LIBRARY

OF THE

UNIVERSITY OF CALIFORNIA

UNIVERSITY OF CALIFORNIA PUBLICATIONS
ZOOLOGY

Vol. 1, No. 7, pp. 227-268, Pis. 20-23 June 1, 1904

THE STRUCTURE AND REGENERATION

OF THE POISON GLANDS

OF PLETHODON

B

C. O. ESTERLY

BERKELEY

THE UNIVERSITY PRESS
PRICE $0.50

UNIVERSITY OF CALIFORNIA PUBLICATIONS

BOTANY.— W. A. Setchell, Editor. Price per volume $3.50. Volume I (pp. 418)
completed. Volume II (in progress):

No. 1. A Review of Californian Polemoniaceae, by Jessie Milliken. Price, $0.75

GEOLOGY.— Bulletin of the Department of Geology. Andrew C. Lawson, Editor.
Price per volume $3.50. Volumes I (pp. 428) and II (pp. 450)
completed. Volume III (in progress):

No. 16. A Note on. the Fauna of the Lower Miocene in California, by

John C. Merriam Price, .05

No. 17. The Orbicular Gabbro at Dehesa, San Diego Co., California, by

Andrew C. Lawson . . Price', .10

No. 18. A New Cestraciont Spine from the Lower Triassic of Idaho, by

Herbert M. Evans ... ... Price, .10

No. 19. A Fossil Egg from Ariz , , . Vi-'m. Conger Morgan and Marion

Clover Tallmon ..... Price, .10

No. 20. Euceratherium, a New Ungulate from the Quaternary Caves of

California, by William J. Sinclair and E. L. Furlong. Price, .10

EDUCATION. -^Elmer E. Brown, Editor. Price per volume $2.50.

Volume I (pp. Notes on the Development of a Child, by Milicent W.

. Price, 2.25

Vol. II (in progress). — No. 1. Notes on Children's Drawings, by Elmer E.

Brown Price, .50

Vol. Ill (in progress). — No. I. Origin of American State Universities, by

Elmer E. Brown . Price, .50

No. 2. State Aid to Secondary Schools, by David
Rhys Jones Price, .75

ZOOLOGY.— W. E. Ritter, Editor. Price per volume $3.50. Volume I (in
progress):

No. 1, The Hydroida of the Pacific Coast of North America, by Harry B.

Torrey Price, ; .00

No. 2. A Case of Physiological Polarization in the Ascidian Heart, by

Frank W. Bancroft and C. O. Esterly .... Price, .10

No. 3. Embryology and Embryonic Fission in the Genus Crisia, by Alice

Robertson . Price, .50

No. 4. Correlated Protective Devices in Some California Salamanders, by

Marian E. Hubbard Price, .20

No. 5. Studies on the Ecology, Morphology, and Speciology of the Young of
Some Enteropneusta of Western North America, by Wm. E. Ritter
and B. M. Davis '. . Price, .50

No. 6. Regeneration and Non-Sexual Reproduction in Sag.irtia Davisi, by

Harry Bea! Torrey and Janet Ruth Mery . . . Price, .15

No. 7. The Structure and Regeneration of the Poison Glands of Plethodon,

C. O. Esterly ... . . Price, .50

UNIVERSITY OF CALIFORNIA PUBLICATIONS
ZOOLOGY

Vol. 1, No. 7, pp. 227-268, Pis. 20-24 June 1, 1904

THE STRUCTURE AND REGENERATION
OF THE POISON GLANDS
OF PLETHODON,/"

BY N^c
C. O. ESTERLY. i£.

It has long since been held that the skin glands of both the
Urodela and the Auura are of two kinds. This distinction was
first made by Ascherson '40 in an investigation of the glands in
the web of live frogs, and was based upon the size, shape, and
location of the glands without regard to function or microscopic
structure. That the skin of Amphibians secretes a substance
other than the well-known mucus, and clearly poisonous, has
been shown by many physiological and toxicological experiments
and investigations (Albini '56; Boulenger '92; Calmels '83;
CappaiTelli '83; Dutarte '89; Gratiolet and Cloez '51-'52;
Hubbard '03; Phisalix-Picot '00), and the facts gained from
experiment are upheld as far as possible by histological evi-
dence. Microscopic examination shows that there is more than
one kind of gland. (Ancel '02; Coghill '99; Eberth '69;
Eckhard '49; Engelmami '72; Hensche '56; Leydig '76 a; Pau-
licki '85; Phisalix-Picot '00; Schultz 89; Seeck '91; Stieda '65;
Szczesny '67; Wiedersheim '86). These have generally been
distinguished as granular (Kornerdriisen) and clear, according
to the appearance of the secretion contained in them, the former
having been almost unanimously looked upon as making the
poison series, the latter the mucous series. The suggestion has
been made, however, that the various glands are only the young
and old stages of one sort of gland (Junius '98), and this ques-
tion will receive further consideration in the present paper.

R15

'2'2S University of California Publication*. [ZOOLOGV

The poison glands are in most cases much larger than (hose
of the mucous variety, and their enormous cells (Riesen/ellen of
Leydig) completely till the interior of the gland so that there is
no lumen. This character distinguishes these glands from the
others, which are provided with a low, cubical epithelium sur-
rounding a capacious lumen. (PI. XX, Fig. 2.) The mucous
secretion filling the lumen is very distinct from the heavily
granular contents of the cells of the poison glands. (PI. XXIII,
Fig. 31.) The two sorts of glands are further distinguished by
other features, chief of which is the staining reaction of the
mucous secretion (Nicoglu '93; Hoyer '90). These observers
used thionin as a specific stain for mucus and found that the
small glands stain rose-red while the others are uncolored.

The foregoing general facts have been determined chiefly upon
the Anura and the various European Salamanders, especially
Triton and Salamaudra. But Pletliodon oregonensis, a salaman-
der found about Berkeley, forms a particularly interesting object
for the study of the poison glands because of their unusual
development on the tail of this animal. This seems to be a
protective character associated to some extent with the ability of
the animal to throw off its tail under stress of circumstances. It
has been shown by experiment that the secretion of the glands
of the tail is poisonous or obnoxious to certain animals, a char-
acter which probably belongs to the dorsal glands (Hubbard, '03) ,
which are very large and much more developed than elsewhere on
the body. In this respect Plethodon appears to resemble Triton
cristatus. (Capparelli '83.) However, the poison glands of
Plethodon are not confined to the dorsum of the tail; much
smaller ones are found 011 its ventral surface and also on
the trunk and head of the animal, intermingled with mucous
glands, which occur in all situations where the poison glands
are found.

The principal question considered in the present paper con-
cerns the changes occurring in the formation of the secretion
and its expulsion from the glands. In Plethodon this involves
the death of the glands, as Seeck ('92), Nicoglu ('93), Vollmer
( '93) and others have shown for other Amphibia. The exhausted
glands are here renewed or replaced in the manner described by

VOL. i.i Esterly.— Poison Glands of Plethodon. ±29

Heidenhain ('98«). Vollmer ('!):!) and Nicoglu ('93) . This
process consists in the growth into the old glands of a new and
smaller gland, which, however, is mucus in character, contrary
to the statements of Nicogln ('93), so that the poison glands
develop from the mucous to the poison variety. This has been
suggested but not definitely shown by Hover ('90) and Junius
('98), and distinctly denied by all other investigators of the
regeneration of these glands. Under the histological structure
of the glands will be considered some new points in the muscu-
lature, especially as to the presence in the epidermis of an
apparatus for closing and opening the duct. The innervation of
the muscles and epithelium of the glands will also receive atten-
tion.

This work was done under the direction of Professor C. A.
Kofoid, and my heartiest thanks are due him for very kind
assistance in every way and for criticism of results.

MATERIAL AND METHODS.

In order to obtain the best insight into the structure of the
glands of the tail, sections in three planes have been made of
that entire organ. The tissue was in all cases perfectly fresh
and was fixed in Zenker's fluid, which has been satisfactory in
all respects. Washing in 70.% iodine-alcohol followed the use of
the fixative.

That bony tissue might not hinder the passage of sections in
any plane through the whole tail, the tissue was subsequently
decalcified in a 5% aqueous solution of nitric acid for from
twelve to twenty-four hours, followed by immersion in a 5%
aqueous solution of sodium sulphate for the same length of time,
and thorough washing in running water for from twenty-four to
forty-eight hours.

Parafflue sections have alone been used, varying in thickness
from 3^ to 10 microns. The sections were fixed to the slide by
the water- albumen method, and in all cases where possible
staining was done on the slide.

A considerable variety of stains has been employed. The
most successful have been Mallory's ('00) connective tissue stain
(acid fuchsiu, phospho-molybdic acid, anilin blue-orange G),

230 University of California Publications. [ZOOLOGY

Van Gieson's haemalum and piero-fuehsin, and the iroii-haema-
toxylins of Benda and Heidonhain. I have found it of consider-
able advantage to increase the percentage of fuchsin in Malic >ry's
stain to as much as 1.5 or 2%. This stain, as a whole, when
successful is very beautiful, but its action varies most unac-
countably. The staining and differentiation will be perfect in
some sections, while in others on the same slide the differential
coloration will fail completely. But the range of application of
the stain seems to be almost unlimited except for purely cytolog-
ical work.

Other stains have been used, such as Mayer's neutral and
acid haemalums followed by eosin, orange G, erythrosin and
Congo red; safranin alone or in combination with light green;
ferric chloride haematoxylin, and such special stains as the
phosphotungstic acid haematoxyliu of Mallory and Cajal's ('03)
silver nitrate-pyrogallic acid method for nerves, Tanzer's orcein
for elastic fibres, and Mayer's muei-carmine as a mucus stain.

As has been said, 'the largest poison glands of Plethodon are
situated on the back of the tail, and in cross sections (PI. XX,
Fig. 1, p.gl.) it may be seen that they lie in that portion of the
skin covering the dorsal half of the tail. Here the greatest
development is in the glands at either side of the mid-dorsal
line, while farther down on the sides they gradually diminish
until they are considerably smaller and not readily distinguished
by their size from the larger mucous glands. (PL XX, Fig. 1,
ni.f/l.) The coloration also of the tail gives a clue to the loca-
tion of the largest glands. The dorsal half of the tail is black
or brown, while the ventral half is orange or yellow, and the
glands under consideration are confined almost entirely to the
darker portion. The mucus glands are found largely on the
ventral side of the tail, but they also occur along the dorsal sur-
face. In this region they lie between the necks of the large glands.

The poison glands form large sacs, extending from the epi-
dermis to the inner layer of the corium. (PI. XX, Fig. 1.) In
shape they are elongated, with oval or even somewhat rectangu-
lar outline. The ducts are short, and the transition from the
body of the gland to the neck and duct is not sharp as in the
mucous glands, which are regularly flask-shaped.

