number of living specimens, I am convinced that they do not bear
Hancock's interpretation, and that they properly belong to the genital system
and not to the circulatory system  as stated by him.  Hancock describes the
walls of the "accessory hearts" as more delicate than the walls of what he
regards as the central dorsal heart.  This is certainly not so in regard to Terebratulina.  In T. septentrionalis, the organ presents all the appearance of a
gland.  The walls are thick and glandular, in fact, no sure evidence of a cavity
within has yet been met with.  It is irregularly pyriform in shape, slightly flattened, and in sorne is attached by a very constricted neck to the genital band
just beneath the flaring margin of the mouth of the segmental organ.
The exterior wall is made up of prominent transparent cells; at the base of
the gland, and also on its walls, masses of yellowish granules in patches appear.
On the genital band also I have seen irregular masses of cells, presenting all the
appearances of the accessory vescicle.  Repeated observations failed to detect
any vascular communication with the band to which it is attached, not the
slightest trace of circulation within its walls has been observed, nor the slightest
evidences of dilatation or contraction, nor evidence of muscular fibre.  I have
repeatedly crushed it beneath the compressor, yet no signs of forcing out contents has been observed. Moreover the organ differed in appearance in different
specimens, and even differed in appearance in the same individual upon the
right and left sides of the body. With the idea at first that the sexes were
united in Terebratulina, I was inclined to regard them as the testes, since they
always occur in the immediate vicinity of the segmental organs.
Where a single pair of segmental organs occur, as in Terebratulina, two
accessory vescicles occur, one to each segmental organ.  Where two pair occur
as in Rhynchonella, four accessory vescicles are found likewise; one accompanying each segmental organ. In Lingula, and Discina, though the segmental
organs are large and conspicuous, and their study rendered comparatively easy,
yet in no case has the accessory vescicle been met with.  So far as we know
then, the accessory vescicle occurs in Rhynchonellidas and Terebratulidse, or in
those groups with the dorsal and ventral plates interlocking, which have no
anus, and in which the ovaries are contained in the vascular sinuses of the
pallial membrane.  The accessory vescicles do not occur in Lingulidas or Discinidce, or in those groups having the dorsal and ventral plates free, possessing
an anal outlet, and which have the ovaries entirely free in the perivisceral
cavity.  That they have nothing to do with the circulation is evident from the
(351)




40
fact that a portion of the pallial membrane separated from the animal, shows
the circulation going on as usual. The vascular bands are also strongly ciliated,
and in Terebratulina, as in Lingula, the perivisceral circulation is probably induced by ciliary action.
That the accessory vescicles, then, belong to the reproductive system, and not
to the circulatory system, there can be no doubt, but just what their function
may be, has yet to be discovered.
I copy from  that inexhaustible work of ClaparBde, the Annelids of the
Gulf of Naples, a figure of the segmental organ of Alciope Cantrainii, and
accompany it with a figure of the segmental organ of Terebratulina septentrionalis with its accessory vescicle attached in the same position.  There is at
least something suggestive in the relations of the two figures, though by this
suggestion I would not throw doubt on Clapar6de's determinations.  With the
impression that the sexes were united in Terebratulina, I was inclined to regard
Fig. 29. t                         Fig. 30.
m           t
i      0  
8       a
Segmental organ of Terebratulina.
s.  segmental  organ.  m. inner
mouth of ditto. o. external orifice of ditto. g. genital band. a.
accessory vescicle. i. intestine.
Segmental organ of Alciope Cantrainii. m. inner
mouth of segmental organ.
o. external orifice of ditto.
t. testes.
the external parietal glands discovered by me as representing the testes.  With
the identification of the spermaries as above described, and the consequent separation of the sexes, we must seek for another interpretation of the glands.
As they are  extremely mucous, and intimately surround the external tubular orifices of the segmental organs, the egg cannot possibly escape without first coming  in  contact with  whatever substance  these glands may
secrete, and it is highly probable that they are instrumental in investing
the egg with some external coat. The glands are very white in color, and are
filled with minute granules which I at first mistook for spermatozoa. It is interesting to observe that they are present in both sexes, but what service they do in
the male it is impossible to conjecture.
(352)




41
In connection with the external parietal glands, it is interesting to recall the
capsulo-genous glands as described by Dr. Lankester  in the common earth
worm. He says " Besides the regular glands developed on the parietes of the
body, the earthworm exhibits numerous glands destined to form the egg capsule,
in which both zoosperms from the spermatic reservoirs, and ova are deposited,
These glands were first detected by D'Udekem.... The white color.
and thick fleshy look which is sometimes observed about the exterior of these
segments, is due to the developement of the capsulo-genous glands. Whether
the capsulo-genous glands have everything or anything to do with the formation
of the egg capsule is very difficult to determine; but the supposition of M.
D'Udekem  is so plausible, and comes from so good an authority, that it cannot
but be received until absolutely disproved."
Embryology.
It seems a little remarkable that of the class of Brachiopods, upon
which so many admirable memoirs have been written, so little should
still be known about the embryology, or early stages of any of its
forms.  Yet all that has been done, thus far, to shed any light on
this portion of their history, points to the unquestionable vermian
characters of the class.
The earliest paper on which we find any reference to the embryonic form of the Brachiopod is by Oscar Schmidt, contained in the
" Zeitschrift fir ges. Naturwissenschaften," 1854 p. 325.  He gives
the following figure of the larval form of a species of Terebratula.
(Fig. 31.)
Fig. 31.      In this figure the body shows a deep constriction
in the centre, the lower end is abrupted truncate,
as if that were to be the point of attachment.  If
that is the case it would correspond to certain Rotifers which also attach themselves by the posterior
segment of the body.
Embryo of a    (See Fig. 5 of Rotifer, on  page  11, 2iielicerta
Brachiopod.                  
ringens by Huxley, showing its first attachment by
the posterior end, the animal at the same time surrounding itself
with a sand case.)
Lacaze-Duthiers was the first naturalist to make known several
stages of the embryology of a Brachiopod.  In a memoir on Thecidium 2 this author gives several figures of the embryos of this
Brachiopod. The body is composed of four deeply constricted seg1 Journal Microscopical Science. Vol. V, 1865, p. 14.
2 Annales des Sciences Katurelles, 4th S6rie, Vol. xv.
(353)




