etd cs

roves

— +t
3 niienernese

SITES

,
ae Lo
. or 7
were pres a Ses
eneteerat,
ss

oe
RS Oh en a acre.
Ses

Pitt eyeias
Agi
+

ra

Bini
i ; Hote :
< tie

Sorte
oe
rt

if

es
= ys Paes i
sete ti tates

\

~ ‘
G s.

bs 3 We =
<, <i iS
Peis

‘BY MEMBERS OF THE UNIVERSITY OF KANSAS.

" LAWRENCE, KANSAS, 7 :
1911. ee: ik womens og

Noe
ae

L

CONTENTS OF VOLUME V.

No. 1.—The Dakota-Permian Contact in Northern Kansas... F. C. Greene.

Further Notes on the Pueblo Ruins of Scott County, H. T. Martin.
Influence of Magnesium Sulphate on the Motor Cells of

the Cerebral Cortex.... H. F. Hyndman and W. E. Mitchner.

cA Study of the Respirateey and Cardiac Activities of the
Skate sagen Intravenous Injection of Salt Solution. ..

Ida Hyde.
5.—Nutrition of the Embryo Sac and Embryo in Certain Labiatz,
F. H. Billings.
-6.—Mitosis in the Root Tips of Podophyllum peltatum, Aute Richards.
7. —Koeberlinia spinosa Zucc. : An Ecological Study of the
ee Anatomy of the Stem and some other parts.. Miriam Sheldon.
- $.—A Case of Absolute Tone Memory................ Archibald Hogg.

9. —The Hackberry Psylla, Pachypsylla celtidis-mammz Riley:
A Study in Comparative Insect Morphology ..... H. B. Stowe.
0.—Investigations Regarding the Phloém and Food Conduction
eee RR ERG CN ua. wey auins sodden Frank U. G. Agrelius.
-—Histology of Townsendia exscapa and Lesquerella spathulata,

Lillian Bunton.
12.—Celi Cihinges t in the Alveoli. of a Carcinoma of the Mamma,
o Earl F. Clark.
13.—The Semnoapsndylous, Amphibia and a New Species of Eryops
4 from the Permian of Oklahoma................ Roy L. Moodie.
‘14.—An Armored Dinosaur from the Cretaceous of Wyoming...
sige Roy L. Mosdia;

<A Contribution to the Soft Anatomy of Cretaceous Fishes,
- and a New Promitive Herring-like fish from the Texas
Pe ee air a ra rg! SO MRE a atalel oie & & Roy L. Moodie.

C. E. McClung and Edith Pinney.
UGE SEE age Arthur Bolles Frisell.

ce eh
CRT ST OS 0 MA RS Se:

j 5 Re ES ae Ne Oh asa ne, heen
ae: eee) roe b Abani

ee Spi y ae

ee
“ ce

aoa. aig

;
see? WAR

Vol. V, No. 12— March, 1910.

2 (Whole Series, Vol. XV, No. 12.) ;
CONTENTS: -

SIN THE ALvEoLI OF A CARCINOMA OF THE MAMMA,
pe Pere Earl F. Clark.

PUBLISHED BY THE UNIVERSITY,

LAWRENCE, KAN.

Ocroser, 1911.

SN,

L64 -

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vou. V, No. 12] MARCH, 1910. VoL. XV.No de

CELL CHANGES IN THE ALVEOLI OF A CARCINOMA
OF THE MAMMA.
BY EARL F. CLARK.
(Contribution from the Zodlogical Laboratory, No. 188.)
Submitted in partial fulfillment of the requirements for the degree of Master of Arts.
Plates XLVI to XLVIII.

HE literature upon the pathological new growths of the
human body scarcely mentions amitosis as a method of
cell increase. It occurred to me to treat in a comparative

way the two methods of cell division, mitosis and amitosis, as
they are found in neoplastic material.

It is not intended to make this paper conclusive, but to make
it the first of a series of case studies from the results of which
a general conclusion may be deduced as to the relative im-

portance of indirect and direct cell division in these new

growths.

I am indebted to Dr. M. T. Sudler for the following history:
“In the first week of October, 1907, she fell, bumping the
breast against a hard object. No further notice was taken of
the incident, however, until the latter part of November, when
she noticed a hard, thin, round lump in the right breast about

_ the size of a silver dollar. It was then freely movable. About

the 25th of December she noticed a tingling in the right arm,
and on January 1, 1908, the arm began to ache. The original
mass increased rapidly in size until the whole breast was in-
volved. Its rate of growth seemed intermittent, some weeks
being much more rapid than others. Upon examination the
entire breast was found to be occupied by a hard, firm mass,
which was intimately incorporated in the connective tissue of
(209)

210 KANSAS UNIVERSITY SCIENCE BULLETIN.

the breast. The nipple was somewhat retracted but not com-
pletely. The mass was movable over the muscle. The axillary
nodes were enlarged and hard. On February 5 the breast, to-
gether with the axillary lymph nodes, was removed. On March
13 a recurrence in the left breast was discovered, which was
removed on March 20. Twelve days later metastatic nodules in
the neck were removed. The patient died soon afterwards
from an apparently general carcinomatous involvement of the
body. From the foregoing history it is plain that this was an
extreme type of the most rapidly growing and malignant of
tumors.”

This material was not prepared especially for fine cytological
work because the discovery of its. excellence for investigation |
was not made until after its fixation. It was fixed in ten-per-
cent formalin and remained in a weaker solution until used.
I found the cell nests somewhat shrunken away from the sur-
rounding stroma and the chromatin either so swollen or so
poorly preserved that I was unable to make out details in re-
gard to the chromosomes. The fixation otherwise was good.
The tissue was mounted in paraffin and cut in sections eight
micra thick. Sections were stained in hematoxylin for twelve
hours after being in a mordant of iron-alum for twelve hours.
They were then decolorized in the iron-alum solution to the
desired extent.

The cells studied appeared in variously shaped pockets or
alveoli surrounded by dense white fibrous connective tissue.
This connective tissue stroma was penetrated more or less by
columns of tumor cells, one to three cells in diameter. In
short, this specimen in parts conformed to the scirrhus while
in others it was more of the medullary type of carcinoma. In
this paper I will confine my study to the cell nests or alveoli.

In the first four figures I chose a typical field from widely
separated parts of the tumor for each drawing, and drew all
the structures appearing in that field. I did this to show the
similarity of the cell content of the tumor mass throughout.
The normal resting cell, with its variation in size, shape and
chromatin content, composed the greater number of cells of
each group. It was considered the normal resting cell because
of its uniform staining quality, its maintenance of cell and
nuclear wall, its content of regularly outlined chromatin nu-
cleoli within a single nucleus. Marked variation in any of these

Bait
Cid

a 5” ¢
“1 ‘

bah SS
i) a +

7 ci

CLARK: CYTOLOGY OF CARCINOMA. 211

characteristics was considered a cell change having as its
ultimate end either the production of new cells or degeneration.

Among the resting cells were others which could be easily
identified as being in the process of indirect division. Al-

- though no detail of the chromosomes could be made out, yet the

increased amount of chromatin over the resting cell, its affinity
for the stain, the irregular outline of the chromosomes, the ab-
sence of the nuclear membrane and the shape and position of
the chromatin masses identical with those of the prophase,
metaphase, anaphase, and telophase leave no doubt as to the
identity of such cells. Equally distributed among the resting
cells were other cells which were clearly abnormal, owing to
their unusual size and their content of two or more nuclei or '
clefts of greater or less depth which either separated or tended
to separate the single nucleus into two or more distinct seg-
ments. These separated segments had all the characteristics
of the nucleus of the resting cell. I have considered such cells
as in the process of direct cell division, although I have been
unable to find a single cell showing distinct separation of the
cytoplasm. Adami thus describes amitosis: “The nucleus
divides without any apparent preliminary rearrangement of
its structure. It becomes elongated, then dumb-bell shape (the
process can be followed in the amceba), the connecting neck
becomes broken across and the two daughter nuclei pass apart,
their separation being in some cases followed by division of the
cell body while in others this further division is wanting and
the binucleate cell is produced.’ Patterson, in ‘“Amitosis in
the Pigeon’s Egg,” says: “In most cases the nucleus elongates
in the direction transverse to the plane of the future division,
after which a constriction in the nuclear membrane appears

about the entire circumference of the nucleus. This con-

striction continues to deepen until a complete division is ef-
fected. A modified form of this type is found in those cases
in which the constriction proceeds from one side.”