VOL. i.] Extrrhj. — Poison Glands of Pie thodtm. i>:!1

It ha.s been shown (Hubliard '().'!) that the swollen appear-
ance of the tails of some animals is due to the increased
development of the poison glands posterior to the well-marked
constriction found just behind the cloaca in such cases. That
this is really true appears in the study of a series of sections of
a swollen tail passing from the tip up to and including part of
the eloacal aperture. In the constriction the dorsal glands are
very small comparatively, and are hei-e no larger on the back of
the tail than on the ventral side. But behind the constriction
their development is much greater, and one may trace the regular
increase in size as the series passes from the constriction back to
the enlarged portion of the tail. Everywhere in the tail, except
in the constriction at its base, the difference in size between the
glands on the dorsal and ventral surfaces is maintained.

As is well known, the bodies of all the glands lie in a spongy
connective tissue, the middle layer of the corium, which in the
region of greatest development of the poison glands is increased
enormously in thickness (PL XX, Fig. 1, m.c.l.), being alone
from one-sixth to one-fourth or more of the dorsal- ventral dimen-
sion of the tail. (Hubbard '03. ) The bottoms of the large glands
rest upon or come very close to the inner layer of the corium.
(PI. XX, Fig. 1, i.e. I.)

The ducts of both mucous and poison glands pass through the
outer corium layer and the epidermis, the long axis of the gland
which passes through the duct and its mouth being perpendicular
to the surface at the point where the duct opens to the exterior.

The histological structures found immediately surrounding
the ducts of the poison glands are in no essential points different
in Plethodon from those in other salamanders. The funnel cells
and their processes (PI. XX, Figs. 1 and 2, fl.c.) are present
as in Triton (Nicoglu '93) and in Salamandra (Ancel '02). The
membrane-like structure lining the duet belongs to a specialized
cell of the epidermis, corresponding to the "stoma cell" of Eberth.
As Nicoglu has shown, the mouths of the glands lie within these
cells, processes of which extend down in the ducts about as far
as the lower limit of the epidermis or a little farther. (PL XXIII,
Fig. 27, p.fl.c.) The prolongations stain black in iron haema-
toxylin, reddish in Mallory's and yellow in Van Gieson's stain.

232 University of California Publications. [ZOOM>QY

In addition to the funnel cells proper, Nicosia lias described the
arrangement of the cells in the epidermis which are to replace
the funnel cells as they are thrown off at the time of moulting.
The same condition is found in Plethodon and does not differ at
all from that in Triton (Nicoglu '03) or in Salamandra (Ancel
'02). (PL XX, Fig. 4; PI. XXIII, Figs. 27, 28, 29, 30. rn>. ('.)

As further evidence that the cells described as replacement
cells really have that function, Plethodon shows that the low fi-
ends of the replacement cells, especially those nearest the duct,
extend inward as do the prolongations of the funnel cells. (PI.
XXIII, Fig. 27, rep. c.) The arrangement of the former very
strongly suggests that they are of the same nature and function
as the funnel cells. And in cross sections of the ducts the
replacement cells are shown rolled one within the other as in
PI. XX, Fig. 4; PI. XXI, Fig. 16, rep. C. The cell first to
replace the one thrown off at moulting immediately surrounds
the duct; the cell next to replace this one lies concentrically out-
side it, and so on. In Mallorv's stain the cell boundaries are
very distinct, and there can be no doubt of the structure as
described either in cross or in longitudinal section of the ducts.

The walls of the gland sacs proper are composed, in many
Amphibia, of a number of elements which have been described
and all of which need not be discussed at length here. In the
most peripheral layer are connective tissue and elastic fibrils, as
is shown by the use of Mallory's connective tissse stain for the
former and orcein for the latter. Nerves, lying in this layer, also
extend over the gland. Inside the connective tissue sheath, MS
it is generally called, lie the muscle fibres, and next to them the
epithelium of the gland.

(See in this connection Drasch '92, '94; Eberth '69; Eck-
liard '49; Englemann '72; Hensche '56; Leydig '76, a, b;
Paulicki '85; Phisalix-Picot '00; Schuberg '03; Schultz '89;
Seeck '91; Stieda '65; Tonkoff '00; Wiedersheim '86.)

Because of the intimate relation between the connective
tissues of the gland wall and those of the corium, it is necessary
to consider in more detail the structures of the inner, middle
and outer layers of the corium. Schuberg ('03) has studied the
corium of Axolotl most minutely. I have confirmed his results

VOL. i.i Esterly.— Poison Glands of Plethodvn. 233

in general in Plethodon, and particularly as to the relation of
the connective tissue bundles of the inner layer of the coriura to
those of the middle layer. He found (p. 222) that columns of
connective tissue pass perpendicularly from the inner into the
middle layer, and seem to serve as mechanical supports for the
glands, since under each one such a column of tissue is found.
The same arrangement occurs in Plethodon except that the
pi'i-pendicular bundles do not stand beneath the glands, but
around them, as can be seen in longitudinal sections of the
glands. (PI. XX, Fig. 5, c.t.b.) In spaces between the large
glands or on the ventral side of the tail, the bundles from the
inner layer of the corium can be seen especially well. The con-
nective tissue fibres from the wall of the gland unite with the
outer layer of the corium which then, lying next the muscle layer
of the gland, passes toward the surface of the epidermis and
ends on the side of the neck of the gland about a third of the
distance between the inner and outer boundaries of the epider-
mis. (PI. XX, Fig. 8; PI. XXIII, Figs. 27, 31.) This appears
in both longitudinal and cross sections of the ducts. In the
latter can be seen a crescent of connective tissue on each side of the
duct between the muscle fibres and one of the replacement cells.
(PL XXIII, Figs. 28, 29; PI. XXI, Fig. 16, c.t.) Ancel ('02, PI.
IX, Fig. 22) seems to have shown the same in longitudinal section.
The elastic fibres pass through the inner layer of the corium
into the middle layer in company with the connective tissue
bundles as Schuberg ('03, p. 231) has described. The elastic
fibres can be followed around the glands, and over them in
tangential sections. The fibres are of nearly the same calibre
throughout and all of them take the same general direction,
from the inner corium layer perpendicularly or sharply turned
toward the outer layer. As in the case with the connective
tissue bundles the elastic fibres pass at once around or over the
large glands, and ar-e not found arranged perpendicularly beneath
them as in Axolotl. (Schuberg '03, p. 232, Fig. 14.) On the
sin-face of the gland they are branched in a few cases; usually,
however, only single fibres of wavy, curving and regular outline
are visible, ending before the outer corium is reached (Schu-
berg '03, PI. XXI, Fig. 9, t /./.).

234 University of California Publications. [ZOOLOGY

Between the connective tissue layer and the gland epithelium
lies the layer of contractile or smooth muscle fibres. These were
first shown histologically by Hensche ('56), though before him
Ascherson ('40) had observed movements of the living glands.
Since this time there has been no doubt of the existence of
muscles in the walls of the poison glands (Coghill '99; Drasi-h
'89, '92, '94; Eberth '69; Eckhard '49; Englemann '72; Heid-
enhain '93 a, b; Leydig '76, a, b; Massie '94; Nicoglu '93;
Paulicki '85; Phisalix-Picot '00; Schultz '89; Seeck '91; Stieda
'65; Szczesny '67; Vollmer '93). As regards the smaller series
of glands the question seems to be open. The absence of con-
tractile fibres on them has been used as a character to separate
them from the large glands. The muscles of the large glands
are arranged in a single layer and have a general meridional
direction on the gland, converging toward the upper pole. The
fibres are usually simple but may be branched (PL XX, Fig. 7) ;
this occurs mostly on the lower part of the gland. Neither do
the muscles form a continuous sheet about the gland; the indi-
vidual fibres are separated by spaces of greater or less extent.
I have not been able to find with certainty muscles on glands
which are mucous in nature.

The nuclei of the contractile cells, contrary to the description
of Nicoglu ('93, p. 437,) and such figures as his and those of
Vollmer ('93), lie in the upper region of the glands just outside
the uppermost gland cells, yet still well beneath the epidermis
(PL XX, Fig. 6; PL XXIII, Fig. 31, »«,«.). The first observer
mentioned has shown (his PL XXII, Fig. 12) the nuclei of the
muscle cells in various locations about the periphery of the glands;
but in Plethodon the nuclei have a constant position as described
and are found only there. In the region of the nuclei the muscle
fibres are considerably larger than elsewhere on the gland, as is
shown in PL XX, Fig. 8, m.f., or PL XXIII, Fig. 31, m.f., so that
the muscle, especially in longitudinal sections of the glands, seems
to bear a flask-shaped expansion. From this point it is possible
to trace a single fibre very nearhr to the base of the gland, and
also outward around the neck of the gland into the epidermis.
(PL XX, Fig. 6.) The connection of the muscles with the
epidermis has been reported by Nicogln ('93) and Heidenhain

vor,. i.] Kxti'i-li/. — Poison Olands of PlethodoH. -•!•">

('93 6), and the arrangement in Plethodon is a similar one
except as regards the presence of the "Schaltstiick" cells described
by them. In Plethodon the "Schaltstiick" is not demoiistrably
present except in one or two questionable cases in all my prepa-
rations, and Vollmer ('93) found that it is very often absent
even in Triton. There can be no doubt, however, that the
muscles send processes into the epidermis. This is especially
well shown in longitudinal and cross sections of the ducts.

The statement that the muscle nuclei of the poison glands in
Plethodon lie only in the necks of the glands, instead of generally
distributed about the periphery as held for other animals, may
be supported by several facts. In the first place, longitudinal
sections of the glands through the duct and mouth show two
nuclei, one at each side of the gland where the sac begins to pass
over into the duct. In sections of the same plane which pass a
little to one side of the duct (PI. XXI, Fig. 10, mf., nui.) may
be seen in some cases the obliquely cut ends of as many as seven
muscle cells each with its nucleus in situ, and occupying exactly
the position relatively of the two lateral nuclei which are shown
in the median section. There can be no doubt of their structure.

Cross sections of glands and ducts are also very instructive
on this point. In such, especially if stained in Van Gieson
(PI. XX, Fig. 12) , there are shown in many cases the light yellow
muscle fibers between the gland cells and the connective tissue,
when the plane of the section passes more deeply through the
gland than the position of the nuclei of the muscles. But when
the gland is cut across at the level of the nuclei, the evidence
gained from longitudinal sections is even more strikingly upheld.
In such cross sections can be seen as many as twelve or fourteen
muscle fibres stained light yellow (in Van Gieson), and in very
sharp contrast to them the brown or black nuclei. And in this
region the section of the muscle is larger than it is deeper in the
gland; this corresponds to the flask-shaped enlargement seen in
median longitudinal section (PL XX, Fig. 8, mf.). If a series
of frontal or cross sections of the tail is studied, it will be found
that while the muscles theinselves can be traced until the bottom
of the gland is reached, nuclei never appear again which are
unmistakably those of the muscle fibres. The only place in

236 University of California Publicnfimis. ! ZOOLOGY

which one can be sure that he is dealing with nuclei of the con-
tractile fibres is in the location above described. Hundreds of
sections have been carefully examined and there has never been
a case of a fully formed gland in which the muscle nuclei are
situated in any position except that described. Not only is this
true in stains such as Mallory's and Van Gieson's but also in
clear nuclear stains like iron haematoxylin .