42
ments, the anterior one small and running back on the second  segment.  On  one embryo he found  two red  eye-spots on  the first
segment, while on another embryo he found four red eye-spots.  In
the same number of rings, the peculiar form  of the cephalic ring,
and if Oscar Schmidt is right, the attachment of the embryo by the
caudal segment, we observe the closest similarity between the embryo
of this Brachiopod and that of Melicerta ringens, figured by Huxley.l
It will be seen by the accompanying figures of the embryo of Thecidium, copied from Lacaze-Duthiers, and the embryo  of lMelicerta
ringens, that in the Rotifer embryo the body is drawn out, while in
Thecidium  it is condensed.
For several years I have endeavoured to secure some data regarding the embryology of Terebratulina, and in the early summer of
1872, I had the good fortune to find Terebratulina spawning.2  The
eggs were round  and ciliated, and had the peculiar pencil of long
cilia so peculiar to the embryos of many worms.  The body was
gradually cut up into three deeply constricted segments, and these at
a later stage had the characteristic vermian contraction, the rings
shortening upon themselves and then gradually expanding.  Though
I made several hundred drawings from fifty different embryos, yet
1 Jour. lic. Soc., Vol. i.
2 Since this paper was in type, I have again visited Eastport, Maine, and have had
an opportunity of studying the embryology of Terebratulina under more favorable circumstances. I have only room here to state that after swimming actively
for a while the segmented embryo becomes attached by the caudal segment which
is to be the peduncle, the middle or thoracic segment increases in diameter, one
portion becoming more prominent; the first or cephalic segment continues to move
and bend on the thoracic segment. Meanwhile the thoracic segment grows rapidly at opposite points, and finally engulfs the first segment by lobes above and
below, these lobes being the dorsal and ventral valves. The mouth becomes apparent, and at the same time two groups of setae make their appearance on the sides
and front of the lobes, these are delicately barbed and decidtous.
The relations pointed out on pages 29 and 30 regarding the cephalic collar of the
Annelids and the pallial membranes of the Brachiopods receives confirmation in this
unlooked for simple developement of Terebratulina, and so far as these forms are
concerned, we can for the first time state positively that the mouth and arms represent the first segment, the pallial membranes with the shells, the second or
thoracic segment, while the peduncle represents the caudal segment.
The first three tubules in the shell are bordered by long delicate hairs, indicating
that these are sense organs as in the tubules of certain crustacea, surrounded also
by hairs as described by De Morgan, see p. 16. Regarding barbed setae, see page 21.
These results were communicated to the Boston Soc. Nat. Hist., June 18th, 1873,
and will soon be published in their Memoirs.
(354)




43
Fig. 32.. 33.        Fig. 34.        Fig. 35.            Fig. 36.
Embryos of Worms.
Fig. 32. Serpula. Fig. 33. Spio. Fig. 34. MIelicerta (Rotifer).  Fig. 35. Pileo-!aria.  Fig. 33. Phoronis.  (Fig. 32 original.  Figs. 33 and 35 from  Claparede.
Fig. 34 from Huxley. Fig. 36 from Dyster.)
Fig. 37.               Fig. 33.                  Fig. 39.
Fig. 40.              Fig. 41.                Fig. 42.
Embryos of BIrachiopods.
Figs. 37, 38, 39. Thecidium  (from Lacaze Duthiers).
Figs. 40, 41, 42. Terebratulina (original).
(353)




44
owing to the difficulty at that time of keeping the water at the
frigid temperature they were accustomed to, the embryos all died.
I saw enough, however, to lead me to believe that they eventually
become attached by their caudal segment.
Thus the three series of observations by Schmidt, Lacaze-Duthiers
and myself, on entirely different forms of Brachiopods, show the
most perfect vermian development.
In regard to the early stages of Brachiopoda, Fritz AMiller 1 has
published some interesting observations on the early stages of Discina,
which he studied at Santa Catharina, Brazil.  The larva is described
as having a perfectly orbicular dorsal and ventral plate, the pallial
membrane gaping all round, and the dorsal plate freely moving, or
sliding back and forth.  Five pair of very stiff setme project from
the periphery, two pair forward and three pair backward, one pair
much larger and stronger than the rest, and these were coarsely
barbed.  The posterior half is occupied by the alimentary canal, two
auditory vescicles, and two eyes. The anterior half is occupied by
four pair of cylindrical tentacles or cirri, strongly ciliated, between
which a rounded knob is situated.  (Possibly related to the rounded
knob in Spirorbis and allied forms.)
The arms or cirri are supported upon a long retractile neck or
esophagus, at the forward extremity of which the mouth is situated.
The larva not only swims by means of the cilia lining the cirri, but
crawls by means of the ventral scale, and pushing itself along by the
larger pair of bristles which have a vigorous motion, often crossing
behind.
These barbed bristles of Discina, Muller finds, are deciduous, and
it is interesting to remember in these comparisons that the larval
worm  has also coarsely barbed bristles which are likewise deciduous.
What could be more annelidan than the description of this larva. It
is true there are no larval worms possessing the dorsal and ventral
plates, though in that degraded worm Sternaspis there is a pair of
ventral plates or scuta, from the edges of which sete project, and by
means of which the worm shoves itself along. (The lines of growth
are prominent on these scuta, as in Discina.)
Since then I have received, through the kindness of Dr. Hagen, a
letter from  Herr Muller, accompanied with a sketch of another
larval form of Discina, in which he describes features similar to those
1 Reichert und Du Bois-Reymond's Archiv., 1860, p. 72. Wiegmann's Archiv.
1861, p. 53.
(356)