Although the process in this material differs somewhat, I
think, from that described by Adami and Patterson for other

_ material, the result is the same. Therefore I feel safe in as-

suming that cells with the nucleus wholly or in various degrees
of separation are in the process of amitosis. Ziegler and

-Vom Rath consider such cells as already in the process of de-

generation. If so, it will not be out of place to consider along
with these division figures structures that appear throughout

212 KANSAS UNIVERSITY SCIENCE BULLETIN.

this material and which are plainly cells in varying degrees of
degeneration, as evidenced by changes in the chromatin, cy-
toplasm and cell wall, and note if there is any similarity be-
tween them.

Figures 5, 6, 7 and 8 are used to further illustrate these three’

cell changes as they appear in various parts of the tumor. In
these drawings I omitted the greater number of the resting
cells.

I will supplement this preliminary account by short de-
scriptions of individual cells in the several plates with inter-
pretations of their variation from the normal.

I have used the term “cell nests” to mean the group of cells
within the alveoli; the term “chromatic cells” to mean those
cells in that condition in the degeneration process just before
the chromatin loses its staining power and when it has swollen
_so much and become so diffuse as to fill the entire nucleus as
a more or less densely staining mass.

OBSERVATIONS.
The Cells of the Nests Change by Mitosis, Amitosis and Degeneration.

In showing the cell changes of these nests I will simply use
a number of figures, describing each one with some amount of
detail and inserting comments as observations suggest.

FIGURE 1 is drawn from cell nest 2. It includes three cells
dividing directly, one dividing indirectly, two cells in the
process of degeneration, and a number of resting cells.

The resting cell is smaller than the dividing cell, is of poly-
hedral shape, has a limiting membrane, and a large, well-
marked nucleus. Its cytoplasm is denser than the karyoplasm
and is of a reticular appearance. The nucleus has a well-
marked nuclear wall and usually contains from one to three
regularly round, deeply-stained chromatic bodies which I will
call chromatin nucleoli. A directly dividing cell is of much
larger size than the resting cell and usually a bit clearer, a
condition suggesting that it has included a large amount of
fluid preparatory to its division. The nucleus divides by the
formation of a groove or cleft on one side. I have not seen a
cell where the cleft appeared on the two sides. The cleft
deepens and the nuclear substance pulls apart to form the

regular outline of the daughter nuclei. The nuclear membrane:

seems to follow the deepening cleft so that as soon as the seg-
ments have separated the karyoplasm is enclosed by its wall,

oe

CLARK: CYTOLOGY OF CARCINOMA. 213

In the cells shown I have drawn them as they appeared in out-
line in one plane, but included the chromatin through the depth
of the cell. Each well-defined segment of the nucleus usually
contains a chromatin nucleolus. In cell A the lower nuclear seg-
ment has a second cleft appearing upon its inner border. Cell
B has two distinct nuclear segments, each one of which has a
secondary groove appearing upon its upper surface. The
upper segment shows no chromatin at all. Cell C shows two
nuclear segments, each containing a chromatin nucleolus. Cell
D shows a central constriction of its nucleus with the mother
chromatic mass in the plane of constriction, but I do not say it
is in the process of division.

Cell # is a representation of the typical indirect division
figures found in this material. It is in the early anaphase.
As I have already said, this material, being fixed in formalin
with no particular care regarding its cytological preservation,
shows no detail characteristics of the chromosomes. I have
used in my counts only the well-marked division figures, from
the close spireme stage on. In these indirect divisions the
cells are larger than the resting cells, the nuclear membrane
had disappeared, and about the chromatic figure the cytoplasm
was clear. I was unable to make out the spindle or any of the
details farther than shown.

Cells F F are two fragments of degenerated cells in each of
which the chromatin has swollen and coalesced into one homo-
geneous, deeply staining mass.

FIGURE 2. Among the cells of this group there is not the
noticeable difference in size usually seen between the dividing
cells and the resting ones. Cell A shows an indirect division
plainly. Cell B shows the nucleus being divided by a cleft
passing in from its upper surface. Each segment contains a
- chromatin nucleolus. Cell C, an otherwise ordinary resting
cell, contains in its cytoplasm, half way from the cell border
to the nuclear membrane, a clear vesicle, regular in outline
and with a definite limiting border indicating that it is filled
with a substance differing from the cytoplasm.

FIGURE 3 shows no new structures, but attention is called to
the regularity of the chromatin nucleoli. Cells A and B show
the nuclei entirely segmented and each segment containing
regular chromatin nucleoli, the lower segment of cell B con-
taining two,

214 KANSAS UNIVERSITY SCIENCE BULLETIN.

FIGURE 4 is a group of twelve tumor cells drawn from cell
nest 1, a small nest of seventy-five cells containing six indirect
and no direct divisions. In this group are shown four of the
six indirect divisions. Cells A and B are noticeable for their
large size, while cells C and D are of the same size as the rest-
ing cells. Cell H is a common-sized resting cell, containing one
chromatin nucleolus but with a hollow clear vesicle bulging
into the wall of its nucleus. Cell F' is a common resting cell ex-
cept that it has its chromatin diffused through the karyoplasm.

FIGURE 5. In this group and in those following I have not
included all the cells that appeared in the field, but simply in-
structive ones and enough resting cells to show a comparison.
The group was drawn from cell nest 3, of 253 cells, which in-
cluded six indirect and five direct divisions. Cell C was in the
fourth tier from the edge and on the peripheral side of the
- group. It was large, being .0125 mm. in its greatest.diameter
and showing a nucleus in two segments, each segment con-
taining a regularly round chromatin nucleolus. The cell was
noticeably clearer than the others. Cell B also has a biseg-
mented nucleus, but its chromatin was scattered in irregular
masses through the karyoplasm. It was also large, being .01
mm. in its greatest diameter. The nucleus of cell A was in
three segments, each part having a regularly round chromatin
nucleolus, and beside this a certain amount of chromatin scat-
tered in clumps through the karyoplasm. It was .0125 mm. in
its greatest diameter. In all my cell measurements I used a
micrometer scale which I have found by measurement to give
the following dimensions: With a drawing board on a level
with the object and at the same inclination as the microscope
stage, using %e2 oil-immersion objective and a 8x Huyghenian
ocular, an object that fills the square of the scale is .0125 mm.
in diameter ; with the % objective it is .03 mm. in diameter, and ©
with a % objective it is .15 mm. in diameter. Three-fourths
of the square equals .0093 mm.; one-fourth of the square is
equal to .0031 mm.; one-half of the square is equal to .0062
mm. Cell D is in the indirect division process. Cell EH is a
structure that is common among the tumor cells and I have
concluded that it is a degeneration result, although it may
occur anywhere in the cell nest among the otherwise vigorous
cells. It is characterized by a large amount of deeply staining
chromatic material, homogeneous and closely massed, having

‘ ve 7

CLARK: CYTOLOGY OF CARCINOMA. 215

no definite form but with a more or less regular outline. This
mass evidently represents the nucleus, which has become filled
with a diffused, swollen mass of chromatin. Its cell membrane
has broken down, allowing the cytoplasm to flow out into the
surrounding intercellular spaces. Notice the regular form of
the chromatin nucleoli in this group of cells.

FIGURE 6. This group of eleven cells was drawn from the
same cell nest as the previous group. Cell A is a large cell,
.0093 mm. in its greatest diameter. In the cytoplasm about the
nucleus are four vesicles, similar to those described under
figures 2 and 4. The chromatin of its nucleus is of regular
outline but is unusual on account of its size, being a trifle over
one-half the diameter of the nucleus itself, at least twice the
amount found in the ordinary resting cell. I think it is a cell
just beginning the degeneration process shown by the swelling
of its chromatin. Cell C is undoubtedly in the process of de-
generation, for its cell membrane has disappeared, allowing

- the cytoplasm to flow out uncontrolled. It is a cell almost as

large as cell A. Its chromatin, which exists as a deeply stain-
ing mass somewhat irregular in outline, was larger in diameter
than that of cell A. The nuclear wall could not be made out.
Cell B is one farther along in the process of breaking up than
cell C, for most of its cytoplasm has gone. The diameter of
its chromatic element is not so much larger than that of cell
C but it is more condensed and of more regular outline. These
three cells, I think, are respectively in advancing stages of dis-

_ integration. Cell D is a cell dividing indirectly, having a chro-

matic mass larger than that of cell C but of a looser arrange-
ment, indicating individual chromosomes. It also maintains its
cell borders perfectly.