That those observers who describe muscle nuclei on the
periphery of the gland sacs have mistaken connective tissue
nuclei for them, seems to me very probable. Nicoglu ('93,
p. 438) says that the nuclei often occupy an eccentric position,
so that even with oil immersions one cannot see that there is
any protoplasm of the muscle cell about them. His description
('93, p. 436) of the flattened narrow nuclei of the muscle cells
applies more to connective tissue nuclei. That these occur in
the walls of the glands has been observed by Paulicki ('85,
p. 158), who says: "An die Driisen treteu gewohnlich
sich nach oben erstreckend bindegewebige Strange init Kernen."
And the figures of Schuberg ('03), especially Fig. 28, show that
this is true for the glands of Axolotl. From these facts and
from my observations on Plethodon it is clear that connective
tissue nuclei closely invest the glands, and evidence is added to
that already brought forward to show the location of the nuclei
of the muscle cells.

The processes of the muscles passing into the epidermis serve
to connect the fibres with the outermost layer of the skin. This
has been shown, as said before, by Heidenhain ('93) and Nicoglu
('93), as well as by Ancel ('02), and there is nothing to be
added to the description given by the former except, as before
stated to mention the frequent non-occurrence of the Schaltstuck
as such. This is a structure described as containing about four
cells which are arranged in a ring about the neck of the gland
at the. lower boundary of the epidermis. The cells form seem-
ingly the principal points of insertion of the muscle fibres, but
this cannot be so in Plethodon where the Schaltstiick is virtually
absent. Otherwise it may simply be said that the upper or outer
ends of the muscle fibres pass into the epidermis and end between
the replacement cells of the funnel. This can be seen fairly

VOL. i.] Esterly. — Poison Glands of Plethodon. 237

well in cross sections of the gland ducts in the epidermis where
the cut ends of the muscles arc seen dose beside the funnel cell
(PI. XXIII, Figs. 28, 20, prol. tii.f.). In good longitudinal
sections of the ducts the muscles (PI. XXIII, Fig. 27, prol.
m.f.) are seen to end between the older replacement cells which
are already elongating into their typical form (same, rep. r.).
Xi. -no-lii ('():!) represents the endings as between the cells, but
Aucel ('02) seems to consider them as special parts of cells. At
any rate he has shown (Fig. 22) the fibrils as within cells in the
epidermis. I have not been able to find such structures as he
shows; there can be hardly any doubt that the prolongations of
the gland muscles into the epidermis end between the replace-
ment cells. Nicoglu and Heidenhain ('93) and Ancel ('02) have
remarked upon the existence of intercellular bridges between the
muscle cell on the one hand and ectodermal epithelial cells on the
other, as Nicoglu says (p. 440), "von ga7iz ahulicher Art wie
zwischen den Oberhautzellen selbst."

I have not found the intercellular bridges in Plethodon
between epithelial and muscle cells, but all the facts concerning
the connection of muscle and epidermal cells have been taken as
evidence of the ectodermal origin of the muscles of skin glands.
This has been so often commented upon that it is useless to more
than call attention to it here. The evidence gained from a study
of the development of the glands shows that the muscle fibres
come from the Malpighian layer of the epidermis (Ancel '02;
Vollmer '93; Junius '98). This, added to the facts already
cited, and coupled with the observations of many investigators
(Engelmann '72; Seeck'91; Heidenhain '93; Nicoglu '93) seems
fairly conclusive that the muscles of the dermal glands are of
ectodermal origin: (Compare also in case of sweat glands,
v. Kolliker '89; Handbuch des Gewebelehre des Menschen, pp.
138 and 258).

The existence of a sphincter or constrictor muscle for the
glands has been claimed by Schultz ('89), who described a band
of muscle fibres running around the neck of the gland beneath
the meridional layer. This observation has been disproved by
Drasch ('94) and Nicoglu ('93), and I have been unable to find
such a structure in Plethodou. And there is no evidence of the

238 University of California Publirntin-ns. [ZOOLOGY

epithelial plug of Drasch ('94) for restraining the contents of
the gland under pressure. Phisalix-Picot ('00) mentions (pp.
•44-45) an orbicular muscle, but gives no description or drawing
of it, so that her meaning is obscure. Dilator muscles for tin-
ducts or mouths of the glands have never been described.

However, both dilator and constrictor muscles occur about
the mouths of the poison glands of Plethodon. These are best
shown in sections of the epidermis parallel to the sin-face, stained
in Mallory's connective tissue stain, which are, of course, also
cross sections of the ducts. All three sets of gland muscles may
very often be seen in one such section (PI. XXIII, Fig. 30, con.
m., dil. m., m.f.; Figs. 28, 29 also). In these cases it will be seen
that the duct (l.<l.) in the epidermis is oval in cross section, and
that at each end of the oval is a triangular mass of fibres, stain-
ing red in Mallory, as do the muscles on the body of the gland.
The fibres converge toward the duct and insert upon the replace-
ment cells nearest the funnel in such a way that by contracting
they will bring the lips of the duct together and so close or
greatly diminish its lumen (PI. X^Q, Fig. 16). The con-
strictor fibres are differentiations of the cell whose large nucleus
(PI. XXI, Figs. 14, 16; PI. XXII, Fig. 28, iute.tf.iiM.),
stands at the ends of the elliptical opening of the duct. The
fibres lie within this cell as can be especially well seen in longi-
tudinal sections of the glands which do not pass through the
duct. Here it appears that the cell of which the constrictor
fibres are a part, together with its nucleus, lies in the deepest
layer of the epidermis immediately upon the outer layer of the
corium. This cell seen in surface view is equal in extent to
several of the neighboring epidermal cells, but in cross section
it is very much flattened (PI. XXI, Fig. 13). Ancel ('02)
has figured such a cell, but gives no clue as to its function.

The dilator fibres belong to the same cell of which the con-
strictors form a part, and are at a slightly lower level seemingly
than the latter. The action of the dilator is two fold. Some
fibres pass around the ends of the oval opening of the duct
(PI. XXIII, Fig. 28, dil. m.; PI. XXI, Fig. 14) and when they
shorten they tend to separate the lips of the lumen more widely,
by pressing at the ends of the ellipse. This is evident when it

VOL. i.] Esterly. — Poixnn (Unmix t>f I'lrfhtiilon. 239

is seen that the mass of dilators is often concave in outline
toward the center of the duct (PL XXIII, Fig. 28: I'l. XXI, Fig.
!.">, ilil.m.), so that in contraction the fibres first mentioned
pull in the general direction of the major axis and toward the
center of the ellipse. Other dilator fibres attach at the edges of
the duct near the end (PI. XXI, Fig. 14), and in shortening pull
in the direction of the minor axis of the ellipse, thus widening
the lumen by spreading its walls at the tips of the oval. (PI. XXI,
Fig. 14. ) The entire effect of the dilator fibres is to make the
aperture of the duct nearly circular, thus offering freer exit to
the secretion. Their action would be to open the duct from the
form shown in PI. XXI, Fig. 16, to that, for example, in PI.
XXIII, Fig. 30.

The fact that the constrictor and dilator fibres lie entirely
within the epidermis need not militate against their having the
function of muscles, for in the case of the intrinsic gland muscu-
lature it has been well established that it has an ectodermal
origin. It is certain that the arrangement and appearance of
the fibres described as constrictor and dilator muscles are such as
to suggest very strongly both that nature and function. The
coloration in Mallory is exactly that of the smooth muscles of
the glands ; and the convergence of the constrictor fibres to their
insertion in a position where contraction would close the duct;
the endings of the dilators in places to be of greatest advantage
in widening it when the muscles contract — all these facts lead
one to conclude that he has to deal with an apparatus for closing
and opening the ducts of the glands.

The muscles of the poison glands, as has been said, immedi-
ately envelop the secretory cells. The entire gland is filled with
enormous cells, the generally recognized "Giftzellen" of many
authors or the "giant cells" of Leydig. In such glands a lumen
does not exist; this is especially well shown in sections of the
tail of a tadpole 38 mm. in length, in which the cell boundaries
are distinct, the secretion not yet being present in sufficient
quantities to obliterate them. There it will be seen that the ends
of the cells are in contact with the middle of the gland, thus
doing away with any trace of a lumen (Nicoglu '93; Seeck '91;
Calmels '83). A glance at the figures will serve to distinguish

240 University of California Publication*. IZOOI.OM

in this respect the poison and mucous glands; the latter have
capacious lumens (PL XX, Fig. 2), often filled with a clear
secretion. The large gland cells each have a number of nuclei
(Nicoglu '93; Drasch '92), not over four in Plethodon. They
are round or oval, of regular outline, and lie normally upon or
very near the wall of the gland, and so at the base of the cells.
The internal structure of the nuclei is simple. There is a scanty
network and few chromatin granules; usually also one or two
nucleoli.

The cells and nuclei of the small or mucous glands are distinct
in every way from those of the poison glands. The cells are low
and cubical and show a filar structure (psendo-filar, Nicoglu '93) .
This is seen in sections stained either with Van Gieson, Mallory
or iron haematoxyliu . The nuclei are smaller than those of the
poison glands, and angular instead of regular in outline. They
invariably stain intensely black in iron haematoxylin, remaining
so when the nuclei of the giant cells have decolorized to a very-
faint gray (PI. XX, Figs. 2, 3; PI. XXII, Figs. 18, 19, 20).

A general comparison of the two sorts of glands might be
instituted in some such terms as these. The poison glands are
very much larger than the mucous glands, and have contractile
walls; the mucous glands lack this character. The extreme
dimensions' of the former on the tail are approximately from 1400
microns in length and 380 microns in breadth to 680 microns in
length by 200 microns in breadth, and half the latter figures on
the body. The mucus glands vary from 93 by 90 microns on the
tail to 60 by 40 microns on the body. This alone, without
closer inspection, would serve to generally distinguish the two
varieties of glands; but in addition the poison glands have no
lumina, the cells and nuclei are much larger than in the other
glands (mucous average about 11 microns in greatest diameter,
poison about 20 microns) and stain differently; and above all
the character of the secretion is vastly different.

As might be gathered from the name often applied to them,
the secretion in the poison glands is composed of granules.
These are of varying size, and the cells are entirely filled with
them. The mass stains from red (PI. XXIII, Fig. 31, sec.) or
reddish yellow to a dark purple in Mallory; in Van Gieson the

VOL. i.] Kstfrhj. — Poison Glands of Plethodon. 241

color is as a whole yellow with a tinge of red. In iron haema-
toxylin some granules stain (I'l. XXII, Fig. 1<S) black: but at
times one can detect in sonic granules a clear outer portion
which takes the counter stain (erythrosin, etc.), while the central
part stains dark black, and others which take only the counter
stain.