45
above mentioned, and  states that the  species has been defined by
Prof. Dunker as D. radiata.
The Brachiopods possess other affinities to Rotifera, beside the resemblances
between the embryos pointed out above.  We give here a definition of the
Rotifera condensed from  Rolleston's Forms of Animal Life, p. cxxxviii, Class
Rotifera, inserting in italics those portions in which some agreement may be
claimed.
"Vcn:  es with a retractile ciliated dislk at the anterior extremity of their bodies.
(To be compared to retractile cirri in young Discina.)  Usually plainly annulated externally, never divided internally by transverse septa.  (Except Lingula.)
In most Rotifera entire body divisible into a' body' proper, and a tail, anterior to
which the digestive and reproductive viscera swith their ducCts are situated. The
body can often be seen when chitinization has not advanced so far as to form a
carapace, not only to be distinctly annulated, but to possess circulcar andl longitldinatl muescles in its walls.  Cilia are never found on the external suzface of the
body, except q)onz the cephalic oryan.
The chitinozs srcface of the integyszent may develope setiformn outctrouths of
various shapes, or the animals naCy secrete or agglltinace a tube for the lodgenent
of their bodies.  The tail is usually annulated when its integument is soft, (quite
marked in Lingzulapyramidata) or segmented when it is indurated.
(The paired claw-like processes is unlike anything found in Brachiopods.
The anus is on the back and not on the side as in Brachiopods.  The differences between the sexes is unlike the Brachiopods, though the ccecal stomach
as in Ascomorpha, Notommata and Asplanchnia finds a parallel in the coecal
stomach  of many Brachiopods.  The jaws and gizzard are again different from
what has vet been observed in Brachiopods.)  Teco or nmore cccal app)lendages
are affixed to the comsmenelcement of the stomach, chich, as also the intestine, is
clothed wcith cilia.  (See Morse, " Early Stages of Terebratulina," and since observed in young Rhynchonella and Discina).  The Rotifera have no heart.  The
periviseral cavitq contains a corpTsculatedlfluid.  No specialized breathing organs.
(save the curious ampullk  on the pallial membrane of Lingula.) The water vascular system has five ciliated infundibuliform orifices, (to be compared with the
segmental organs in Worms and Brachiopods).  The reproduction in Rotifera
by means of Awinter eggs possibly finds a parallel in the statoblasts of Polyzoa."
Features of dissolution.
Believing the Brachiopods to be true worms, every feature of relation, no matter how  trivial, becomes important in these comparisons.
The mere fact of Lingula pyramidata agglutinating  a sand  tube is of
small moment in  itself, since such  a feature might of course be a
matter of secondary  acquisition.   The  unquestionable  fact  however, that many  groups of worms are notable tube builders, and  this
peculiarity  is almost entirely  confined  to the worms, the feature of
tube  building  in Lingula becomes a matter of importance in  these
(357)




46
comparisons.  Therefore the points to follow, trivial in themselves,
cannot, with justice, be overlooked in these comparisons.
Dr. Williams, in his elaborate work on the British Annelilda., in
describing the dissolution of Arenicola and Nais, says in regard to
the former, that the division occurs somewhere within the middle
third of the body, though sometimes the head is detached, and sometimes the tail.  This process, both in the Nais and Arenicola, occurs
in July or August. The cephalic and caudal portions continue for
some time to writhe in the sand.  Towards September the fragments
disappear by decomposition, the parts turning black.  He further
says that the sand of the sea-shore, and the water of the fresh water
pools are thickly strewn with the mutilated bodies of these worms.
" It is a catastrophe which every fall involves the whole community."
He believes therefore that these Annelids are annuals.  " They are
born during the latter months of one summer, and survive the winter, attain the maturity of growth, reproduce the species, and die by
the spontaneous subdivision of their body into fragments."
In studying Lingula p yranidcta, I had come to the conclusion that
with this Brachiopod at least, their duration of life did not probably
exceed one year.  Of over one hundred specimens of Lingula collected by myself in June, and as many more collected by Dr. Elliott
Coues in July, I had remarked as a noteworthy fact, that specimens
varied but little in size.  I did not meet with a single young specimen, or, rather, a single small specimen.  Furthermore the shells in
all cases presented the same features of newness.  There were no
erosive or parasitic growths upon them (though late in the fall there
is no reason why hydroid growth might not occur on that portion of
the shell exposed).  This fact, coupled with the absence of even a
small specimen, for which I particularly searched, led me long ago to
believe that they were all of the same age, and that their life did not
exceed one year.
I brought several specimens home with me to Salem, Mass., and
kept them alive during the summer months; imitating as far as possible the conditions in which I found them, keeping them  in the
same sand in which they were collected.  They all died within a
few days, during the last of September, and in their death they repeated almost precisely the features described by Dr. Williams in
Arenicola and Nais.  Spontaneous division occurred  between the
thorax and peduncle.  The thorax was thrown out on the surface of
1Report of the Brit. Asso. for Ad. of Sci. 1851, p. 248.
(358)




47
the sand, while the peduncle lay embedded in the sand.  The thorax
lived for a few days in a weak condition, the seta moving feebly, and
the dorsal plate slightly oscillating.  This portion finally turned
black, and as one after the other perished in this way, I removed
them from the vessel, in order that their decomposition might not vitiate the water, and thus imperil the lives of the survivors.
Several days after they had all been removed in this way, curiosity
led me to turn up the sand, to find the condition of the peduncles,
and judge of my astonishment at finding the perivisceral fluid contained in the peduncular cavity still rapidly circulating in several,
though the region at which the separation had occurred was blackened by decomposition.  In a few days more this circulation ceased,
and decomposition envolved all parts.  Here we have the most complete relation between the dissolution of this Brachiopod and that
of the Annelid described by Dr. Williams.
Opinions of Authors concerning the Relations of the Brachiopoda.
Naturalists are sufficiently well acquainted with the relations repeatedly pointed out, as existing between the Brachiopoda and Polyzoa,1 and there is no need of again repeating them here.
It is also a matter of history that the Polyzoa were placed with the
Mollusca solely on the relations which were supposed to exist between
them and the Tunicates, and afterwards the relations recognized between the Polyzoa and the Brachiopoda.
Again, there can be no question that at the outset the association
of the Tunicata with the Mollusca arose from  the relations supposed
to exist between the external sac or tunic, with the two apertures of
the one, and the shells and syphons of the other.  Aristotle2 dwells
on this resemblance where he says, in speaking of them as Mollusca,
"They are the only kind whose whole body is enclosed in the shell,
and that shell of a substance between true shell and leather.  They
are attached to the rocks by their shell.  They have two separate
1 In the affinities of the Polyzoa with the Vermes, the curious genus Phoronis
offers an important link. Claparede thinks the affinities of Phoronis are with the
Gephyreaqs on the one hand, and the Polyzoa on the other. Kowalevesky has
discovered that the larva of Phoronis is an Actinutrocha, and possibly the young
Sipunculoids that Schneider saw resulting from a transformation of Actinotrocha,
are early stages of Phoronis. Certainly no one studying the characters of Phoronis and Crepina can fail to see many very intimate relations between these forms
and the hippocrepian Polyzoa.
2 Fourth book of " History of Animals."
(359)