FIGuRE 7. This group of eleven cells was drawn from cell
nest 6 in the region of a necrotic area. Cells A and B are in
the fifth and sixth tiers respectively from the free edge of the
tumor on the peripheral side of the group. I will use this
group and the one following it, figure 8, to show changes in the
cells passing inward from the free edge of the cell nest to a
necrotic area. On the side peripheral to cell A there is a small
necrotic spot two or three cells in diameter, which accounts
for cells A and B on that side of the group. Cell A is a good
specimen of such cells as C and B of figure 6 or cell F of figure
5, It shows the swelling of the chromatin and the disin-

mie. 3: KANSAS UNIVERSITY SCIENCE BULLETIN.

tegration of the cytoplasm well. While there has been a
swelling of the chromatin in each of these cells there has been
an equally great shrinkage of the nucleus, so that at this stage
there has been a real loss of material to the nucleus. Cell B is
peculiar in that its nucleus, while practically the same size as
normal, stains very dark; not solid, as in A, but granular.
There is also a faint indication of its chromatin nucleolus re-
maining, not indicating so much that it has become diffused
through the karyoplasm as that the karyoplasm itself assumes
a new staining power. The cell wall of cell B is broken down
in one place, showing signs of disintegration. Cell C, a directly
dividing cell, is .0093 mm. in its greatest diameter, while cell
D is just a trifle less than .0125 mm. The average diameter of
the resting cells of this group is near .007 mm. So I think
that a cell preparing for direct division increases in size to a
marked extent.

FIGURE 8. This group of sixteen cells is taken from cell nest
6, the same nest that group 7 is taken from, and drawn from a
zone of cells to the inner side of group 7, that is, between
group 7 and a necrotic core which lies in the central part of
the nest. These two figures show well the transition from the
more peripheral nondegenerating cells to the entirely necrotic
mass of cell debris which fills the central portion of so many
of the cell nests. No dividing cells were shown. Cell A, the
most peripheral of the group, is the only one showing a chro-
matin nucleolus. It is an ordinary resting cell. In cells B
and B’, a type of many cells found at this zone about the ne-
crotic area, and for that reason described as characteristic,
no chromatin nucleolus could be made out through the depth
of the cell, but in each of these the chromatin was scattered
through the karyoplasm in the form of a loose network, stain-
ing lightly. The cell and nuclear membranes were preserved
perfectly. Passing inward toward the center of the nest is a
field of cells which I have called “chromatic cells,” and which
are perfectly characteristic of a zone of cells that immediately
surrounds the necrotic core of the cell nest. They are a
transition stage between the normal cells and the wholly de-
generated ones. They resemble closely the disintegrating cells
described in figures 5, 6 and 7, but seem to have another history
of formation. As in the other cells, the nucleus is filled with a
homogeneous chromatic mass, the outer layers staining darker

PS a ae ee
: PY sad

Ct ee ee ee Tat” Pees
ee

CLARK: CYTOLOGY OF CARCINOMA. 217

than those farther toward the center. In this field the stain
was not as intense as that in the cells of the previous figures but
was of a grayish shade, and passing toward the necrotic area
the light in the gray increases until in cells D and D’ there is
just enough black remaining to outline the nucleus well. From
this stage it gradually fades out until the cells are entirely ne-
crotic and do not stain at all. The cell membrane in this
group could not be decided upon satisfactorily. In cells A,
B and B’ it was perfectly intact, while in cells F it seemed to
be entirely gone. In other cells, as C, C’, the cell outline was
definitely preserved, while yet again in cells D and D’ the cell
boundary was irregular and indefinite. But I am safe in say-
ing that when the cell has so far degenerated that the color |
has disappeared from the nucleus the cell membrane has also

disappeared and disintegration immediately follows. But in

many of these cells the cell membrane has at least partially
gone while the chromatin stains deeply. The nucleus of cell
B’ is 0.004 mm. in diameter while the chromatin mass of cell
C is near 0.003 mm. in diameter, showing an appreciable
shrinkage of this element in the degeneration process.

The Mitotic Divisions Occur in General Upon the Periphery of the
Alweolus, the Amitotic Nearer the Center.

I will use camera lucida drawings, together with tables, to
prove this. I counted the number of cells, the number of di-
rect divisions and the number of indirect, and also counted the
tier of cells from the periphery of the cell nest in which each
division occurred. I hoped by this method to arrive at some
conclusion as to the rapidity of growth of the tumor, in order
to compare it later with the growth of the adult and em-
bryonic mammary tissue. I also wished to decide as to the
relative frequency in occurrence of the direct and indirect di-
visions and to note, if possible, whether there was any com-
parison between the relative proportions of these and the sort
of nests in which they occurred. I also wished to decide if
there was a well-founded relation between either of these
methods of cell division and the necrotic areas indicating that
one or the other was related to degeneration, or that they
occurred equally in the same region showing that they were
equally vital methods of cell growth.

CELL NEST 2 was the non-necrotic limb of a necrotic body.
It was counted up to the beginning of the necrotic area. I can-

218 KANSAS UNIVERSITY SCIENCE BULLETIN.

not give the tiers of cells in which the various divisions oc-
curred in this nest.

No. of cells. ° Direct divisions. Indirect divisions.
563 9 f 16
Percentage of cells in. divisions. 2.3.05 ie oe eee rl. a SUaaets Shape 4.42
of. cells in. direct divislons wi sec. ide sea eee Pe soe
i of cells in indirect divISIONS (\ 060. csac a bee eee eee 2.90
ee of divisions which are direct «2:30 2) ieee 36
Ks of divisions. which:.are. indirect) .o54 a2. didwe see 64

CELL NEST 3. This nest apparently is a small nest in di-
ameter but it has four “chromatic cells,’”’ which indicates that
it is somewhat degenerating.

Tier LAO 82 abs Oe of ease: en Av. Total
Direct’ Dy s.c.ko 3. Ce ge ey eae Lee ree git eee eae ee 6.2 5
Indirect Deeks Ss ese Rae ees <2 go ee 2.3 6

No. of cells. Direct divisions. Indirect divisions.
253 5 6
Percentage of: cells in. divisions: .'.6592..54. tA 2. SR eee 4.34
of cellsin: direct divistans. «sis siersseess onto. Bev aie eee 1.90
eb: of ‘cells: in: Indirect’: divisions: «ia. <s soe wn et ee ee eee 2.40
* of divisions ‘which aré direct sis 6. 29s shale hee 45.40
We of: divisions: which are: indirect: isis ss sie Soren 54.50

CELL NEST 4. This was a nest with a necrotic center .125
mm. to .1125 mm. in diameter, while around this on the edge
of the nest was a zone of cells varying in width from .05 mm.
to .037 mm. which was undegenerate and which constituted
the number of cells counted, so that the total number of cells -
counted does not show the real size of the nest in point of num-
ber of cells, one-half of the area of the nest being degenerate.

Tier y hee eae foe aur een Bee ees GE Sn is Av. Total
Direct Def. De SR? ee eee Fi eat 3
Indirect D. ...... Cree We ye Renee erro come beauty beer ae one. 52 ee

No. of cells. Direct divisions. Indirect divisions.
515 12 8
Percentage of. cells in divisions i5 ocd boeken bind Seg Ge 3.88
of cells in’ direct divisions: ... 05... 0... ve ceo ee 2.30
wt of cells: in indirect. divisions.) . 225. 23a See 1.50
$ of divisions: which: are directs..i\:\...) 87s eee 60
“ of divisions which are indirect: oo... 3c: 6 a ee 40

CELL NEST 5. This was one of the smaller sized nests and
was quite vigorous, showing no degenerating cells at all.

Tier 1.2.83 4-6 67 8 $40 11 “ao Soe

Direct D. .......- ple OR SE | See RZ. 00 00
Indirect BD, of B sos a OE ee ae

CLARK: CYTOLOGY OF CARCINOMA. 219

. of cells. Direct divisions. —_— Indirect divisions.
106 ; 00 4

ASS C of cells in divisions “ee ee eee ee wee eee eee ee 3.77
of cells in direct OE EERE GO EEE tS See ee

of cells in indirect divisions ........... erate. Rae i
‘of divisions whieh are direct | 2005 G26 ie. ccc ccs

of divisions which are indirect been e nese ee ee eee ees .100

bl NEST 6. This was a cross section of a large alveolus,
75 mm. in length and .1758 mm. in width. It was
oval in outline. There were two degenerate spots

ps SE ee Soe RE Os Isat! Av. Total
Te aoe . i 4 1 . 3 . 2 1 o* . 5.0 12
Pree v

Bee ee ; ae 5.2 4
Direct divisions. Indirect divisions.