The mucous secretion, on the other hand, reacts very differ-
ently, as does the cytoplasm of the mucous cells, which can be
easily distinguished from their secretion. Here the reactions are
typically those of mucus. Mallory's stain, which colors mucus in
the sublingual of a cat a clear blue in two minutes, stains in the
same wiiy both the cells and the secretion of the small glands.
This same stain beautifully differentiates the mucus in the goblet
cells of the oesophagus and intestine of Plethodon. I have not
been able to obtain the reaction in these gland cells with thionin ,
in which Nicoglu places so much confidence as a mucous stain.
Hubbard ('03) has had the same difficulty. However, inucicar-
mine, a specific mucous stain, gives the mucous reaction after
twelve or twenty-four hours in both the glands of Plethodon and
the sublingual of the cat. The use of Van Gieson's stain clearly
differentiates the small gland from the large ones. In the former
the cells and the secretion are stained a clear red or pink,
without a trace of yellow as in the poison glands. Orcein also,
which has been described as a mucous stain, colors the cytoplasm
of the mucous gland cells a deep brown, and has absolutely no-
effect on the granular secretion of the poison glands. The iron
haematoxylin is of little use in revealing the mucous nature of the
small glands, since they take only the counter stain except for
the nuclei. These become a deep black as already said. But
this method at least serves to distinguish the two sorts of glands
aside from the nuclear staining, in that the secretion of the small
glands never takes the haematoxylin, as do the granules of the
large glands.

From these distinctions as to the primary character of the
two classes of glands, we are led to consider the histogenesis of
the secretion. It has been generally held that this process is not
the same in the mucous and poison glands. Seeck ('91), p. 55,
holds that the secretory cells are of two sorts, "solche die als.

242 University of California Publications.

Zt'lli'ii erhalten bleiben und Driisensecret secern tren (Schleim-
driisen), und andere, cleren Protoplasma sich in fVinkorniges
Driisenseeret metamorphosirt wobei die Zellen vollkommen auf-
gebracht werden, zu Gruude gehen, so dasz man ilm> in Zerfall
begriffenen Kernen in Driisensecret findcn kann (Korner-oder
Giftdriisen)." Nicoglu ('93), p. 447, finds that the cells of the
poison glands "wenn ihre Stuude gekommen ist, wandelu sie sich
in toto in Secretniasse um." But up to this time they act as
other gland cells in elaborating and retaining a secretion in their
interior, as the pancreas cells do zymogen granules. Sclmltz
('89) does not think that all the cells of a gland are destroyed at
the same time, but such as do form a part of the secretion mass
must be regenerated; indicating that they are destroyed in the
process of secretion. Drasch ('94) merely states that the poison
glands of the salamander, if completely emptied, pass entirely
away, and are replaced by new glands. Observations of the
glands at various times after emptying show regressive changes
in all the layers. Vollmer ('93) also has described the process
of solution of the Leydig cells after strong electrical stimulation
of the glands, and has made careful statements regarding the
appearance of the emptied glands. The conditions in Plethodon
almost duplicate those he has described.

It seems pretty well founded, then, that the poison gland
cells pass bodily into the secretion mass. But a distinction
should be made here, as Nicoglu has done, between the secretion
mass as that thrown out, and the secretion material, which is
the formed substance in the cells. There is no evidence of the
disintegration and solution of cells in the full but not discharged
gland. It is only when for some reason the glands are emptied
that the degenerative processes are discerned. Otherwise the
formed secretion is retained within the cells, which remain in a
normal condition at such times.

This review of the literature describes very well the processes
which go on in Plethodon; PI. XXI, Fig. 17, will show the
appearance of a gland on the day it was emptied. It has
shrunken greatly in size; as compared with others of the same
animal which, for some reason were not emptied, from three
hundred microns in diameter, say to one hundred microns. The

VOL.!.] Vxtfi-li/. — PoixoH Glands of Plethodon. . 243

nuclei which in full glands lie at the bases of the cells, are in tliis
case in the inner parts of the cells, and are larger and clearer
and in a state of disintegration. In some places only outlines
or shadows of nuclei can lie seen. Often they became shrunken
and irregular in outline when the gland is emptied. The entire
appearance of emptied glands would lead to the conclusion that
their time of functional activity is at an end.

The mucous glands, on the other hand, never reveal such
changes. It seems correct to say that the processes there are
like those in milk glands, where parts of the cell bodies ai-e
thrown oft' as secretion, while the remaining portions in time
repeat the same processes of secretion. Nussbaum ('82, p. 302;
speaks of the heads or inner portions of the mucous gland cells
of Salamander as discharged on stimulation.

If it is true then, as it seems to be, that the poison glands are
changed bodily into the mass of secretion, we must look to some
source for their replacement, if the animal is to have their con-
tinued protection. Nussbaum's conclusions should be cited here
('S2, p. 336) as bearing on the general topic of death of gland
cells through secretory activity, and their renewal. He says
secretion consists in the formation and elaboration of the mother-
substance of the secretion material, the changing of this in the
cells and in emptying the secretion when ready, out of the cells.
''Wie alles Lebende aus uns unbekannten Ursachen abstirbt
und neuen Generationen Platz macht, so gehen auch nach einer
gewissen Zeit Driisenzellen zu Grunde und werden von leben-
skriiftigen Nachbarzellen ersetzt. Sterben aller Zellen gleich-
zeitig ab, so 1st die Driise vernichtet wie eine Protozoen Colouie.
Die Secretion mag wohl die Zelle abniitzen; die Zelle
wird altern. Der Ort der Secretion ist aber nicht gleich bedeu-
tend mit Zellentod; er ist eine energische Lebensthatigang."

In this particular case of the skin glands of Amphibia, a
definite process of replacement goes on, occurring in Plethodon
in the way described for other salamanders by Nicoglu ('93),
Heidenhain ('93) and Vollmer ('93), and not as Junius ('98)
claims, by entirely new origin. The former observers find that
inside the old poison glands there lies a second smaller gland,
possessing a lumen. This small sac is to replace the older gland

244 University of California PnblirutioHs. [ZOOLOGY

and lies always between the musculature and epithelium of the
latter. Nicoglu ('93) iinds that the new glands possess "all the
epithelial parts of the old gland with the exception of the
"Schaltstiick." Whether this statement is to include also the
muscle fibres, he does not say; his figures show muscle cells
lying upon the ingrowing gland, but there is no reference to
prove that they belong to it rather than to the old gland.
However, Vollmer ('93) says that the new gland contains both
gland cells and smooth muscle fibres, which arise as does the
gland bud, from the Malpighian layer of the epidermis.

The place of origin of the replacement glands is found by
Nicoglu ('93) and Heidenhain ('93r/) in the very small, flattened
cells immediately adjoining the Schaltstiick and lying inside the
gland. Vollmer ('93) on the other hand concludes that the
place of origin of the new gland "ist das Keimlager des Rete
Malpighi. Auch die von Heidenhain erwahnten unscheinbarer
Zellenelemente, denen er die Bildung der Driisenknospe zusch-
reibt stammen vom Rete Malpighi." There is no reason, he
says, why the new glands inside the old ones should not differ-
entiate as do the first glands in the course of their development.

In Plethodon the method of renewal of the worn-out glands
is as these authors have described, but there is no evidence
showing the source of the replacement glands, and the subject
must be dismissed with the above references to the literature.

But whatever the source of the new glands, there can be no
doubt that in every old gland without exception there is a small
sac or replacement gland. This is always found in those glands
which have not been discharged (PI. XXIII, Pig. 31), as well as
in those which have been and show the most extensive degen-
erative phenomena. In this respect Plethodon seems to differ
from Triton (Vollmer '93). This author states that the growth
of the new gland is initiated when the old glands are emptied.
Nicoglu ('93) mentions the fact that the old poison glands con-
tain the smaller sacs, but does not say definitely whether or not
the destructive processes must have set in before the the new
gland makes its appearance. But in Plethedon the presence of
the replacement gland is not dependent on the secretory processes
in the large glands. The former are present in the glands of an

VOL. i.] E.itcHij. — Poison Glands of Plethodon. '24'}

animal thirty-eight mm. long which are not filled with secretion.

We have to deal then, in these cases, with the regeneration
of a gland by a gland. Individual cells are not broken down,
and then renewed by the growth of new cells as Schultz ('8!))
maintains, and as seems to be implied by Calmels ('83), who
finds that the young gland cells are indifferent elements which
may develop into either poison or mucous cells, so that a gland
may be poisonous only in part.

The question, however, as to whether a poison gland is
replaced only by a poison gland is still to be considered. May
not these be renewed by glands which to begin with are mucous
in character? That is, may not a specific poison secreting
epithelium be replaced through mucous cells, and gland by gland
instead of cell by cell ? These inquiries have been raised by
Xicoglu, and he says ('93, p. 425) that a mucous cell never goes
over into a poison cell, or vice versa, and Schultz ('89) also says
that mucous glands are always only mucous glands, and poison
glands only poison glands (p. 33), and therefrom we are to sup-
pose that the same is true of the individual cells, as he finds that
cells replace cells.

Still the evidence gained by a study of the poison glands of
Plethodon indicates rather strongly that we have to deal with
<i production of poison glands from mucous glands fiilirely.
Nicoglu has already shown that in Triton a mucous gland
may sometimes replace a poison gland entirely, but he
very strongly opposes the idea that the function of such a gland
ever changes. He holds (p. 435) that the condition of mucus
within poison gland is a functional adaptation, because the
animal needs more mucous glands than are on hand. Everything
goes to show that in Plethodou, on the other hand, the occasional
method of regeneration described by Nicoglu is the only one.
The replacement glands already described stain blue without
exception in Mallory, which has been shown to be a mucous stain.
The contrast between the blue of the mucus and the red of the
granular secretion is very sharp (PI. XXIII, Fig. 31). The
mucous reactions described for Van Gieson, orcein and mucicar-
inine, are shown invariably in the replacement glands as in the
mucous glands outside, and the correspondence of the replacement

'J-H> University of Cnl-ifm-niii I'tiliHriilions. [ZOOLOGY

glands stained in iron haematoxylin with the other mucous glands
is just as complete (compare I'l. XX, Figs. 2 and 3).

There can be nothing clearer than the reaction of the new-
glands to ATallory's stain. The blue color is present in every case
as shown by a study of hundreds of glands. In the very large
glands on the back of the tail the ingrowing glands never reach be-
yond a certain size, such as is shown in PI. XXIIT, Fig. 31. This
may possibly be due to some effect of the poison which would
hinder the growth of the small gland, or, as seems more likely,
the new gland does not develop because it is hemmed in and
hindered in its growth by the pressure of the large amount of
secretion in the old gland. Drasch ('94) has made this sugges-
tion previously, bnt does not say where the replacement glands
are located. But in all the small poison glands which lie along
the sides of the tail and also on the dorsal and ventral surfaces,
particularly in the constriction, can be seen all stages of devel-
opment of the mucous glands within them, from the small buds to
new glands which have almost entirely replaced the old ones.
The small poison glands differ from the largest ones in no other
respect than in size, and for that reason it seems fair to conclude
that the processes of regeneration going on in them are charac-
teristic, and typical of those believed to occur under certain
circumstances in the large glands of Plethodon, and as observed
in other salamanders. There are many cases to be seen in
Plethodon in which some glands are so far replaced by a new
mucous gland that only a faint crescent of granular secretion can
be seen, the rest of the contents being mucus. In other cases"
the amount of granular material is a little greater, and in still
others we may see the gland half granular and half mucus
(PI. XX, Fig. 3; PI, XXII, Figs. 18, 19, 20). In all these the
granular portions stain as do the same parts of the large glands,
while the remainder reacts to Mallory and the other stains as do
the small sacs in the large glands and the mucous glands outside
of these.