48
openings, which are very small and difficult to notice; the one to
take in, the other to eject the water," etc.
The branchial sac was also believed to be the homologue of the
gills of the Lamellibranchiate Mollusk, though Hancockl has shown
that the branchial sac of the Ascidian is not the anatomical equivalent of the gills of the Lamellibranch, but is a portion of the alimentary canal.
Milne Edwards, in his splendid memoir on the composite Ascidians,2  in  speaking  of their molluscan  affinities, said  that  these
relations were far less intimate than was usually believed, and that
they departed from the Mollusca in their mode of circulation, in the
itetamorphosis which their fry passed through, and more particularly
in the singular feature that most of them  possessed in multiplying by
gemmation.
Now since the Polyzoa are placed with the Vermes by Gegenbaur
and others, and indeed were long ago placed there by  Leuckart,
while the Tunicates have been assigned to the Vermles by many of
the most eminent German investigators, while others still would place
them  at the foot of the Vertebrate series, it is unnecessary for me
to consider the affinities of the Brachiopods through their relations
with these groups.
Kowalevsky. Kupffer, Schultze and others, assign the Tunicates a position
at the base of the Vertebrate series, through the unquestionable affinities of
certain of th    fr   t Amphi, a          their fors to Amphioxus, as  ell as theisingular embrological relations with the Vertebrates.3  In this connection it is instructive to note that
Gegenbaur sees a relation between the branchial sac of Balanoglossus, and
the branchial sac of an Ascidian.
In the year 1848, the far seeing Leuckart was inclined to believe that the
Tunicates formed an intermnediate class between the Echinoderczata and the
Vermies, while others traced a resemblance between certain Nesnertean larva
and the eualy stages of the Echinoderms, causing Huxley to unite the Echinodermata with the Articulates. And lastly, a Tornaria described by Miiller,
Krohn and Alex. Agassiz, was taken to be an unquestionable Echinoderm larva.
Now thanks to the brilliant investigations of Mr. Agassiz,4 this Tornaria turns
out to be the young of that odd worm Balanoglossus; though Mr. Agassiz finds
a wide gap between the Tornaria of Balanoglossus and the Echinoderm young,
vet he admits the striking resemblance between the two. He says, " This
remarkable type recalls the Tunicates, from the nature of its gills and mode of
I Annals Nat. Hist. 4th Series, Vol. v., p. 196.
2 Memoires de lAcad. des Sciences. T. xviir, 1842.
s Early Stages of an Ascidian. Proc. Bos. Soc. Nat. Hist., Vol. xiv., p. 351.
4 The History of Balaenoglossus and Tornaria. Iem. Amer. Acad., Vol. ix., p. 434.
(360)




49
formation.  It has, like Echinoderms, a ring canal; its larva is eminently
Echinodermoid, allied to Star-fish larve, which in their turn are more closely
allied to the larvm of Holothurians and Crinoids, than to those of Echinoids and
Ophiurians."
Aside from  the roundabout way in which the Brachiopods have
been entangled with the Mollusca (their preposterous comparisons
with Anomia will be considered presently), it is interesting to observe how often certain features of the Brachiopods have been compared to the lower Articulata by those who have made special
researches upon them.  Thus Dr. Gratiolet,l who studied Lingula
anatina, says the organization of the arms resembles the branchia of
certain Crustacea.  He also expresses the opinion that the Brachiopoda are allied to the Crustacea in respect to their vascular system,
and not to the Mollusca; least of all to the Tunicata.  Again, after
recounting the peculiar character of Lingula in the annulated hairs,
developed fiom veritable glands, the structure and arrangement of
their muscles, their arms and other features, Gratiolet says, the
Brachiopods are very far removed from the Lamellibranchs, and have
no kind of relation to the Tunicates.
Lacaze-Duthiers,2 in speaking of the oviducts of the Brachiopods,
recalls the fact that in Bonellia (of which he made an elaborate
study) there are similar openings, which are the genital openings,
through which the visceral fluid can escape. He also queries whether
there are not two kinds of circulation in the Brachiopods, as in
Bonellia.  On the development of the young, studied by himself in
Thecidium, he observes points entirely unlike anything existing in
the Lamellibranchiates.
Burmeister compared the gills of Lingula to the gills of Lepadm.
Gegenbaur, in his " Outlines of Comparative Anatomy," points to
certain worm-like features in the Brachiopoda, and repeatedly calls
attention to their worm-like genitals.
Dr. Williams,3 in his elaborate work on the British Annelids, calls
attention to the outlying affinities of the Vermes, recalling Dentalium,
Chiton, Amphioxus, but no where alluding to any approach of the Annelids to the Brachiopods.  None of the above authors, however, had
ever suggested the removal of the Brachiopoda from the Mollusca, nor
had they, or Owen, Vogt, Huxley or Hancock, ever made the slight1 Jour. de Conch. 2d Series. Tome ii, pp. 237, 252, 257.
2 Annales des Science Naturelles. 4th Series. Tome -xv.
3 Report of the Brit. Asso. for A. of S. 1851. p. 164.
4                                      (361)