Barete 12 4
EES A eer 2.36
of cells in direct divisions ....................4-. ve RD
of cells in indirect divisions ...................66: . 0.60
of divisions which are direct .........+.+..+++++++. 75
of divisions which are indirect .............-...4. 25

a T.. This was a small, oval, non-necrotic nest.

oe 24 8 eT 8 8 10-21 Av. Total

ee ey «200, 08
Pee ee eB Dies

Direct divisions. Indirect divisions.
fs ; 00 5
Of cells in divisions .......<.........+ ber pes pease =|
of cells in direct divisions ey ee ee eee 0
et cells in indirect divisions ........... pT pe ope . 4.06
- of divisions which are direct .......... ScMiGUs bie haat wie

ous divisions which are indirect .......... rear ran cae

eS gs io i Av. Total
Tp aS Soe Stes Eva ce anes eer art 42.8

eck < Peer ca ac ee e
_ Direct divisions. Indirect divisions.
ae 2
(SEES RE MEE .. , 8.84
_in direct divisions toy hs Tet ig ae ass eee ERE 2.60
MN ee ed 0 yin, we nly es ROE eg
which are direct .................. keh) «(Oy
isi oo Bes ee 9 ny sacar ae

220 KANSAS UNIVERSITY SCIENCE BULLETIN.

In the following cell nests I have taken the serial sections of
the same nest and counted them in the same way as I did the
single sections of the previously counted nests. I hope in this
way to in a manner corroborate my results and also to study
the history of the cell nest farther.

CELL NEST 10. This nest was .4125 mm. in length, .0875
mm. in width. It was a nest that showed occasional spots of
six to eight cells that had broken down, an average of two of
these spots in each nest. The direct divisions occurred largely
in the upper, wider end, while the indirect occurred largely in
the lower, narrower end, but I have not illustrated this by
sketches. It was therefore taken as a nest, non-necrotic, but
near the stage when it would rapidly have become so, indicated
by the degenerate spots.

CELL NEsT 10, A.

Tier. 1°28 £8 6.27) B90. TT: AY DPD: ae: eee
Divert De ee ee es 5a) ee 8 ne Fe.
Indirect D; 1° 6 2° 8 Be) AA Sasi S siecle aa ae 14° 682

CELL NEsT 10, B.
Dirset (Do 241 222 bi AD 3a ae ok
Indirect:D,.2:. 85-3..8 2.6 . 11 600
CELL NEST 10, C.
Direst. Dae 8 1S ee aes 3.8 7 ive
Indirect. D2 2 = 2.852 es 84 9> BST
CELL NEstT 10, D.
Direct Bes Ld 822.8 46 10 x ae
Indirect D232 22... ; 2.4 ot 9 655
CELL NEsT 10, E. :
Divrett Di ee 8 Base ead ae ee ee 9 oe ar
Indizect.D. 1: A 2 te et gictag er Mak i 7 654
CELL NEST 10, F.
DigectDe 2 iS 424. 28 Rome Pence anes ly eae Ar oe
Indivect:D: 2° 2.3. av 2 are gi eeige kg oe ee ne 14. +650
CELL NEstT 10, G.
Direct D. . a eee ope hae os We ecg he 1 mo co
Indirect D. 3 qe gs ee Bee a Oe ee aes 15 564
No. of cells. Direct divisions. Indirect divisions.
4212 64 79
Percentage Of celia Th GRyMIONG fy ce ek eno cee eae ee 3.39
of cells in direct divisions: -=. 7s.--2) 2 1.50
~ of cells in: indirect divisions... i a ee 1.80
ig of divisions: ‘which are direct... oc. ee ee 44
& of divisions Which are. indirect 2)... ¢ bk a ee ee 56

The direct divisions averaged 1.68 rows nearer the center
than the indirect divisions.

LARK: =CYTOLOGY OF CARCINOMA. 221
iLL NEST 11. This nest was probably one-half of a half

CELL Nest 11, B. ‘ite
Roe ay Eye Bee
et eee oe ce) ee eee

aac Nest 11, ae ee
ih 66 7.89 10 MA DD. LD. Cells.
Bea we RE Cres 3 384
ab ae

51 pegs
J) eto ae

CEL Nesr 11, C.
ee ere a pee gle Foy,
i A ee eR. ec a ee

CELL NEST 11, D.

: ee ie rite atest 18 8 os soe
ee eo... A (B88

Orta, Nese 1 E.
t

ey | Cau Nese 11, 6. ees
an Ses a SAS Y 10 Mase Telus

ee ee ee
CELL NEst 11, I.

| WES ces Samii wre S| oh ae Sse
ist 5. igo a ey Sa Oe ee eee

Cet Nest 11, J.

ary
bo
bob

222 KANSAS UNIVERSITY SCIENCE BULLETIN.

CELL NEST 8. The work upon this nest was done with
greater care than that upon the previous nests. I first made a
camera lucida sketch of the nest in outline, then went over the
nest carefully with the high power 8x compensating ocular
and marked each division I found upon the outlined sketch. I
‘then used an 8x Huyghenian ocular, upon the diaphragm of
which I had placed a micromillimeter scale, and went over the
field again, carefully counting the cells by the aid of the mi-
crometer scale, and verified the previous locations of the divid-
ing cells and located any new ones that I may have missed in
the first survey. I thus corroborated my work and made it
doubly sure by locating each dividing cell upon an outline
sketch. I also used these sketches in the hope that together

with the tables I might farther establish the apparent relation

of directly dividing cells to the degenerating areas. Where
this relation has been especially noticeable I have made high-
power camera lucida drawings of the particular field to show
some detail about the cells I counted, simply drawing the more
important structures in relation to them. Using these large
drawings and marking their position upon the smaller sketches
one cannot avoid showing in the same way the slide does the real
relation of the various dividing cells to these necrotic areas.
In these outline sketches I only outlined the nest, tracing into
any irregularities among the cells along the edge, so that, un-
less otherwise marked, rifts in the nests will mean spaces left
where the cells have apparently been pulled apart by external
means. The areas that are entirely necrotic will not be dis-
tinguished from those that show definite degenerating changes ;
_ cells that are in the “chromatic cell” stage I have counted as ne-
crotic. These areas are shown by cross-hatched work. The in-
direct divisions are shown by a dot or “>” while the direct
divisions are shown by a circle. Starting with figure 10 the
sections are in serial arrangement. This cell nest 8 is only a
portion of the entire alveolus, showing at the right side of the
sketch just the end of a necrotic plug, and leaving the greater
part shown in the vigorous condition. I used this nest to con-
trast, if possible, the two divisions in these two areas.

eee A Rp A en

=

ne re

CLARK: CYTOLOGY OF CARCINOMA. 993
Fic. 10. CELL Nest 8, A.

3 4 8 6 % O26 1 Av. DD. ID. Cells.

. 1 3 . fe: 2 me ite Bs: * . 5.4 8 oe eee

4 1 2 2 osiee ens ee ‘, a 8.2 o. 15 534

Fic. 11. CELt Nest - |

2 : Ber oor Bree es Jog t Ma 9 r ook

4 wile Ce ESE a ree eae bl Panis 15 511

re ons 32. Sel Neor 8, Cc.

1 £ as 1 2 Giek alee idecre a 5 15 on he

Sv e2e 1 5 ar that See > al glen 1 554

Fic. 14. Creit Nest 8, C.
Phat «py Ane ih oie eas 7. 10 as Re
ya 5 alee Ae ae Sealing Be Pager: ; 16 555
CELL NEst 8, D.

Thais ince”) (emi pene Hy denote 8 2 ‘ ae
Mae Ba -1 BR 1G (1686
tte?) SUBNEST 8, D.

<n iy ULE ee eg oe 45 2 ee
Reece te BE nist oO i478
Bae _ CELL NEst 8, E.

“ont ke SGN. See eer Bae Wil ee ek saat
ca 74 4. See OO as ee BRD
ee - CEeLy Nest 8, F.
ee eee eB CRBS OR es 5 t eeise § 35 ee eee
ek 2S B.S . Soe Seis a Ce
ss Fig. 16. Cen Nest 8, G.

Se ae eae | 3 Cp LO) Gat tN ee, | eens meee ree
tig Ate et ee ee Smear ake 8 es 13 916
ist Fie. 18. CeLt Nest 8, BH:

> as Eee : hiya “aa AR detest i; twat 1) P See
a ae 5 BB gs eS BOT
“Fig. 19. Cet Nesr 8, I.