To sum up the foregoing we may say that the small glands
within the large ones react like known mucous glands to Mallory 's
stain and mucicarmine, and in the same way so far as the nuclei
of the replacement and mucous glands of the tail are concerned,

VOL. i.] Eslrrly. — PO'IXOH Blonds of Plethodon, 247

to iron haematoxylin. That is, both the mucous cells and those
of the replacement glands stain blue in Mallory. red or pink in
Van Gieson, and both have a fibrillar structure. The mucous
reaction is also given with inucicarinine. And finally, the nuclei
of the ingrowing gland fundaments always stain intensely black
in iron haematoxylin, as do the nuclei of the mucous glands.

The facts just related have been gained entirely from a stud\
of preparations made from material taken from unstimulated
animals, that is those not irritated prior to immersion in killing
fluids. The evidence along this line is stronger and more con-
vincing in the case of an animal which, without stimulation of any
kind other than such as might have occurred in nature, got rid
of a great deal of the secretion in the glands of the tail and then
cast that organ off, as if it could be of 110 further use. The
animal in question, when first observed, was seen to be entangled
head down between some pieces of bark in the terrarium in
which it was confined. This seemed to irritate the salamander
very much, for when it freed itself it began moving quickly
about, swinging its tail from side to side like an angry cat. The
tail, during this time, became covered with a very abundant
white secretion, After about five minutes of such behavior on
the part of the animal, when I merely touched the tail it was
suddenly thrown off, the break being in the constriction back of
the cloaca.

The tail was put into Zenker's fluid after about fifteen min-
utes, and sections made later. Here the likeness between the
fundaments in the empty poison glands and the mucous glands
could not be more complete. In all the stains used the appear-
ances are exactly the same. The cells of the mucous glands are
much higher than in other animals seen, stain a lighter blue in
Mallory, and have a vesicular structure approaching granular,
rather than the filar structure usually seen. Even so, the
replacements glands cells are their exact counterparts, and show
the same reactions to Van Gieson, mucicarmine, and iron haema-
toxylin, as well as Mallory 's stain.

It seems hardly possible that the cells of the mucous glands
could have so changed their structure and appearance in fifteen
or twenty minutes, though the increase in height and consequent

248 University of California Publications. [ZOOLOGY

diminution in size of the lumen of the gland, together with the
vesicular structure of the cells, would lead one to think that they
are in the way of becoming granule or poison cells. But what-
ever the interpretation put upon this appearance, and to what-
ever source it is due, it must be admitted that the fundaments in
the old poison glands have undergone the same processes and
their histological characters are now exactly similar to those of
the mucous glands.

Further evidence that the glands are originally all of the
same character may be gained from the literature. Ancel ('02),
who has followed very closely the development of the skin glands
in salamander, considers that the large glands represent organs
more completely differentiated than the small glands toward a
special functional adaptation, though both in early development
are absolutely alike (pp. 269, 283.) Jiuiius ('98) believes
that there is but one kind of gland in the skin of the frog and
probably of all Amphibia, and that the various glands of the
authors are young and old forms or developing stages of them.
He says further that in the frog he has not seen the regeneration
described by Vollmer and Nicoglu, and declares that atrophied
glands are replaced by wholly new ones developed by down-
growths of epiderm cells into the cutis. According to him,
small glands represent young stages of large ones, and the for-
mer are equivalent to the non-contractile or mucous glands, while
the latter are the dark, contractile, granule or poison glands.

Again, Hoyer ('90, p. 354) finds that in some poison glands
of the salamander single cells or groups of cells lying between
the non-staining large granular cells take on a red-violet color
in thionin (which he employs as a specific mucous stain). He
makes the suggestion merely: "Moglicher Weise deutet dieses
eigenthiimliche Verhalten auf eine genetische Beziehung der in
den Driisenzellen enthaltenen mueinahnlichen Substanz zu dem
giftige Secrete." And finally, the observation of Phisalix-Picot
('00) that the secretion of the mucous glands of the Salamander
is poison, seems to me to bear along this line of a correlation
between the so-called mucous glands and the poison glands.

Evidence in this direction also, further than that already
advanced seems to be indicated in the poison glands of Plethodon.

VOL. i.] Etttfrly. — Poison Glands of Plethodon, 24!)

Here there is very frequently a distinct blue tinge to the <jni»nl<n-
secretion. This may possibly be because the metamorphosis
from ''a mucus-like substance to the poison secretion" is not
entirely completed. At any rate one is impressed with the like-
lihood that there is mucous material in the poison glands outside
of that contained in the replacement glands.

In the discussion of the replacement of the poison glands by
those of the mucus variety, it has been shown that every large
gland has within it the fundament of a new gland which to all
stains for mucus except thionin gives the mucous reaction, and
which is also the exact counterpart of the small glands having
the mucous secretion. The fact that only in poison glands of
smaller size are found evidences that they are entirely replaced
by mucous glands, maybe explained on the ground that there the
amount of granular secretion is not sufficient to mechanically
hinder the growth of the new replacement gland. The actual
transition stages from mucous to granular secretion have not been
observed in my material.

If we make the assumption in view of these facts that the
glands of mucous character in the poison glands develop only into
mucous glands on the death of the latter, we are forced to one of
two conclusions: either that the small glands outside the large
ones, especially in Plethodon on the dorsal surface of the tail,
become the poison glands, or, on the other hand, that when the
latter are once destroyed there is no return to such struct-
ure except by developing anew according to the embryonic
type.

The latter process is going on continually in large as well as
in small animals, as can be readily seen by inspection of sections.
Hut it seems that the fundaments are all alike to begin with
(Ancel '02) ; as this author says, the solid gland buds in which
a cavity is formed do not undergo further important morpho-
logical transformations, and constitute the mucous glands. Those
which remain solid, however, continue their development in other
ways and form the poison glands (p. 269) . It seems to me that
this is equivalent to saying that in embryological development
the poison glands pass through a mucous stage to reach their final
form and character. It certainly lends evidence to the view

250 University of California l*ubrH-<iii<>ns. [ZOOLOGY

expressed, that the glands which are to replace the worn-out
poison glands are originally mucus in character.

There is no reason to believe, however, that the replacement
glands are functionless during the life of the poison glands in
which they lie. Even the smallest replacement glands have
distinct ducts and epithelium, and in some cases it is absolutely
certain that they have elaborated a secretion similar in every
respect to that of the mucous glands.

It is very probable that under all ordinary conditions the small
glands in the large ones secrete mucus, and in thi> sense art-
adaptations; not because the animal through some unusual
external conditions has come to need more mucous ghinds as
Nicoglu ('93) says, but rather because under normal environ-
ment there is always need of more mucus than can be secreted
by the glands outside the poison glands, especially when the
latter are so closely crowded together as on the back of the tail
in Plethodon. And much evidence goes to show that under
stress of necessity such mucus secreting glands become by
replacement the more highly specialized poison glands and take
on a particular function, that of forming a substance protecting
the animal from its enemies (Hubbard '03.)

The nerve supply of the skin of Amphibia has been a favorite
subject of study for many years. Most investigators have limited
themselves to the terminations in the sense organs of the skin
and in or on the ordinary epidermal cells (Pfltzner '82; Cauini
'83; Preukel '86; Massie '94; Herrick and Coghill '98; Coghill
'99) . The innervation of the glands has received less attention.

Eckhard ('49) first showed that the glands could be emptied
by stimulating the anterior roots of the cerebro-spinal nerves,
but did not consider the structure of the nerve endings. Eberth
('69) found that there is a network of very fine fibres close upon
the glands; Englemann ('72) came to the same conclusion and
showed farther that from the nerves about the gland fine twigs
are given off to the contractile cells. Openschowski ('82)
describes a network of nerves surrounding the glands, as well
as an iutracellular net; but from his figures it is hard to believe
that the structures he shows are nerves. Drasch ('89) also experi-
mentally proved the efficacy of nerve stimulation iu obtaining

VOL. i.] l-lxferly. — Poimm Qlands of Plfthodon. '1~>\

from the glands, as docs I'hisalix-I'icot ('00). Ebertli
and Hunge ('92) liave described free nei-ve fibres which seem to
end with knobs outside the epithelium of the ball of the thumb
of the male frog- Loeb ('9(5) has also shown how closely the
glands of Amblystoma are connected with the central nervous
system. In 189S Herrick and Coghill were able to show the
existence an intimate connection of nerve fibres with the walls
of the glands, but were unable to discover the exact relation
of the fibres to the gland cells. They also described the
plexus of nerves beneath the corium as being composed of two
sorts of fibres; larger ones connected with the nerve bundles
of the central system, and smaller ones which in part, at least,
originate in ganglion cells in the corium. Sehuberg ('03) has
criticised the results of these authors, contending that many or all
of the nerve bundles described are really connective tissue
bundles, and that the ganglion cells are the "Mastzellen" he
himself figures.

^lassie ('!)!)) continuing the work of Herrick and Coghill,
(•(insiders the same arrangement of fibres beneath the corium,
and also shows that nerves end on the muscles of the "ental"
glands. He finds that nerve fibres passing from the nerve bundle
plexus under the corium are intimately connected with the ental
glands, and seem distinct from the nerves supplying the muscles.
"It seems, therefore, that there are two groups of nerves passing
to the glands of the ental series; the one attaching by the typical
endings to the enveloping muscle cells, the other ramifying
promiscuously over the surface of the gland." (p. 59.)

In the study of the nerves of the poison glands of Plethodon,
three methods have been relied upon; namely, the silver nitrate-
pyrogallic acid method of Cajal, and Mallory's phosphotungstic
acid haematoxylin and fuchsin-orauge G-anilin blue methods.
The last named gave most excellent results, while of the other
two Cajal's was only indifferently successful.

The haematoxylin of Mallory stains only the sheaths of the
nerves and so it is of no value in tracing the axis cylinders, since,
as is well known, the nerves lose the medullary sheaths on pass-
ing into the corium. Beneath the corium, however, the nerves
can readily be followed by this method. In some instances fibres

2.V2 ['Hirerxity of California Publications. [ZOOI/IOY

are shown rniuiing for long distances beneath the corium, and
branches can even be seen to turn toward the epidermis, but all
traces of them arc lost as soon as they enter the corium.