50
est allusion to Prof. Steenstrup's views on the subject, and until quite
recently I had thought that to myself belonged the entire credit of
the views advanced in this paper, until I was made acquainted with
the fact that twenty-five years ago Steenstrup had not only considered
the Brachiopods as worms, but had placed them  near the tubicolous Annelids.  Before presenting the views of this distinguished
naturalist, it is proper to go back to the relations that many naturalists, as Agassiz, Deshayes, Owen and others, believed to exist
between Terebratula and Anomia.
In the light of our present knowledge of the subject, it seems as
incomprehensible that such views were held, as that the Cirrepeds
were ever included with the Mollusca, and it is still more a matter
of astonishment that to this day there are a few naturalists who have
a vague idea that Anomia forms a sort of connecting link between the
Brachiopods and the Lamellibranchiates.1
In the year 1853, Forbes2 wrote as follows: "Linnsus included in
his genus Anomia the species of Terebratula.  Mislead by a false
analogy, he considered these very different Mollusks to be organized
on the same plan, and the perforation of one of the valves in each, to
be of similar origin.
"Lamarck, in like manner, fancied that in Anomia he saw a passage
into Terebratula and the Brachiopods; and some anatomists even
believed that they had discovered transitional characters. A close
examination shows that there is no relationship of affinity between
them, but only a resemblance through formal analogy.
"The parts which seem, at first glance, in each to be identical, prove
not to be homologous upon investigation. Anomia has really very
close relations with Pecten, and is connected to the latter by the
curious genus Hemipecten of Reeve.  The perforations in one of the
valves of Anomia is chiefly a greater extension of the auricular sinus
in Pecten; and when the very young fry of this genus shall have been
carefully observed, we believe they will be found spinning a byssus,
which passing through this in the first instance, before a portion of
it becomes attached, eventually becomes detached with a part of the
adductor muscle, and forms the opercular process."
Lacaze-Duthiers,3 in an elaborate memoir on Anomia, refers to
1 At one time Vogt and Agassiz believed the dorsal and ventral plates of the
Brachiopoda to be right and left, like the bivalved shell of the Lamellibranch 
2 British Mollusca. Vol. ii., p. 322.
3 Annales des Sciences Naturelles, 1854. 2d Series, v., p. 35.
(362)




51
these statements of Forbes and Hanley, and expresses his belief in
their correctness. In speaking of their peculiar asymmetry induced
by this byssal modification, he aptly calls them  the Pleuronectes
among Mollusks.
In the year 1847, Steenstrup presented similar views before the
Royal Danish Academy.' And in the same year, during the meeting
of the Scandinavian Naturalists in Copenhagen, he made a similar
communication, which was published in their Report for that year.
On both occasions he militated against the suggestions of Owen,
Agassiz, Deshaves, and other naturalists, who considered Anomia as
a connecting link between the Lamellibranchiates and the Brachiopods.  He showed that Anomia was not so abnormal as was generally
supposed, and that the foramen in Anomia had no correspondence to
the opening of the valve in Terebratula, but, on the contrary, was
homologous to the notch in Pedum, Pecten, and certain other bivalve
Mollusks, and that the plug in Anomia was simply a calcified byssus,
and that it passed through this notch and held the shell fixed to some
object. He showed also that the muscle attached to this plug was a
foot muscle, corresponding to the muscle which goes to the sheath of
the byssus in certain other Lamellibranchs; and concluding with the
statement that the Terebratula and all the Brachiopods might necessarily be considered as not only widely removed from the Lamellibranchiates, but as having no sort of relation to the Mollusca at all.
Last year, in a brief examination of the early stages of Anomia, I
had the pleasure of amply confirming the predictions of Forbes and
Steenstrup, namely, that the plug in Anomia represented simply a
modified byssus.
The following extract is taken from  my short paper on the Relations of Anomia.2  "The smallest specimens examined are quite
orbicular, the upper or left valve is very tumid near the nucleus, the
lower, or right valve is flat, and somewhat smaller than the upper
valve.  The foramen, or sinus, is not closed, but opens on the anterior
border of the shell. The chief point of interest, however, is seen in
the nucleus, or that portion of the shell first formed, when the animal was free and roving. This early condition of the shell is distinctly marked at the beak in both valves. It is yellowish in color,
and marked with numerous, very regular concentric lines of growth,
while the remaining portion of the shell is colorless, or white, with
1 Seethe Proceedings for that year, pp. 74, 75.
2 Proc. Boston Soc. Nat. Hist. Vol. xiv., p. 152.
(36a)




52
irregular lines of increment. The nucleus is oblong oval. The urnbones are nearly central, though nearer the anterior margin, and the
shell is more globose behind.  Both valves of the nucleus appear
equally convex, and no sign of a
Fig. 1.      Fig. 2.      sinus or perforation is visible in
either valve.  On the free edge of
the right valve, directly under the
umbo, a distinct notch is seen, the
lines of growth indicating  it, and
showing that the edge of the shell
is not absorbed to form this notch.
l~,((?f{~_'\ " —~ "~'[  It will be noticed that this marFig. 3.                             ginal notch appears in that valve
Fig.   _ 3.vwhich is below, and which afterwards presents the opening for the
passage of the byssal plug.
"The condition of the shell at
this time clearly indicates that the
/f E   li t   animal is not only already attached,
but has fallen to one side, and while
Fig. 4.   \\                        in this position has added a few
more lines of increment to its larval
shell, as no sign of this notch is seen
on the left or free valve.  Soon,
however, the peculiar and rapid secretion of a different shell growth
takes place; the lines of increment
|   a     are no longer regular, nor so conFig. 5. I                            Fig. 1. Right or lower valve of Anomia, showing notch in ventral pallial
margin, caused by byssus. Diameter one
sixty-fourth inch.
  _/^~~        ^Fig. 2. Left, or upper valve of Fig. 1.
Fig. 3. Foramen commencing to form.
The black portion shows proportions of
left valve.
Fi //ff{\~  \\\ 1Fig. 4. A more advanced stage of right
Fig. 6.       V      _A             valve with foramen almost completed.
Fig. 5. Left, or upper valve of Fig. 4.
Diam. one thirty-second inch.
Fig. 6. Showing still later stage, with
foramen completed, and nucleus still visible.
(364)