Peewee 2 8, LG hec Le vy aioe

Pee ad i 8B ee ee
cells. Direct divisions Indirect divisions.

: le 121 179

CS oe Sree, |
cells in direct division ................. petal
f cells in indirect division .................. cae ca ee
' divisions which are direct .................-.08- 40.33
Be see whine - + Sige See ee ee pie Pals Oe

294 KANSAS UNIVERSITY SCIENCE BULLETIN.

DESCRIPTION OF OUTLINE SKETCHES AND DRAWINGS.

CELL NEsT 8, “A” (fig. 10). The amount of necrotic material
in this figure is greater than in the following ones, the necrotic
plug showing less and less in each consecutive section. There
are no divisions shown in this figure near the necrotic plug,
but the nearest two are direct. The greater part of the divid-
ing cells are in the limb of the nest.

CELL NEST 8, “B” (fig. 11). The number of divisions pres-
ent in the distal part of the limb is noticeable. There are three
direct divisions immediately adjacent to the necrotic plug and
one nearer than the nearest indirect division.

CELL NEST 8, “C’”’ (fig. 12). In this nest there was an un-
usually large number of dividing cells. At the point of the
arrow in figure 12 is the center of the group of cells drawn on
a large scale in figure 13. About the arrow point are the four
‘direct divisions and the two indirect divisions shown in the en-
larged group. This group of nineteen cells includes a larger
area under the lens than any previously drawn. I drew all the
cells of importance under the: field of the camera lucida, in-
cluding a field .05 mm. in diameter, while the previous groups

drawn have included an area only .08375 mm. in diameter. Itis |

located on the edge of the necrotic area and includes the whole
width of cells lying between this area and the free edge of the
cell nest, cell F' being in the first row of the nest, while cell B is
among the necrotic cells. The direct divisions of this section
are deep in the center of the nest, while the indirect are
largely very close to the edge, seven cells out of twenty-five
dividing indirectly lying in the first row and no directly di-
viding cells nearer than the second row, and then only one.
A general survey of these sketches together with the tables
strongly indicates that the cells on the edge of the tumor
increase almost, if not wholly, by indirect divisions, while those
deeper away from the edge increase largely by direct divisions
but occasionally by indirect. I want to emphasize by the large
drawings the intimate association of the directly dividing cells
with the degenerate ones. Cells H, H’, H” are the “chromatic
cells’? that I have described previously under figure 13. Sur-
rounding these cells closely are cells A, B, C, D and E, plainly
in the process of direct division. Cell B is peculiar in that it
shows no chromatin nucleolus in either segment of the
nucleus. This same occurrence was noticed in numbers

A Slough eae —
} 2 7 3 3 ,
, 5: PR 4 ‘ 2 el = ‘
ah x : ee! ;
< ’ i " = siete ahi ec a ‘ :
- i ety ee ee . Bog , e few ae Som ie hx His . ‘ oe ‘
a dealt a ere eee Rasen Ree Me Pte |) St eee 2 os RAN ater he
eee Aa) ee ee oe < "Eeg “I ee et ee ie *

vor,

—%

< y * 2
ee ee Oe eee ET ae ey on Be ae

ee
fz.
:

CLARK: CYTOLOGY OF CARCINOMA. 225

of resting cells, B, B’, figure 8, etc., just outside the zone
of “chromatic cells’ H, H’, H”. Its chromatin had wholly
or largely disappeared, and if any remained it was loosely
diffused through the karyoplasm. Cell F is another cell show-
ing no chromatin in any of the segments of its nucleus. It
evidently was in decay for it was very translucent. The frag-
ments of its nucleus showed no chromatin at all. It may be that
a cell such as B, which has vigor enough to cause its nucleus
to segment but not to cytoplasm, gradually loses its staining
ability and its definiteness of outline and later becomes merely
a mass of debris through a stage represented in cell EZ. Cells
F and G are two cells in the indirect division process. The di-
rectly dividing cells are much larger than the resting cells or
those dividing indirectly. Cell A is .0093 mm. in its greatest
diameter, cell B is .007 mm., cell C is .0093 mm., cell G is .0062
mm., cell F is .0062 mm., cell EZ is .0093 mm., cell D is .0085
mm., while cell J, an average resting cell, is .0062 mm. in its
greatest diameter. This cell shows the breaking up of the
chromatin and its diffusion through the karyoplasm, possibly
preceding cells H, H', H”. Its nucleus is .005 mm. in diameter,
while the diameter of the “chromatic mass” in cells H, H’, H’”
is .0031 mm., showing evidently a shrinkage in the latter.

FIGURE 14. In this section two indirect divisions are shown

_ located on the very edge of the necrotic plug, and two more

near. There are also close to the edge of the nest three direct
divisions, still Showing a greater proportion of direct divisions
near the necrotic area than indirect, when compared with the
total number of divisions of each in the section. But this
shows that indirect division may occur on the very edge of a
degenerating area.

_ At the lower extremity of the limb of the regular section
following is found a circular group of cells, which in the next
section shows a connection with the larger body. This is the
first appearance of this group, but in the previous sections
there is a circular space indicating where it has been but from

a which it has in some way been lost. There is a cleft entirely

through the regular section, but this was where the cells had
pulled apart and not where they had degenerated. The direct
divisions are intimately associated with the necrotic spots. In
the middle segment there are two indirect divisions on the edge

226 KANSAS UNIVERSITY SCIENCE BULLETIN.

of the necrotic area, but usually they are near the free edge of
the nest.

In the succeeding section the subnest has joined the limb of
the main body and it was all counted together. ‘There are
three necrotic areas in this section. In the lower one the two
directly dividing cells of that region are on the edge of the
degenerated area.

A later section shows the nest in two segments, broken apart
from each other. A pluripolar cell division shows at one place.

The group of cells drawn under high power in figure 15 is
from this section. Figure 15 shows a group of five directly and
one indirectly dividing cells. In this drawing and those made
hereafter I shall indicate the segments of the nucleus as they

appear through the depth of the cell, demonstrating their over-

lapping. The border line used was drawn.at the edge of the
cells which presented no marked degeneration changes, but it
is not a hard, fast line, a few vigorous cells jutting out farther
than the line and a few chromatic cells being behind it. Cells
A, B, C, D and E' show well-marked direct divisions. Cell A
is on the necrotic border. Cell B, a very small cell for a di-
rectly dividing one, is a half cell width from the necrotic
border. Cell C is a trifle closer than B to the border. Cell D
is in the third row from the border. It is peculiar in that its

chromatin nucleoli in each segment are not staining deeply and ~

the chromatin is diffused through the karyoplasm. Cell EF is in
the fourth row from the necrotic edge. Cell F is an indirectly
dividing cell and is in the same row of cells as cells C and B.
Cells G and G’ are the “chromatic cells” described under
figure 13. They mark the edge of the necrotic plug and are in
various stages of degeneration, as indicated by the disintegra-
tion of their cytoplasm and the loss of the staining power of
their chromatin. This sketch together with the smaller one
shows well this collecting of large numbers of the directly di-
viding cells about these necrotic areas. There are mitotic cells
here, but not in. such proportion in comparison to their totai
number as the amitotic.

FIGURE 16. In this section there are more direct divisions
than indirect. There is a noticeable change in the number of
indirect divisions in the lower part of the nest, in what was
termed a “subnest” in a preceding section. In one section
there are nine indirect divisions, two direct. In the next there

cogs ; ‘
=¥ he ee
ee SS Oe ee, ee

hei

CLARK: CYTOLOGY OF CARCINOMA. 227

are eight indirect divisions, two direct. In the next there are
five indirect divisions, three direct. In a later (fig. 16), there
are five indirect divisions, four direct. In another (fig. 19),
there are four indirect divisions, three direct. This shows
consistently in this case that as the center of the nest is ap-
proached in the sections the number of direct divisions increase
in proportion while the indirect decrease in proportion. In the
region of the arrow point of figure 16 an enlarged drawing was
made (figure 17). I used this to show more detail about the
cells shown in figure 16. Cell A was on the edge of the necrotic
area. Its chromatin was small in amount and gathered in two
clumps in one segment and three in the other. Cell C is an
ordinary dividing cell. Cell D is in the second row from the
necrotic edge. The cross-hatched work shows the necrotic
area, this area being entirely necrotic and showing no transi-
tion cells along the edge.