The other method of Mallory gives like results as far as the
distribution of the nerves beneath the corium is concerned. In
cross sections of the tail it is often possible to trace a fibre from
the roots leaving the cord out to the corium. Sometimes this
may be seen in one section; in many cases two or three neigh-
boring serial sections will show the same. The plexus beneath
the corium is shown best, as a whole, in frontal sections of the
tail. Here itwill be seen that the nerves are very HIIIHO'OHX, and
with the method in hand can be traced to their connections with
the cord. There can be no question as to the presence of the
nerve-bundle layer of the plexus that Herrick and his pupils have
shown; but as regards the stratum of glauglion cells, it seems
to me that Schuberg's criticism holds good. At any rate neither
of Mallory' s methods reveals such a structure, and this would at
least seem strange in view of the beautiful staining of other nerv-
ous elements. In cross sections of the tail, Mallory's fuchsin
method shows nerves running in or immediately beneath the
inner corium layer. At times several fibres are in view at once,
being, however, of different sizes.

Within the corium the distribution of the nerves to the glands
is not apparent in sections which pass through the gland, owing
to the exceedingly small size of the fibres. But when the peri-
phery of the gland is just denuded, the nervous elements are
shown very clearly. In such cases it will be seen that there is a
feltwork of many verg fine fibres closely investing the gland, end-
ing upon the muscle fibres and around the nuclei of the gland cells.

The endings upon the muscles are shown both by Cajal's
method and Mallory's fuchsin stain, and in some cases are typi-
cal (PI. XXII, Figs. 25 and 26) as described by Huber and
Dewitt ('97) and Coghill ('99) . That is, they are equipped with
terminal expansions or bulbs which lie on the muscles. In many
cases fine branching fibres can be clearly seen lying upon the
muscle layer. These pass over ultimately into the finest of
slender twigs which without terminal expansions always lie on a
muscle fibre and end there (PI. XXII, Fig. 26.)

VOL. i.] Estrrly. — Poison C/'/nds of PHJimJon. 253

The fibres in the perinnclcar endings are of much the same
character as those of the muscles. There are many instances
which are very clear of basket structure about the nuclei of the
large glands (PI. XXII, Figs. 21, 22. PI. XXIII, Fig. 30).
[ have not been able to discover connections between the fibres and
the nuclei, though in at least one ease (PL XXII, Figs. 23, 24)
I lie fibres end in knobs which lie directly on the nucleus. The
latter seems usually to be surrounded only by a basket of fine
fibres. Bethe ('94) has described three sorts of endings on cells.
Of these he finds that in the unicellular glands of the frog's
palate one frequently finds under the nucleus a small blue knob
which is connected with a fibre. The latter cannot, however, be
followed farther.

In the case of the gland cells under consideration, there can
be no dmibt that the nuclear basket is connected with nerve
fibres. That there should be a nerve supply to the gland cells,
seems evident from the experiments of Drasch ('89), Eberth
('49) and Loeb ('96) on Amphibian glands, and we have in
Plethodon histological evidence of such supply. The well-
known influence of the nervous system on the secretion of sweat,
for example, may be also mentioned in this connection. Herrick
and Coghill ('98, p. 51) have suggested the possibility of a con-
nection between the nerves enveloping the glands, and the gland
cells, but were not able to demonstrate it.

The objection may be raised that we are dealing here with
elastic instead of nerve fibres. This does not seem possible for
several reasons. The elastic fibres, as has been said, show very
little variation in size, and never, as shown by staining in orcein,
reach the excessive fineness of the nerve fibres. The branching
of the elastic fibres is much less frequent than that of the nerves,
and, in clearest distinction the former, as seen upon the glands,
take an almost uniform direction even in branching, straight
toward the epidermis, while the nerve fibres cross and recross and
branch in all directions, and the finest twigs show varicosities
which are never seen on the elastic fibres. The general effect of
the brown fibres in an orcein stain is entirely different from that
of the red ones in Mallory's stain, and leaves no doubt of the dis-
tinction here set forth between the elastic and nervous fibres.

254 University of California Publications. [ZOOLOGY

SUMMARY.

1. The skin glands of 1'lrtliodon oref/oiii-nxix, as of most
Amphibia, are of two kinds: granular and mucous. The two arc
distinguished by the character and staining reaction of their
secretions, and by other histological features, as well as by the
sizes of the glands.

2. The bodies of the large glands possess an investing mus-
culature, and in addition the ducts have both dilator and con-
strictor muscles lying in the epidermis.

3. The granule glauds are poison in character.

4. In the elaboration and ejaculation of the secretion the
poison glands are destroyed.

5. Renewal takes place by the growth into all the old glands
of a new and smaller gland, which is mucous in character. Tin-
presence of this smaller sac is not dependent upon the removal
of the secretion of the large glands, for whether this occurs or
not, the fundament giving the mucous reaction is found in all
glands; in those which show no degeneration as well as in those
where it is wide-spread.

6. The growth of the new gland is dependent upon the
removal of the secretion about it. There is evidence that even
in case the glands are hindered in their development, they still
secrete mucus. But when not hemmed in by the heavy granular
contents of the large glands they grow and take the place and
very probably assume the function of the old glands which they
replace.

7. Both musculature and epithelium of the granule glands
have a direct nerve supply. The gland cells are surrounded by a
basket work of fibres, which in some eases have terminal expan-
sions lying on the nuclei. The muscles are supplied by nerves
with typical endings of expansions or bulbs, as well as by fine
twigs without terminal expansions.

Zoological Laboratory,

University of California,
April 29, 1904.

VOL. i.] l-'xii'flij. — Poison Glands of Plethodon.

BIBLIOGRAPHY.

Ancel, P.

1902. Etude du developpement des Claudes de la peau des Batraciens et

CM particiilier de la Salamandre terrestre. Arch, de Biol., Vol.
IS, ]i)>. 257-291. Pis. VIII-IX.

Ascherson.

1S40. I'elier die Hantdriisen der Frosche. Arch. f. Anat. u. Phvsiol.
(Miiller) 1840, pp. 15-24. PI. VI.

Barfurth, D.

1891. Zur Regeneration der Gewebe. Arch. f. mikr. Anat. Bd. 37, pp.

40li-491. Pis. XXII-XXIV.

Bethe, A.

1894. Die Nervenendigungen im Gaumen and in der Zunge des Frosches.
Arch. f. mikr. Anat., Bd. 44, pp. 185-206. Pis. XII-XIII.

Biedermann, W.

1886. Xur Histologie und Physiologie der Schleimsecretion. Sitzungber.
der Wiener Akad. der Wiss., Bd. 94, 3 Abth., 2 plates. (Cited
from Centbl. f. Physiol., 1887, p. 90.)

Boulenger, G. A.

1892. The Poisonous Secretion of Batrachians. Nat. Sci., Vol. I, pp.

185-189.

Cajal, R.

1903. MSthode nouvelle pour la coloration des neuroflbrilles. Compt.

rend. Soc. Biol., T. 55, No. 13, pp. 1565-1568.

Calmels, G.

1882. Evolution de 1'epithelium des glandes a venin du Crapaud. Compt.

rend. Ac. Sci. Paris, T. 95, No. 24, pp. 1007-1009.

1883. Etude histologique des glandes a venin du Crapaud. Arch, de

Physiol., T. 1, 3 series, pp. 321-362. PI. VIII.

Canini, A.

1883. Die Endigungen der Xerven in der Haut des Froschlarvenschwanzes.
Arch. f. Anat. u. Physiol., 1883, pp. 159-160. PL IV.

Capparelli, A.

1883. Recherches sur le venin du Triton cristatus. Arch. Ital. Biol.,
T. 4, Fasc. 1, pp. 72-80.

Coghill, G. E.

1899. Nerve-termini in the Skin of the Common Frog. Jour. .Comp.
Neurol., Vol. 9, pp. 53-64. Pis. IV-V.

256 University of California Publication*. [ZOOLOGY

Cope, E. D.

1889. The Batrachi;i of North America. Bull. U. S. Nat. Mus. No. 24,
86 plates, 120 figures.

Dewitz, H.

1884. Verschiedenes Aussehen der gereizten und ungereizten Driisen im

Zehenballen des Laubfrosches. Biol. Centrbl., Bd. 3, pp.
558-560.

Drasch, O.

1889. Beobachtungen an lebenden Driisen mit und ohne Eeizung der
Nerven derselben. Arch. f. Anat. u. Physio]., Physiol. Abth.,
1889, pp. 96-137. Pis. IV- VI.

1892. Ueber die Giftdriisen des Salamanders. Verh. d. Anat. Gesel., auf
der 6 Vers. zu Wien, pp. 225-269. Pis. XII-XV.

1894. Der Ban der Giftdriisen des gefleekten Salamanders. Arch. f.
Anat. u. Physiol., Anat. Abth., 1894, pp. 225-2(59. Pis. XII-XV.

Dutarte, A.

1889. Recherches sur 1'action du venin de la Salamandre terrestre.
Compt. rend. Ac. Sei., Paris, T. 108, No. 13, pp. 683-685.

Eberth, C. J.

1869. Untersuchungen zur normalen und pathologischen Anatomie des
Froschhaut. Leipzig. (Cited by Hoffmann, Bronn's Klass. u.
Ordn; Seeck '91.)

Eberth, C. J., und Bunge, R.

1892. Die Endigungen der Nerven in der Haut des Frosches. Anato-
mische Hefte, Abth. Arbeiten, Hft. 5, pp. 175-203. PI. XI, 14
figures in text.

Eckhard, C.

1849. Ueber den Ban der Hautdriisen der Kroten und die Abhilngigkeit,
der Kntleerung ihres Sekretes vom oentralen Nervensystem.
Arch. f. Anat. u. Physiol., 1849, pp. 425-428.

Engelmann, T. W.

1871. Ueber das Vorkommen und Innervation von contractilen Driisenzel-

len in der Froschhaut. Pfliiger's Archiv., Bd. 4, pp. 1-3, '72.

1872. Die Hautdrusen des Frosches, Ibid., Bd. 5, pp. 498-538, PI. VII.

Fraisse, P.

1885. Die Regeneration von Gesveben und Organen. Kassel u. Berlin,

1885, 164 pages, 3 plates. (Cited by Vollmer, 1893, p. 406.)

Frenkel, S.

1886. Nerv und Epithel am Froschlarvenschwanz. Arch. f. Anat. u.

Physiol. 1886, pp. 415-430. PI. XIII.

VOL. i.] Esirrly. — /W.w/i Qla/nds of Plethodon. 257

Gratiolet, P.— Cloez, S.

1851. Notes sur les propriete's vSne'neuses de 1'humeur lactescente que

seVretent les pustules pittances de la Salamandre terrestre et du
Crapaud commun. Compt. rend. Ac. Sci. Paris, T. 32, pp.
692-695.

1852. Nouvelles observations sur le venin content! dans les pustules

cutanSes des Batracims. Ibid, T. 34, pp. 729-732.

Heidenhein, M.

1893d. Die Hautdriisen der Amphibien. Sitzungsber. d. VViirzb. physik.
med. Gesel. No. 4, p. 52. (Cited by Vollmer, '93, p. 407.)

18936. Ueber das Vorkoramen von Intercellularbriicken zwischen glatten
Muskelzellen und Epithelzellen des ailsseren Keimblattes und
deren theoretische Bedeutung. Anat. Anz. Jahrg. 8, pp. 404-
410. 1 figure in text.