53
spicuous, and the shelly matter is white. The left valve adds concentric layers around its entire margin, not excepting the hinge
margin. The lower valve appears to grow from its posterior and lower half, the successive accretions being produced around the byssus.  This growth for a while seems to take place exclusively from
the posterior half of the shell, limited in front by the byssus, and even
after this growth has increased to twice the diameter of the embryo
shell, only a slight increase is noticed on its anterior margin, this latter addition being slightly reflected.  The left or upper valve grows
more rapidly, so that its margin overlaps the right valve at all points.
As the animal increases in size, the foramen increases also, and its
earlier boundaries are consequently absorbed.
"It will be seen by reference to the figures, that the growth of the
perforated valve is first posterior and downward, from the posterior
half of the shell; it then grows forward, avoiding the byssal plug, and
by successive additions surrounds the byssal plug, and ultimately
reaches the umbones of the larval shell, and even beyond and behind
this region.
" From these facts it is obvious that at an early stage the animal is
free, and for a time locomotive; that it has an elongate, oval, bivalve
shell, with close and regular lines of accretion, and that during the
latter stage of this growth it becomes attached by a byssus passing
from between the valves of the shell, as in IMytilus; that before the
growth of the larval shell is completed it drops over to one side, since
one valve only shows the notch upon its margin, and that so soon as
this growth ceases, a new growth takes place, looser in texture, and
white in color, as above described."
The figures are here given which accompanied the paper from
which this extract is taken.  At another time I hope to meet with
living examples of Anomia in this early stage.  The specimens from
which the above studies were made, I found upon dried sea weed, but
the shells told the story plainly enough, as will be seen by the figures.
Let us again revert to Steenstrup's views regarding the affinities of
the Brachiopods.  In the Proceedings of the Royal Danish Academy
for the year 1848, pp. 88, 89, Prof: Steenstrup again refers to the
supposed deviations of Anomia from the Lamellibranch, and showed
that they were not only closely related to the other Monomyarians,
but had no sort of approximation to Terebratula, or the other Brachiopods. In that meeting he indicated what he considered the true
place to which the Brachiopods should be assigned.
(365)




54
Having shown that they were isolated from tle Lamellibranchs, he
there said if the Brachiopods were rightly looked upon, they would,
according to his opinion, not find their place in the Molluscan series
at all, but in the series of annulated animals, and just in the class of
Annelids.  In this class, however, he might not ranrge them  as a
particular order, but as a link, or row of particular families, which
one after the other came near to the now living Serpula, going
slowly, without any abruptness, over to them.  Steenstrup believed
that if the relationship existing between the shells of Serpula and
the tests of Thecidia, Crania, etc., was a natural one, then the type of
the animal of Serpula might be recognized in the Brachiopoda.
At that time, Steenstrup believed that the Itippurites were Annelids too.  This view he informs me he has since abandoned, though
he regards them  as a very alienated type from the Lamellibranchs.
It seems strange, that in all the elaborate memoirs on the Brachiopods, no allusion has been made to Steenstrup's views on the subject.
These allusions to the affinities of the Brachiopods in a language of
which I was entirely ignorant, never came to my knowledge until
nearly two years after my first views of the Annelidan affinities of
the Brachiopods were published in " Silliman's Journal."  These
views he presents every year to the students at the University at
Copenhlagen, for yhich he tells me he has been highly blamed, but to
which lie still adheres.
In a leter to me he writes: -" You remember that Seneca says,
we ought always to go'Non pecorum modo, quo itur, sed quo eundum
est.' As to the last half of this sentence I must be quite silent, but
as to the first half I shall add that I have been highly blamed, that I
did not follow the common' quo itur' path."
Conclusion and Recapjitulation.
The oMollusca have always proved a stubborn group to define, and
simply because certain forms were placed with them, by general consent, which did not belong there. Milne Edwards first broke the spell
by rendering the affinities of the lower classes less antagonistic by separating the Mollusca into two great divisions, the Mollusca and Molluscoidea. This act was promptly adopted by Prof. Dana as before alluded to in the first part of this paper.
However vague and ill-defined the Vermes may be, there seems to
be no wider gap between their extreme forms than between Amphi(366)




55
oxus and IlVan, or the Lerneans and IIymenoptera, or between Aspergilluim  and Loligo. Were it not that Protula and Autolytus
increased by transverse division, one might look at low worms with
these anomalous features of reproduction, as remotely separated
from the Chatopodous groups.
In considering the assemblage of remarkable characters in the
Brachiopods, we must recognize in them  a truly ancient type, and
consequently a synthetic, or comprehensive type.  Thus while we do
not find them  in all their characters resembling any one group of
worms, I have endeavored to show that all their features, to a greater
or less degree, are shared by one or the other of the various groups
of the Vermes, with one or two features shared by the Arthropods.
It is important to rernalk in this connection that most of the
ancient groups differ froom  present groups with which they are associated.  Thus the Trilobites are widely unlike modern Crustaceans,
Milne Edwards and Van Beneden suggesting their affinities with the
Arachnids.  Tetrabranchiate Cephalopods are widely separated from
the Dibranchiate Cephalopods.  Crinoids are widely unlike modern
Echinoderms.  In other words, among the Mollusks, Echinoderms
and Crustaceans are ancient types widely different from the modern
types with which they are correlated.
So in worms we should expect to see ancient types, while presenting a high organization, yet differing from  present groups to which
they are unquestionably related.  And from the high complication of
structure of the Brachiopods, Tetrabranchiates, and other ancient
types, it would seem  that in their culmination in ancient times
they had the samie relation to animals living then as the higher
groups of present times bear to their associates.  As to the more
ancient forms of Brachiopods, it is probable with them, as with
other groups, that their lower members were soft bodied, and
the argument that has been urged, as militating against Darwin,
that animals of high complication of structure occur in the older
groups, becomes valueless, when we consider that the lower forms of
their respective groups are more often soft bodied, and that complicated forms of earlier times, were also culminating forms of pre-existing groups. In the light thrown upon the history of man by the
wonderful discoveries in Archseology, where we meet with traces of
an ancient civilization, with complicated language and manners, we
can surely believe in savage hordes pre-existing, from which this
ancient civilization has been evolved.
(367)