FIGURE 18. There are but two dividing cells in the upper
segment of this nest, while in an earlier there are fourteen, the
number decreasing with each section until this one. They
have increased steadily in the middle part of the limb, however,
from eleven to nearly twenty in the section shown in figure 18.
This carefully worked out will demonstrate that at the same
time the cells in one part of the nest are rapidly dividing those
in another part are not dividing at all. There is either an
unequal stimulus to growth in the nest or an unequal negative
cause.

FIGURE 19. This nest is the same one as that in the preced-
ing figure, but in this section the two parts have been sepa-
rated. At the point of the arrow is the location of figure 20,
and here are shown the three dividing cells drawn in figure 20.
In figure 20, cell A is plainly among the degenerating cells in
the necrotic plug, and is surrounded by “chromatic cells,’”’ no
vigorous cell being near. It is a large clear cell with a biseg-
mented nucleus and showing but very little chromatin. Its
appearance indicates that it is a degenerating cell. Cell B
lies just outside the border of vigorous cells. Its nucleus is bi-
segmented and it stains much deeper than the cytoplasm. The
chromatin nucleoli stain very faintly. Cell C is a large clear
cell lying in the third row of cells from the necrotic edge. It
stains lighter than the other cells about it and shows a small
amount of chromatin and that stained faintly. Cells A and B

Out of the 294 direct divisions in all = rests onl,
in the first row on the edge of the nest.
Out of the 357 indirectly dividing veils 63.
first row on the edge of the nest.
The direct divisions in all the nests iviniede? j
nearer the center than the indirect divisions. =

CELL NEsT 12. This nest was: sketched Cie
forty-two serial sections.

CELL Nest ae
Tier es ae, Nee tame tice Gee eg
Divot... 5.50 coicketan ae os ee
: Tnidivect Dy os, Se ee
| Cent Nest 12; II.
Direct D. eee ee ewww eee . Bae . . . er
ladinect “Dy as44c4ies. a ae
Cenn Nest 12, II.
FAGIZOCE “Do oe cee Pe eee: He EP amare Te CS)

Direct Bic iues sce oe ee eee
Tesdliveet) BD es ioe gs I ee og

Direct. Diss uss sais sakes a ee eee san Ee
Indirect D.. ..-....026, 2 8 10 0 cet ee ae ees

Direct D. .<.. ss. ee eA ec | < ea
Tudireet DR 01.03) 5040 205 ge Ee oa

Direct D. Beebe <r x Br ROR ote:

Indirect’ D. “cincsepiaiestice! Sabri ais es
CELL Nest 12, VIII.

Direct. Bye.) :h52F bi a oe ee eee same
Indirect D. tesseteteee oO Be Chea? Syl are ak

ele _ CELL Nest 12, IX. |

Direct Ty. Oe a dM Se al . ey 4 1 . gai stadia 3.5

Indirect Dis eee 1 oS sacar enpran eign
CELL Nest 12, X.

Direct Dee eee % id 2 *‘ + Py

Indirect D, ore rere geese 2 2 e ’ * t ’

° ee eee eee

Re ep

ia
een “Se Fy
TSS. W Ae ie) 9; Kote

od seme ee ee ee
os

a He §

Pes 6 88h eh Se 6
eee eww eee
=
ee
ee ay ee ae eee
eis

a: oe

eee ee we ee

Ya ail Ae dad D Rtg
ar

eee ee eee

ag eo ae wee

a ee ee ee er es

eee eee ewe

08 Ce © 20 6 t9 50

eno) Se ee ee ee

eee © @ © ahe\eve
ps

PS Wt a ee ae

Se oe

GLARK: CYTOLOGY OF CARCINOMA.

. . . ‘a Shem ‘j . 3

& ee

CELL Nust 12, XV.

eee
¥ 3 be 2 . 1 1 as
Ceti Nest 12, XVI.

oe ee . ° 1 . s :
So te ee ae ee ate

Cen eee Fe XVII.
eat te
CeuL Nest 12, XVIII.
Boge ay gf oe ek Bk
Cet Nesr 12, XIX.
Sock BS Ot eo ae

. . i : = : ries 1 4

a ee
CeLL Nest 12, XXI.

Sy te ty: Gee ae

At, OP RSS

Cent Nes 12, XXII.

. r,

ae ee aa |
CeLL Nest 12, XXIII.
2 1i4)

es eae

—
CeLL Nest 12, XXIV.
as ae Se
CeLL Nest 12, XXV.

SG Se. seal Om

ae... 7) 3 r 5 ¢ 4

eee

one

215

ae

209

ee

254

oe

owe

449

230

Direct D.

Indavect 2Di ts oes

Direct D.
Indirect D.

Direct D.

Indirect D.

Direct D.

Indirect D.

Direct D.

Indirect D.

Direct D.

Indirect D.

Direct D.

Indirect D.

Direct D.

Indirect D.

Direct D.

Indirect D.

Direct D.

Indirect D.

Direct) Des... eee

Indirect D.

Direct: Dei. eee
Indirect D.

Direct D.

Direct D.

Direct D.

Indirect D

eee eee we ee ee

Indirect D.

Indirect D.

KANSAS UNIVERSITY SCIENCE ; ULLATIN

tee kes Bip oes 1 ee: ‘ ia :

. . ets ee

ge

NH wor nce

i seins
®
ie
rt

<

=

Q

=
OW. Ee Sas
weer
Hp

ee

Eee eid ss ch yea wee pea aren ys
Me ee
CELL Nest 12, XXX. fepeede ee

vee bk OR sha yt ae bak sake coe eh agen
CELL NEsT 12, XXXI. cg

1.0
ee eee Ss ee ey at se ees ket

CELL — est XXXII.

Ee Are 1 POS ee ec amee °: Sema |
et etre ee tes me SRN eG

CELL NEST 12, ‘XXXII.
NPE RAP OS Manes MiP Re Se 1 ee 3
RPE rede Fae eee Reon cave yee cts ,
CELL NEsT 12, XXXIV.

emi ema Bal Ge ete eye ty Cees
eee Pea fast e Gar ek Se mus

Daas hal Sar ee te) Sie oa nay re ty

ee very ee

eee eee www

:
CELL

Be es,
ao

eo

: fe:

Oe

ee
ee P age a

oe eee ewe ewe

me i poe
Bu.

Ti
eee oS
3

em

ee)

hee 6

CLARK: CYTOLOGY OF CARCINOMA. _ 281

Cet Nest 12, XLI.

oe ec oe din - eee BO. fis Eva ks
Sey eee tS B oyoee ee ee, | OB ; 5 6371
Cent Nest 12, XLII. |
By veges ine: set ae ere Me! <a, Sage
Meret Dies. es isk) 4D BEE 28 F 10 421
No. of cells. Direct divisions. ‘Indirect divisions.
———-:12,228 Edel 81 Rees eS SOB 236) 3
age of cells in divisions ........ Se yates ss = vs oi Ac yeay 2,"
s@f Celle in direct divisions ....0..... 2.60 .sseeeeese 6
ae ye Inuleet ‘Givione 146 e ES 2.3
of aiitions which are direct ....... bait they ii)

of divisions which are indirect ............. Se eo 77.9

sin indirect division :
gage ae of Percentage of Pibsobbiaaiet
ls in cells in cells in
a direct division. indirect division.
Pik cee 4.42 1.5 2.9
Fae Fane nega 57 ' 1.9 2.4
Bee Pons ees 3.88 2.3 1.5:
np ee Mae if — 0.0 8.77 .
pee wr 2.36 ss 6
4.06 0.0 4.06 —
4.01 1.6 | 2.45:
3.34 2.6 at
3.39 1.5 fa. foie
8.59 1.9 1.6
Pe awe. Cee cal eee Sah a: ; 2.3

2 writer of this paper wishes here to acknowledge his in-
c : s to Doctor McClung, whose kind suggestions made
th work Saseible. and to Doctor M. T. Sudler for the material

e -— data upon which this study is based. st”
CONCLUSIONS.

The percentage of cells in division shows no etetian’ to
ercentage of amitoses or mitoses.

The percentage of mitoses is high in small non-necrotic

The percentage of amitoses is high in large necrotic nests,
ating that it is the division associated with degeneration.

itosis is the form of division occurring in the necrotic

ite of division in the various nests is inconstant.
ectly dividing cells lie nearer the center of the cell
Sei, Bull., Vol. V, No. 12. |

232 KANSAS UNIVERSITY SCIENCE BULLETIN.

nests and thus nearer the necrotic areas than the indirectly
dividing cells. -

7. The indirectly dividing cells lie nearer the periphery of
the nest than the directly dividing cells.

8. From these observations it may therefore reasonably be
concluded that amitosis represents the final effort of the cell at
reproduction, and that shortly after its occurrence ieee
tion ensues.