Hensche, A.

1856. Ueber die Driisen und glatten Muskeln in der aiisserem Haut von
Rana temporaria. Zeit. f. wiss. Zool., Bd. 7, pp. 273-282.

Herrick, C. L., and Coghill, C. E.

1898. The Somatic Equilibrium and the Nerve-endings in the Skin.
Journ. Comp. Neurol., Vol. 8, pp. 32-57. Pis. V-IX.

Hoffmann, C. K.

Amphibien. Bronn's Klassen und Ordnungen, Bd. 6, Abth. 2, pp. 347-
367.

Hover, H.

1890. Ueber den Nachweis des Mucins in Geweben mittels der Farbe-
method. Arch. f. mikr. Anat., Bd. 36, pp. 310-374.

Hubbard, M. E.

1903. Correlated Protective Devices in Some California Salamanders.
Univ. of California Pubs., Zoology, Vol. 1, No. 4, pp. 157-168.
PI. XVI.

Huber, G. C., and Dewitt, Mrs. L. M. H.

1898. A Contribution to the Motor Nerve-endings and on the Nerve-end-
ings in the Muscle Spindles. Jour. Comp. Neurol., Vol. 7.
pp. 169-230. Pis. XIV-XVIII.

Junius, P.

1898. Ueber die Hautdriisen des Frosches. Arch. f. mikr. Anat., Bd. 47,
pp. 136-154. PI. X.

Kallius, E.

1896. Endigungen sensibler Nerven bei Wirbelthieren. Ergebn. d.
Anat. u. Entwicklungsgesch. (Merkel-Bonnet), Bd. 5, pp. 507-
561.

'J")S University of California Publications. [ZOOI.OM

Leydig, F.

1876a. Ueber die allgemeine Bedeekungen der Amphibieu. Arch, f. mikr.

Anat. Bd. 12, pp. 119-241.
18766. Die Hautdecke mid HautsinBe8organe der 1'rodelen. Morph.

•liihrb., Bd. 2, pp. 287-318. Pis. XVIII-XXI.

1892. Zum Integument niederer Wirbelthiere abernmls. Biolog. Centbl.,

Bd. 12, pp. 444-467.
List, J. H.

1889. Ueber den feineren Ban Schleim secernirender Driisenzellen nobst

Bermerkungen uber dem Secretionprocess. Anat. Anz., Jain';,'.

4, pp. 84-94.
Loeb, J.

1896. Sur Theorie des Galvsnotropiwnng, III. Ueber die polare Erre-

gung der Hautdriisen von Amblystoma durch den constanten

Strom. Pfliiger's Ai-chiv., Bd. 65, pp. 308-316. 12 (inures

in toxt.

Macallum, A. B.

1886. Nerve-endings in the Cutaneous Epithelium of the Tadpole. Quart.
Jour. Micr. Sci., Vol. 26, pp. 53-70. PI. VI.

Mallory, F. B.

1900. A Contribution to Staining Methods. Jour. Exper. Med., Vol. 5,
No. 1, pp. 15-20.

Massie, J. H.

1894. Glands and Nerve-endings in the Skin of the Tadpole. Jour.
Comp. Neurol., Vol. 4, pp. 7-12.

Nicoglu, P.

1893. Ueber die Hautdriisen der Amphibien. Zeit. f. wiss. Zool., Bd.

56, pp. 409-487. Pis. XXI-XXIII.

Nussbaum, M.

1882. Ueber den Thatigkeit der Driisen. Arch. f. mikr. Anat., Bd. 21,-
pp. 296-351. Pis. XV-XVIII.

Openchowski, Th.

1882. Histologisches zur Innervation der Driisen. Pfliiger's Archiv., Bd.

37, pp. 223-233. PI. VI.
Paulicki.

1885. Ueber die Haut des Axolotls. Arch. f. mikr. Anat., Bd. 24, pp.
120-171. Pis. VIII-IX.

Pfitzner, Wilh.
1880. Epidermis des Salamandres. Morph. Jahrb., Bd. (i, pp. 4()0-f>L>(i.

Pis. XXIV-XXV. Also Journ. R. Micr. Soc. (2), Vol., pt. 1,

pp. 218-224.
1882. Nervendigungen ira Epithel. Morph. Jahrb., Bd. 7, pp. 726-745.

1 plate.

VOL. i.] Kxli'rly. — 1'iiixoii Gttmds of Plethodon. _•">!>

Phisalix, C.

1889. Nouvellcs experiences sur le vcnin de hi Siilamandre terrestre.
Coiupt. rt-nil. Ac. Sci. Paris, T. 109, pp. 4(15-407.

Phisalix and Langlois.

1889. Action physiologique du venin de la Salamandre terrestre. Compt.
rend. Ac. Sci. I'aris, T. 109, pp. 482-48,').

Phisalix-Picot.

1900. Recherches embryologiques, histologiques et physiologiques sur les
glandcs a vcnin de la Salamandre terrestre. These de doctoral
en me'decine. Paris, 1900.
Roeber, H.

1869. Ueber das elektromotorische Verhalten der Froschhaut bei Reizung
ihrer Nerven. Areh. f. Anat. u. Physiol. 1869, pp. 633-649.

Schuberg, A.
1903. Untersuchungen iiber Zellverbindungen. Zeit. f. wiss. Zool., Bd.

74, Hf. 2, pp. 156-325. Pis. IX-XV.
Schultz, P.

1889. Ueber die Giftdriisen der Kroten und Salamander. Arch. f. mikr.
Anat., Bd. 34, pp. 11-57. PI, II.

Schultze, F. E.

1867. Epithel— und Driisenzellen. Arch. f. mikr. Anat., Bd. 3, pp. 137-

202. Pis. VI-XII.
Seeck, O.

1891. Ueber die Hautdriisen einiger Amphibien. Inaug.-Diss. Dorpat.

Stieda, L.

1865. Ueber den Bau der Haut des Frosches. Arch. f. Anat. u. Phys.,

1865, pp. 52-67. PI. I.

Szczesny, O.

1867. Beitrage zur Kenntniss der Textur der Froschhaut. Inaug.-Diss.,

Dorpat. (Cited by Seeck, '91.)
Tonkoff, W.

1900. Ueber die elastischen Fasern in der Froschhaut. Arch. f. mikr.

Anat., Bd. 57, pp. 95-101. PI. VII.
Vollmer, E.

1893. Bin Beitrag zur Lehre von den Regeneration speciel der Haut-
driisen der Amphibien. Areh. f. mikr. Anat., Bd. 42, pp. 405-
423. Pis. XXIV-XXV.
Weiss, O.

1899. Ueber die Hautdrusen von Bufo cinereus. Arch. £. mikr. Anat.,
Bd. 53, pp. 385-396. 3 figures.

Wiedersheim, R.

1886. Lehrbuch der vergleichenden Anatomie, p. 25.

Zalesky, S.

1866. Ueber das Salamandarin. Das Gift der Salamandra maculata.

Med. Chem. Unters, Bd. J 1. (Hoppe-Saeyler) Berlin, '66, pp.
85-116. (Cited by Hoffmann Klass. u. Ordn.)

260 rith'frsity of California Publications. [ZOOLOGY

LIST OF ABBREVIATIONS USED IN THE PLATES.

con.m. — constrictor muscle fibres.

c.t.b. — connective tissue bundles.

c.t.l. — connective tissue layer of gland walls.

c.t. ---connective tissue in epidermis.

e.w. — cell walls.

d. — duct of gland.

dil.m. — dilator muscle fibres.

el.f. — elastic fibres.

ep. — epidermis.

ep.m.c. — cell containing musculature of duct.

fl.c. — funnel cell.

i.e.l. — inner layer of corium.

l.d.— lumen of duct.

l.gl. — lumen of gland.

m.b. — muscle bundles.

m.e.l. — middle layer of the corium.

m.f. — muscle fibres.

m.gl. — mucous gland.

m.n. — muscle nucleus.

n.c. — nerve cord.

n.e. — nerve endings.

n.fl.c. — nucleus of funnel cell.

n.f. — nerve fibre.

nuc.m.c. — nucleus of mucous cell.

nuc.p.c. — nucleus of poison cell.

nue.eji.m.c. — nucleus of muscle cell in epidermis.

o.c.l. — outer layer of corium.

p.fl.e. — processes of funnel cells.

p.gl. — poison glands.

pig.— pigment.

prol.m.f. — prolongations of muscles into epidermis.

rep.c. — replacement cell (and nucleus).

rep.gl. — replacement glands.

sec. — secretion.

All the figures were drawn with the Abb6 camera lucida.

ERRATA

P. 251, 1. 31: For fuchsin-orange G-anilin blue,
read fuchsiu-orangeG- anilin blue.

P. 259, under Zalesky 1866: For Bd. J 1, read Bd. 1 ; for Saeyler, read Seyler.
P. 264, under description of Fig. 16: For X 1650, read X 825.
I'. LliiG, under description of Fies. 21, etc.: For X 1850 «•<•/•-• '

PLATE XX.

Fig. 1. — Cross section of entire tail, showing position on dorsum of large
poison glands (p. (/I.) and the mucous glands (m.gl.) chiefly on the
ventral side. Diagrammatic except in outlines and proportions
of parts. Van Gieson.X 22

Fig. 2. — Mucus gland from ventral side of tail, showing large lumen (l.gl.),
and dark staining, angular nuclei (nuc.m.c.). Lower part of
funnel cell (fl.e.) shown in epidermis (ep) which is not repro-
duced entire. Benda's iron haematoxylin. X 342

Fig. 3. — Poison gland (p.gl.) of small size partly replaced by ingrowing
mucous gland (m.gl.). Funnel cell (fl.c.) shown in epidermis
(ep.); nuclei (nuc.m.c.) of mucous gland darkly stained as in
Fig. 2. Benda's iron haematoxylin. X 342

Fig. 4. — Outline drawing of cross section of duct of poison gland showing
replacement cells of the funnel (rep.c.) rolled one within the
other, the funnel cell (fl.c.) and the lumen of the duct (l.d.)
Mallory's conn, tissue stain. X 875

Fig. 5. — Portion of lower part of poison gland showing bundles of connec-
tive tissue (c.t.b.) passing from the inner layer of the corium
(i.e./.) to the connective tissue layer of the wall of the gland
(e.t.l.) Nuclei (nue.p.c.) and walls (e.tc.) of gland cells. Secre-
tion not shown in detail. Mallory's conn, tissue stain. X 342

Fig. 6. — One side of median longitudinal section of duct of poison gland
showing muscle fibre (»»./.) and its nucleus (m.n.) and the pro-
longation of the fibre (prol.m.f.) into the epidermis (cp.).
Compare with PI. IV, Fig. 27. Mallory's conn, tissue stain.
X 280

Fig. 7. — Branching muscle fibres (m.f.) from lower part of gland. Mal-
lory's conn. tiss. stain. X 342

Fig. 8. — Longitudinal section of poison gland through the mouth showing
two expansions of muscles (m.f.} in which the nuclei lie, and
portions of muscle fibres. Nucleus of funnel cell (n.JI.c.) at
duct (d.). Secretion of gland not shown. Ferric-chloride
haematoxylin. X 342

[262]

BULL. DEFT. ZDDL.UNIY CAL.VDL.I

] PLATE XX

. >•-.:-• .-. ".u. .