The  Tunicates, Amnphioxus, and many important groups, which
have taught us so much regarding the affinities of classes supposed
to be widely separated, had no hard parts to preserve, and nothing
by which  we could make out their structure.  Even the worms, with
the exception of the Chaetopods, have left us nothing, and the last
named group are known to us only by their tracks, spines, and tubes
left in the rocks.  Even  the hard leathery peduncle of the  fixed
Brachiopods has, in one or two instances only, been preserved.'
The earliest forms of Brachiopods thus far met with, are those
with thin, diaphanous shells (as in the Lower Lingula Flags), and
even before these forms existed we can legitimately conceive the
primary existence of certain Brachiopods, with the peduncle endowed
with high functional importance, and the anus terminating posteriorly,
as in those Annelids, like Lysilla, where the caudal portion is apodous, scarcely annulated, and devoid of seta.  Later appeared Brachiopods with an  anterior termination of the intestine, as  in  Phoronis and certain sipunculoid worms, and finally the peduncle became
attached, and the anus obliterated.
In regarding the relations of the Brachiopoda; let us suppose that nothing
was known of such a form as Lingula, and that the only forms of which we
had any knowledge, were the short and wide forms, like Chonetes. It would
certainly be considered a wild supposition to suggest the possible existence of a
Brachiopod, in which the shell should be translucent and. elastic, four times as
long as wide. A peduncle partially annulated, nine times as long as the body,
capable of sharp vermian contortions, having a rapid circulation within, incased in a sand tube, with the animal living free in the sand, and having a
limited power of locomotion. Surely it would. tax the powers of the imagination more to conceive this form than to simply endow a sedentary Annelid, like
Prot.uia, with the cephalic collar extending beyond the arms instead of surrounding the base, and secreting a dorsal and ventral plate, with the anus
terminating at the side instead of behind.
The presence of prominent dorsal. and ventral plates enclosing the arms from
which the arms in some cases are protruded and the consequent developement
of special muscles to move these plates, form the only marked differences between the Brachiopods and the Worms. In the Gephyreans as in the common
Sipunculus. for example, where the anterior portion is highly contratile, ponderous retractor muscles are developed with widely expanded bases, as in the
Polyzoa and in many Brachiopods.
In the Museum of the Geological Survey of Canada is a specimen of Eichwcalicia
subtrigonalis Billings, from the Black River Limestone, Lower Sil., showing the
silicified peduncle nearly as long as the shell, photographs of which IMr. Billings
sent me.
(368)




57
In most cases the Brachiopods are fixed, and it is well known that
with the radical change in the habits or conditions of any group, a
corresponding change is noticed in their structure.  Thus among the
Crustaceans the feature of attachment is accompanied with many
changes in the structure of the animal which mask their proper affinities, as in the Cirripeds, and notably in the Lerneans.  Similarly in
certain low worms, internal parisitism is generally accompanied with
a loss of gills, setae and cilia.1
On the same grounds, we should naturally expect to see great and
striking differences in those worms which are attached.   Consequently those Annelids which are fixed, while in most respects
resembling the free Annelid, yet differ from them  in being edentulous, the mouth or head, in some cases, supporting a crown of cirri,
oftentimes  springing  from  two  cephalic lobes, which  may  be
developed into a closely twisted spiral.  There is also a marked
degree of cephalization, the anterior rings forming a thorax, and
supporting brancheia, while the caudal region is often apodous, and
without sete, and in some cases not even annulated, and often separated from  the thorax by a deep constriction.
So in the Brachiopods, while in every feature of their internal
structure betraying their annelidan affinities, and while the errantian
forms with their long vermiform  and annulated peduncle, their locomotion by means of setse, and their power of fabricating a sand tube,
unite them clearly with the fixed and higlily cephalized Chlatopods;
yet in those groups that are attached, a remairkable concentration is
seen, and many features are presented which have heretofore obscured the affinities of the group.
To sum up the whole then. - Ancient Chstopod worms culminated
in two parallel lines, on the one hand, in the Brachiopods, and on
the other, in the fixed and highly cephalized Chaetopods.  The divergence of the Brachiopods, having been attained in more ancient times,
a few  degraded features are yet retained, whose relationships we
find in the lower Vermes; while from their later divergence the fixed
and cephalized Annelids are more closely allied to present free Choetopods.
And so we must regard  the Brachiopods as ancient cephalized
Chcetopods, while Serpula, Amphitrite, Sabella, Protula and others,
may be regarded as modern (later) cephalized Chcetopods.
Rolleston. Forms of Animal Life.
(869)




58
The following characters of the Vermes, mainly compiled from Rolleston's
"Forms of Animal Life," are here given, with the characters of the Brachiopods in parallel columns.
Vermes.                                Brachiopoca.
Bilaterally symmetrical, depressed        Bilaterally symmetrical, depressed
or flattened, or circular, never flat-   or flattened, or circular, never flattened laterally.                          tened laterally.
In many cases, above and below            In many  cases, above and below
similar, or these regions distinguished    similar, or these regions distinguished
with difficulty.                          with difficulty.
Most free, some attached.                 Some free, most attached.
Locomotor  muscles  closely  con-         Locomotor  muscles  closely  connected with  integumentary system,   nected  with  integumentary system
not only on ventral, but on dorsal and    not only on ventral, but on dorsal and
lateral aspects of body wall.            lateral aspects of body wall.
In some (Annulata) annulated ex-          Never annulated externally, except
ternally, and  divided  internally by    the peduncle.  Asingle dissepiment
dissepiments, in others not divided.      in Lingula.
Two layers of muscles in  body            Two layers of  muscles in  body
walls,                                    walls.
Digestive canal straight, or seldom       Digestive canal straight, or seldom
convoluted.                              convoluted.
Suspended in perivisceral cavity by       Suspended in perivisceral cavity by
partial dissepiments consisting of del-   bands consisting of delicate tissue.
icate tissue.
Peculiar   depuratory   apparatus,        Peculiar  depuratory   apparatus,
characteristic of the entire sub-king-   characteristic of  the  entire  class.
dom. This apparatus in the Annu-   This apparatus taking the shape of
lata taking the shape of bilaterally    bilaterally symmetrical tubes, opensym:mletrical tubes, opening  exter-   ing externally, and  communicating
nally, and comrnunicating with the    with the perivisceral cavity by ciliperivisceral cavity by ciliated infun-   ated infundibuliform orifices.
dibuliform orifices.
Nervous system  consisting at most       Nervous system consisting at most
of simple cesophageal collar, to which    of a single cesophageal collar.
a few accessory ganglia, or chain of
ganglia, may be appended.
In some the bilateral elements of         The bilateral elements of the nerthe nerve chain widely diverging.        vous system widely diverging.
Generative products in  most, set        Generative products set free in perifree in perivisceral cavity.             visceral cavity.
(370)