9. Amitosis, mitosis and disintegration comprise he a
changes in the alveoli.

10. Cells in direct and indirect division are larger than the
resting cells.

11. In one stage of degeneration of the cells the chromatin
becomes diffused through the karyoplasm and so swells that it
fills the nucleus as a deeply staining, homogeneous mass. In
another process of degeneration the chromatin loses its staining
power and the cell becomes clear.

12. There is a shrinkage of the nuclear bulk in degeneralten:

13. A border of chromatic cells lies between the vigorous
cells and the cell debris of the necrotic area and marks the
limit of degeneration; such cells are themselves degeneration
results.

BIBLIOGRAPHY.

C. M. CHILD. Studies on the Relation Between Amitosis and Mitosis.
Biological Bulletin, vol. XII, No. 2, January, 1907.

J. GEORGE ADAMI. General Pathology.

J. E. S. Moore and C. E. WALKER. First Report on the Oytolanicnt In-
vestigation of Cancer, 1906.

ALFRED STENGEL. General Pathology.

J. T. PATTERSON. Amitosis in the Pigeon’s Egg. Anatomischer Anzeiger,
Bd. 32, 1908.

CONTENTS: nes

AMPHIBIA AND A NEW SPECIES OF ERYOPS
OF OKLAHOMA, . } ; Roy L. Moodie.

LAWRENCE, KAN.

October, 1911.

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vou. V, No. 13.| MARCH, 1910. » bypime teg peaer

THE TEMNOSPONDYLOUS AMPHIBIA AND A NEW
SPECIES OF ERYOPS FROM THE PERMIAN
OF OKLAHOMA.
BY ROY L. MOODIE.
(Contribution from the Zodélogical Laboratory, No. 189.)
Plates XLIX to LIV.

UR knowledge of the fossil Amphibia begins with Dr. J. J.
Scheuczer’s discovery of two skeletons in the Miocene of
Oeningen, which he referred to as “Homo diluvii testis

et Theoskopos,” and which later Hermann von Meyer desig-
nated “Andrias scheuczeri,” thus keeping the same idea in the
generic name as Scheuczer had expressed in his letters and
descriptions. The discovery was made about 1725, so far as I
can ascertain. His letter to Sir Hans Sloane in 1726, which
was translated in the republished volume of the early trans-
actions of the Royal Society of London, is the first definite
mention I find of this discovery in the literature which is ac-
cessible to me.

Exactly one hundred years elapsed from the time of Scheuc-
zer’s discovery of the giant salamander skeletons to the time
when Dr. Georg Friedrick Jaeger discovered in the Triassic of
Germany some labyrinthodont remains which he figured and
described in his folio published in 1824 at Stuttgart with the
title “De Ichthyosauri sive Proteosauri Fossilis Specimenibus
in Agro Bollensi in Wurtembergia.” He figured the occipital
portion of the skull and teeth of what he, in 1828, called ““Mas-
todonsaurus.” In 1833 Doctor Jaeger pointed out the syn-
onymy of Salamandroides with Mastodonsaurus. Previous to

(235)

236 KANSAS UNIVERSITY SCIENCE BULLETIN.

Doctor Jaeger’s discovery, however, was that of the form
Pygopterus, the history of which is given below.

The literature on fossil Amphibia for the next twenty years
following Jaeger’s last mentioned paper deals almost entirely —
with the Tertiary Amphibia, although in 1844 and 1845 Her-
mann von Meyer published some interesting papers on the more
ancient amphibians. It is interesting to note that in von
Meyer’s “‘System der fossilen Saurier,” 1845, he does not men-
tion a single temnospondylous amphibian.

Our knowledge of the Temnospondylia dates from the year
1847, in which year Dr. August Goldfuss described Archego-
saurus, the first known and for many years the only representa-
tive of this order of amphibians. Doctor Goldfuss gave as the
title of the paper in which he described Archegosaurus “Ueber
das zlteste der mit Bestimmtheit erkannten Reptilien.” It will
thus be noticed from the literature that the early writers re-
garded the labyrinthodonts and their allies as reptiles. It was
Quenstedt who in 1850 made the first contention that “Die
Mastodonsaurier im gruenen Keuper Sandstein Wurtemburgs
sind Batrachier,” and from that time dates the development of
our knowledge of these extinct forms as Amphibia. Three
years after Quenstedt’s paper appeared, Vogt upheld the same
contention in his essay “‘Archegosaurus und all Labyrintho-
donten sind Amphibien nicht Reptilien.”

In 1850, after the discovery of the Archegosaurus and its
identification, Doctor Jaeger published an interesting essay en-
‘titled “Ueber die Uebereinstimmung des Pygopterus lucius Ag.
mit dem Archegosaurus dechenii Goldf.” In this essay Jaeger
brings out the interesting facts of the early discovery of the
_Archegosaurus. I give the following translation of the first
paragraph of Jaeger’s essay: ‘In the collection of the late
Professor Storr, of Tiibingen, which was received in 1817 at the
Imperial Museum of Natural History at Stuttgart, there was
found a skull enclosed in a spherosiderite nodule. This skull
was listed in 1777 as No. 192, in the printed catalogue of the
Pasquay collection (which was the foundation of the Storr col-
lection), without giving the locality, as a fish. skull which had
been discovered in a non-fissile gray shale. Agassiz, at whose
disposal the entire collection of fossil fishes was soon after
placed, designated this specimen on a label in his own hand-
writing at first as Aspidorhynchus nov. sp. On a label written

aM v

MOODIE: PERMIAN ANPHIBIA. 237

by him later it is designated as Pygopterus lucius. In the first
part of the second volume of the Fossil Fishes (Poissons fos-
siles), page 10, Agassiz placed this skull in the family Sauroidze
as the second species of the genus Pygopterus, with the short
description: ‘Pygopterus lucius Ag., une téte seulement, dont
la machoire supérieure est plus allongée. Houille de Saar-
briick.’ In the same work, part II, page 78, this specimen is
designated as the fifth species of Pygopterus, with the descrip-
tion, ‘une téte avec des dents trés acerées, de la houille de Saar-
briick. L’original se trouve au Musée de Stuttgart.’ He re-
gards the same specimen as the third species on page 162 in the
synoptic table of the family Sauroide arranged in the order of
the formations of the Carboniferous which contained Sauroide.
It is also given this place in the general table of fossil fishes
which appeared in 1844, page xxxvi.”
Jaeger affirmed the supposition of Agassiz that the speci-
men was from Saarbriick, from the similarity of the nodule to
those found at that place. He then proceeded to show that the
skull described by Agassiz as Pygopterus lucius really shows
reptilian characters which Jaeger regarded as intermediate
between those of a crocodile and an Jguana. He then compared

_ it with the skull of a young crocodile and also with the skulls of

various lizards. He found that the measurements and propor-
tions of the skull compare favorably with those of the crocodile,
and that the teeth have the form and size of those of a newly
born crocodile. The position of the orbit is also identical with
that of the crocodile. He made the important observation that
the nostrils have the lateral position of the lizards. He then
compared the specimen with other known fossil forms, 7. e.,
Zygosaurus, Rhinosaurus and the labyrinthodonts, and found

considerable differences. He then made the following interest-

ing observations regarding Archegosaurus: “At the meeting of
naturalists in Aachen in 1847 Goldfuss described the remains in
one of the many spherosiderite nodules, which contained a rep-
tile skull, to which he gave the name A7rchegosaurus. He estab-
lished three species for the genus, which he described more
fully in a later publication. Only one species of this genus,
Archegosaurus decheni, was described and figured in the
Neues Jahrbuch fiir Mineralogie, 1847, Tab. VI, with a line
drawing.” Jaeger then pointed out in a conclusive manner the
identity of Pygopterus lucius with Archegosaurus, He re-

238 KANSAS UNIVERSITY SCIENCE BULLETIN.

garded Archegosaurus as allied to the Crocodilia but having
characters of the Batrachia as were exhibited in the labyrintho-
donts, and on account of the geological position of the fossils
he supposed that Archegosaurus would really belong with the
labyrinthodonts.