PLATE XXI.

Fig. 9. — Elastic fibres (el.f.) on surface of gland. Gland wall (c.t.l.) in
section indicated; also nuclei of gland cells. The elastic fibres
pass through the inner layer of the corium (i.c.L). Tanxcr's
oreein. X 342

Fig. 10. — Section through upper pole of gland at one side of the duct, show-
ing cut ends of muscle fibres (»«./.) and their nuclei (m.n.).
From cross section of tail. The nuclei in this figure correspond
in position to that shown in PI. XX, Fig. 6, and to the enlarge-
ment of the fibres shown in Fig. 8. Mallory's conn, tissue stain.
X342

Fig. 11. — Tangential section through wall of gland and the mouth, from
frontal section of tail. Muscle fibres (»(./.) and nuclei (/«.«.)
shown. Funnel cell(./Le.) lining duct and some secretion (sec.)
in lumen of duct (l.d.). Mallory's conn, tissue stain. X 342.

Fig. 12. — Cross section of gland from frontal section of tail, at level of
muscle nuclei (m.n.). Compare with Figs. 10 and 11. Van
Gieron's stain. X 400

Fig. 13. — Cross section of epidermis at upper pole of poison gland, showing
deep lying epidermal cell (ep.m.c.) which contains the con-
strictor and dilator muscles of the duct. Mallory's conn, tissue

X342

Figs. 14 and 15. — Cross sections of ducts at level of cell described in Fig. 13,
showing constrictor (con.m.) and dilator muscles (dil.m.). In
Fig. 14, only the outer ends of the constrictor fibre appear. In
both figures are shown the ends of the muscle fibres (»«./.) of
the glands, in the epidermis, and the connective tissue (c.t.)
outside the nuscles. The nucleus of the epidermal muscle cell
is shown in Fig. 14. Mallory's conn, tissue stain. X 875

Fig. 16. — Description as for Figs. 14 and 15. But one set of constrictor
fibres shown; lumen of duct (l.d.) nearly closed. Mallory's
conn, tissue stain. X 1650

Fig. 17. — Longitudinal section of nearly empty poison gland. Secretion
(sec.) very small in amount, cell walls (c.w.) distinct, nuclei
clear and of irregular shapes. Semi-diagrammatic in unim-
portant details. Benda's iron haematoxylin. X 34^

BULL. DEFT ZQDL.UNIV CAL.VDL!.

[ESTEHLV] PLATE XXI.

16

C D E DEL

tis^T1

Siv

7/>>

PLATE XXII.

Figs. 18, 19, 20. — Stages in replacement of small poison gland (p.gl.) by
mucous glands (m.gl.) from sides of tail. Mucous nuclei dark.
Secretion (sec.) shown in poison part only. Benda's iron-
haematoxylin. > 342

Figs. 21, '22, 23, 24. —Tangential sections of poison glands, showing nerve
endings (ii.e.) on nuclei of poison cells (nuc.p.c.). Mallory's
conn, tissue stain. Figs. 21, 22, 24. X 1850. Fig. 23. X 875

Figs. 25 and 26. — Tangential section of wall of poison glands, showing nerve
endings (n.e.) on muscles (»»./.)• Fig. 25, Mallory's conn,
tissue stain. Fig. 26, Cajal's silver nitrate-pyrogallic acid.

XH75

[2661

BULL. DEFT Z DDL. UN iv CAL.VDL!.

[ESTERLY] PLATE XXII

L'l

23

\ nuc.pr.

PHOTO -LITH.HRITTaN &RITi:a

PLATE XXIII.

Fig. '27. — Median longitudinal section of duet of poison gland, showing pro-
longation of funnel cell (p.Jt.e.), prolongation of muscles
(l>rol.m.f.) into the epidermis, and the c-onnective tissue (c.t.)
outside them. Replacement cells (rep.c.) shown with processes
extending down as far as funnel cell. Mallory's conn, tissue
stain. X 1650

Figs. 28, 29, 30. — Cross sections of ducts, showing funnel cells (/.<•.), gland
muscles (prnl.m.f.), connective tissue (c.t.) at sides of duct (</.),
and constrictor and dilator muscles (con. in, dil.m.). Mallory's
conn, tissue stain. * 1650

Fig. 31. — Longitudinal section of poison gland, showing small mucus gland
(m.gl.) inside it. Large gland 440 microns by 180 microns;
small gland 90 microns by 43 microns. Mallory's conn, tissue
stain. X 1650

Fig. ',12. — Nerve-endings (n.e.) about nucleus of poison cell (WHO.;;. c.). Mal-
lory's conn, tissue stain. X 1650

[268]

BDLL.DEPT.ZDDL.UNIV CAL.VDLI.

,

(tit m. S

\

Tig. 28

/„•,,/',„/:

I''

•

» — ^^ -* ~m * •• » •"» "

[ESTERLY] PLATE 'XXIII.

Fig-

cttf.m.

•— con.ni

Fig. 30

PHOTC-l-ITH BRITT

UNIVERSITY OF CALIFORNIA PUBLICATIONS-iCONTlNUED,

The following series in Graeco-Roman Archaeology, Egyptian Archaeology, Ameri-
can Archaeology and Ethnology and Anthropological Memoirs arn publications from the
Department of Anthropology:

GRAECO-ROMAN ARCHAEOLOGY.

Vol. 1. The Tebtunis Papyri, Part I. Edited by Bernard P. Grenfell, Arthur
S. Hunt, and'J. Gilbart Snn :• -00, Plates 9, 1903
, . . Price, $1'

Vol. 2. The Tebtunis Papyri, Part 2 (in preparat;<

EGYPTIAN ARCHAEOLOGY.

Vol. 1. The Hearst 'Medical Papyrus. Edited ; Reisner a;..

Lythgoe (in press).

AMERICAN ARCHAEOLOGY AND ETHNOLOGY.

Vol. 1. No. 1. Life and Culture of the Hupa, by Pliny Earle Goddard.

Pages 88, Plates 30, September, ! . . Price, 1.25

No. 2. Hupa Texts, by Pliny Earle Goddard. Pages 290, March,

1904. . . . Price, 3.00

Vol. 2. No. 1. The Exploration of the Potter Creek Cave, by William J.

Sinclair. Pages 27, Plates 14, April, 1! . Price, .40

No. 2. The Languages of the Coast of California South of San

Francisco, by A. L. Kroeber. Pages 72, June, 1904. Price, .60
No. 3. Types of Indian Culture in California, by A. L. Kroeber.

Pages 22, June, 1904 .' Price, .25

ANTHROPOLOGICAL MEMOIRS.

Vol. 1. Explorations in Peru, by Max Link PI

No. 1. The Ruins of Moche.
No. 2. Huamachttco, Chincha, lea.
No. 3. The Inca Buildings of the Valley of Pisco.

PHYSIOLOGY.— Jacques Loeb, Editor. Price per volume $2.00. Volume '
progress):

No. 12. On the Production and Suppression of Gl . in Rabbits through

Electrolytes (second communication rtin H. Fischer.

No. 13. The Infuence of Saline Purgatives on Loops of Intestine removed-, ,
from the Bodv. Bruce MacCallum. /

one

No. 14. The Secretion of Sugar into the Intestine caused by Intravenous I
Saline Infusions, by John Bnu M I allum.

No. 15. On the Influence of the Reaction of the Sea-Water on the Regenera-
tion and G'.owth of Ttibularians, by Jacques 1 <•

No. 16. The Possible Influence of the Amphoteric Reaction of Certain <,

Colloids upon the Sign of their Electrical Charge in the Pres- I ,
ence of Acid and Alkalis, by Jacques Loeb.

No. 17. Concerning Dynamic Conditions • contribute toward the { cover
Determination of the Morphological I'v -;UP \

! first communicatioi'

No. 18. The Action of Cascara Sagr;>,da la preliminary cotnmunicat1
John Bruce MacCalh

No. 19. Artificial Parthenogenesis and Regular Segmentation in an Annelid
(Ophelia-, by G. Bullot.

UNIVERSITY OF CALIFORNIA PUBLICATIONS-(CONTINUED)

PHYSIOLOGY.— Continued.

No. 20. On the Action of Saline Purgatives in Rabbits and the Counter-^
action of their Effect-by Calcium, (second communication), by
John Bruce MacCallum.

No. 21. On the Local Application of Solutions of Saline Purgatives to the
Peritoneal Surfaces of the Intestine, hy John Bruce MacCallum.

In
one

PATHOLOGY. u. Taylor, Editor. Price per volume $2.00. Volume 1

proj

No. 1 . On the Quantitative Separation of the Globulins of Hemolytic Serum,
; Special Reference to the Carbon Dioxide Group, by
Clarence Quinan.

No. 2. Hydrolysis of Protamine with Especial Reference to the Action of

Trypsin, by Alonzo .'Engleben

No. 3. On the Synthesis of Fat Through the Reversed Action of a Fat- ]
Splitting Enzyme, by Alonzo Englebert Taylor.

No. 4. On the Occurrence of Amido-Acids in Degenerated Tissues, by
Alonzo Englebert Taylor.

No. 5. On the Autolysis of Protein, by Alonzo E, Taylor.

No. 6. On the Reversion of Tryptic Digestion, by Alonzo E. Taylor.

In

one

cover.

v in -;

> one
J cover.

CLASSICAL PHILOLOGY.— Edward B. Clapp, Editor. Price per volume $2.00.
Volume I (in progress) :

!. Hiatus in Greek Melic Poetry, by Edward B. Clapp.

ASTRONOMY. — W. W. Campbell, Editor.

Publications of the Lick Observatory.— Volumes I-V completed. Volume
VI (iri progres-

No. 1. A Short Method of Determining Orbits frorri Three Observations,
A. O. Leuschner.

, 2. Elements of As - > iA, by A. O. Leuschner and Adelaide

M. Hobe.

No. 3. Preliminary Elements of comet 1900 III, by R. H. Curtiss and
G. Dall.

Contributions from the Lick Observatory,— Nos. I-V.

Lick Observatory Bulletins.— Volume I (pp. 193) completed. Volume II

;ss).

AGRICULTURE.— Bulletins of the Department of Agriculture. Nos. 1-146.

Biennial Reports of the Department of Agriculture.

Publications of this Department are sent gratis to citizens of the State of California.
For detailed information regarding" them address Professor E. W. Hilgard, University
of California, Berkel - * ' . : ..,• •.

UNIVERSITY CHRONICLE.— An official record of University life, issued quarterly,
edited by a committee of the faculty. Price, St'.OO per year. Current
volume No. VI. __

Address all orders, or requests for information concerning the above publications
(except Agricultural) to The University Press, Berkeley, California.

U C BERKELEY LIBRARIES

NON-CIRCULATING BOOK

UNIVERSIT-
5m-9,'26