59
Possessing  chitinous outgrowths,       Possessing chitinous and calcareous
either as scales or plates, hairs or   outgrowths, as scales or plates, and
spines, the latter being secreted by    chitinous outgrowths, as hairs  or
setigerous follicles.                   spines, the latter being secreted by
setigerous follicles.
Cuticle perforated by minute pores.     Cuticle perforated by minute pores.
Perivisceral cavity lined in some       Perivisceral cavity lined by delicate
cases by delicate ciliated membranes.   ciliated membranes.
An extensive vascular system, con-      An  extensive  vascular  system,
taining a colored fluid representing    which may represent the pseudo-hoethe pseudo-hemal system.  The cor-   mal system. The corpuscular nutripuscular nutritive true blood is con-   tive true blood is contained in the
tained usually  in the  perivisceral   perivisceral cavity alone.
cavity alone, and in a few instances
is found penetrating the pseudo-hemal system.
Embryos  distinctly  transversely       Embryos  distinctly  transversely
segmented. In some cases attached    segmented. In most cases attached
by caudal portion at a later stage.     by caudal portion at a later stage.
In some groups intestine having no      In some groups intestine having no
anal outlet.                            anal outlet.
Sedentary Annelids.
Usually having body divided into        Having  body  divided  into  two
regions, the thoracic and caudal.       regions, the thoracic and caudal.
Springing from the head, in many,       Springing from the head two arms,
two arms, often twisted into a closely   often twisted into a closely wound
wound spiral, and sustaining ciliated   spiral, and sustaining ciliated cirri.
cirri.
Encircling  the head, or base of        Encircling the head and arms, a
arms, a flaring membrane split upon   flaring membrane split upon the sides,
the sides, and notched, in the dorsal   and often notched in the dorsal and
and ventral median line.                ventral median line.
Mouth unarmed.                          Mouth unarmed.
Mostly fabricators of sand tubes        One, Lingulapyramnidata, fabricates
which invests the body.                 a sand tube, which invests the body.
Gephryreans.
Never definitely segmented.  Anus       Never definitely segmented. Anus,
opening anteriorly in many.             when present, opening anteriorly.
(371)




60
I desire to express here my deep gratitude to John E. Gavit, Esq.,
who has constantly aided me in getting material together in our Eastport dredgings, who has freely placed at my disposal his extensive
microscopical apparatus, and who has, by timely suggestions, aided
me in many ways.
My thanks are also due to Dr. J. R. Nichols for the constant use
of a Smith &  Beck Binocular; to Dr. Elliott Coues, U. S. A., through
whose attention and kindness I was enabled to study living Lingula
on the coast of North  Carolina; and to Prof. J. W. Dawson, for
specimens of Rhynchonella, and for the loan of a stout dredge; and
to Dr. P. P. Carpenter for accompanying me to the mouth of the St.
Lawrence, and for aiding me in dredging for Rhynchonella.  I am
also under obligations to Prof. A. E. Verrill, Prof. A. Hyatt, Thomas
Bland, Esq., and Edward Burgess, Esq., for many favors.
To the Editors of " Old and New " I am  indebted for the use of
the engraving which adorns the title page; and to Mr. C. A. Walker
for the pains taken in engraving my hasty drawings.
EXPLANATION OF PLATE I.
Figures I to 7, inclusive. Lingzula pyramidata Stimpson, natural size, from life.
Figure 1. Representing two specimens as they appear partially protruding from
the sand, showing the sand disturbed by them about their burrows.
Figures 2, 3, 4, 5 and 6, different specimens in various positions after having been
removed from the sand in which they were found; Figure 3 showing the animal
while in the act of occillating the dorsal plate; Figure 4, with peduncle straightened.
Figure 7. Representing bottom of earthen bowl, in which eight specimens had
been kept alive in sand; the sand covering the bottom of the bowl an inch and a
half in depth, above which the Lingulas protruded, and below which they would
partially disappear with a quick jerk when alarmed. On removing the sand they
presented the appearance as here given. Sand tubes had been made by them, adhering to the bottom of the bowl, quite unlike those made by them when free in
the sand. Into these tubes the Lingulas had partially receded, as represented by
the figure; a. represents a deserted sand tube.
Figure 8. Protula media Stimpson, natural size, reduced from a drawing made
by 3Mr. Emerton, kindly loaned by Prof. Verrill.
Figure 9. Showing thorax and left arm enlarged from same drawing; the pectinated character of cirri are not fully shown. This drawing is inaccurate in not
showing the calyx or membrane encircling the base of cirri, and in not properly
showing the pectinated character of the cirri. The correct and beautiful drawing
of Mr. Emerton's will be published by Prof. Verrill in the Proceedings of the Connecticut Academy of Sciences.
Figure 10. Serpuzla crater, from Claparede.
The last three figures are given to show the distinct separation of thorax and
caudal portion.
(372)




PCX,  Bost Sec. Nt  t. Hist. l, tVol. XV.la
S'(ys.HSB:,    \     SYT I  POSITI3ON  OF  rRAC   XS.I   PO)A.




























EMBRYOS OF TEREBRATULINA.