Huxley in 1862 described Pholidogaster pisciformis from the
Carboniferous of Scotland. In the next year he distributed the
ancient Amphibia into the two groups, the Archegosauria and
the Mastodonsauria of von Meyer. In the former group he
placed Pholidogaster, to which it manifestly does not belong.
Huxley was himself doubtful as to its location, for he says:
“Archegosaurus, of course, takes its place among the Archego-
sauria; and Pholidogaster, I suspect, must go with it, though its
vertebre are far better ossified” ; and in a footnote he makes the
interesting statement: “It seems to me probable that the verte-
bral centra of Archegosaurus may really have been osseous
rings, such as are found in embryo frogs and salamanders,”
which shows that he was on the right track but did not quite
comprehend the nature of the vertebree of Archegosaurus.

Our knowledge for the next few years of our history is
advanced but little, and that only by sundry discussions on
Archegosaurus. In 1866 Owen placed Archegosaurus in his
subclass “Dipnoa,”’ under the order “Ganocephala.” In the
same year Haekel retained Owen’s classification, except that
he displaced Owen’s “Dipnoa” with “Phractamphibia.”’ Cope,
in 1868, retained essentially the same classification as proposed
by Owen. In 1874 the committee of the British Association
for the Advancement of Science reported on the extinct Am-
phibia, and placed Archegosaurus by itself in group VI, the
Archegosauria. |

No further additions were made to our knowledge of the
Temnospondylia until 1875 and 1878, when Cope described
Cricotus from the Carboniferous of Illinois and Trimerorhachis
from the Permian of Texas. In the next year Gaudry described
Actinodon from the Permian of France. In 1882 a decided ad-
vance was made in the separation by Professor Cope of the
rhachitomous forms such as Eryops, Actinodon, Zatrachys, Tri-
merorhachis and other genera, which have subsequently proven
to be synonyms of the genera mentioned, into a distinct group.
Cope proposed the term Rhachitomi for this subordinal group,
and designated two families, the Eryopide and the Trimeror-

MOODIE: PERMIAN ANPHIBIA. 239

hachidz as its members. Later in the same year Cope de-
scribed Acheloma from Texas. In 1885 Lydekker added other
genera and families to the group, notably Gondwanosaurus,
which he placed in the family Archegosauride. In 1884 Cope,
and in the following year Fritsch, contributed further to the
knowledge of the Temnospondylia by the description of the
forms Diplospondylus and Cricotus from the Permian of Bo-
hemia and Texas. In 1888 Zittel proposed to place all temno-
spondylus amphibians under the suborder Temnospondyli in
two groups—A, those with rhachitomous vertebre; B, those
with embolomerous vertebree—for which two groups Cope had
previously, 1885, proposed the terms Rhachitomi and Embo-
lomeri, but had regarded them as coérdinate with the group
Stegocephali.

Case, Broili, and others have in recent years given additional
information on the forms previously described and have de-
scribed some new forms. In 1909 Dr. S. W. Williston, in his
essay on “T'rematops milleri from the Permian of Texas,” was
enabled to give a restoration of the skeleton of that form. In
the present bulletin the writer will describe an interesting new
embolomerous amphibian under the name Spondylerpeton
spinatum, from the Coal Measures of Mazon Creek, IIlinois.
This is the oldest of the known Temnospondylia.

We know nothing of the ancestry of the Temnospondylia.
They occur in the Carboniferous and Permian, and if the Ster-
eospondylia are highly specialized Temnospondylia, as seems
probable, they also occur in the Triassic, but this is yet to be
proven.

The geographical distribution of the Temnospondylia is as
follows: In the Carboniferous and Permian of North America;
Europe; and Asia, in the Permocarboniferous of India, Gond-
wanosaurus. So far as the writer is aware they occur nowhere
else unless some of the little known forms from South Africa
will, on further search, reveal temnospondylous characters. It
_ is entirely too soon to arrive at any conclusions as to the paths
of migration of this group. They followed the way which was
open to other land animals, whatever that may have been.

Nor can we say in what region the first of the Temnospondy-
lia lived. Judging from the remains so far discovered their
place of origin must be located in North America. The an-
nouncement of the discovery of the eryopoid form from Pitts-

240 KANSAS UNIVERSITY SCIENCE BULLETIN.

burg, Pa., by Case, and the discovery by the writer of the
cricotoid form in the Mazon Creek fauna, as well as the pres-
ence of Carboniferous Cricotide in Illinois and Kansas, all
point to an origin of the temnospondylous group at an earlier
date in North America than in any other region of the globe.
The embolomerous forms are by far the older.

The species of fossil Amphibia from the Carboniferous,
Permian and Triassic are so little known that it is difficult to
give a definite list of the species of the Temnospondylia. The
list given below is believed to be approximately accurate. It
does not differ in essential outlines from that given by Zittel
in 1887. The species are distributed in the various groups to
which, at the present time, they seem to belong:

CLASS—AMPHIBIA, LINNE, 1758.

ORDER—TEMNOSPONDYLIA, Zittel, 1887.
Suborder—RHACHITOMI, Cope, 1885.

FAMILY—ERYOPIDZ Cope, 1882.

Eryops megacephalus Cope, 1877. Permian of Texas.
erythrolithicus Cope, 1878. Permian of Texas.
ferricolus Cope, 1878. Permian of Texas.
reticulatus Cope, 1881. Permian of Texas.

“2? platypus Cope, 1877. Carboniferous of Ohio.

“ ~ latus Case, 1905. Permian of Texas.

“ _willistoni Moodie, 1911. Permian of Oklahoma.
Trematops milleri Williston, 1909. Permian of Texas.
Acheloma cumminsi Cope, 1882. Permian of Texas.
Ichthycanthus ohiensis Cope, 1877. Carboniferous of Ohio.
Gondwanosaurus bijoriensis Lydekker. Bijori, India.
Zatrachys apicalis Cope, 1881. . Permian of Texas.
conchigerus Cope, 1896. Permian of Texas.
crucifer Case, 1905. Permian of Texas.
micropthalmus Cope, 1896. Permian of Texas.

serratus-Cope, 1878. Permian of Texas.

Actinodon frossardi Gaudry, 1880. Permian of France.
“2 latirostris Jordan. Permian of France.
. » (?) Anisodexis enchodus Cope, 1897. Carboniferous of Ohio.
imbrecarius Cope, 1882, Permian of Texas.
*Rhytidostaus capensis Owen, 1884. Karoo Beds, South Africa.
(?) *Euchirosaurus rochei Gaudry. Permian of France.

FAMILY—TRIMERORHACHIDZ Cope, 1882.

Trimerorhachis bilobatus Cope, 1883. Permian of Texas.
conangulus Cope, 1896. Permian of Texas.

es insignis Cope, 1878. Permian of Texas.

i mesops Cope, 1896. Permian of Texas.

ae leptorhynchus Case, 1902. Permian of Oklahoma.
Cricotillus brachydens Case. Permian of Oklahoma. é
Archegosaurus austricus Makowsky, 1876.
decheni Goldfuss, 1847. Permian of Germany.
latifrons Gein. & Deich. Permian of Saxony.
Zs ornatus Woodward, 1905. Permian of India,

“

MOODIE: PERMIAN ANPHIBIA.. 241

Platypodosaurus ricardi Twelvetrees, 1880. Permian of Russia.
= stiickenbergi Trautschold. Zechstein of Russia.
Cochleosaurus bohemicus Fritsch, 1876. Permian of Bohemia.
Gaudrya latistoma Fritsch, 1885. Permian of Bohemia.
Chelyosaurus vranyi Fritsch, 1878. Permian of Bohemia.
*Sphenosaurus sternbergi Fitzinger, 1840. Permian of Bohemia.
Sparagmites lacertinus Fritsch, 1885. Permian of Bohemia.
*Sclerocephalus hauseri Goldfuss. Permian of Germany.
*Sclerocephalus credneri Fritsch. Permian of Bohemia.
*Sclerosaurus bavaricus Branco. Rothl. Ohmbach.
*Melosaurus uralensis H. von Meyer. Kupfersandstein.
*Osteophorus roemeri H. von Meyer. Rothl. Schliessen.
*Zygosaurus lucius Eichwald. Permian of Russia.

Gonioglyptus huxleyi Lydekker from the Gondwana beds of
India also possibly belongs in this family.

So many genera and species of extinct Amphibia are based
on fragments of crania or limbs that an exact diagnosis of the
forms is impossible. The species marked with a star (*) are
uncertain as to position. Many of the species listed under the
Trimerorhachide are placed in this family on the authority of
Lydekker, who included them in his family Archegosauride,
which was preoccupied by Trimerorhachide. Doctor Williston
regards Trematops milleri Will. and Acheloma cumminsi Cope
as representatives of a family distinct from the Eryopide for
which in 1910 he proposed the name Trematopside. He also
regards the species of Zatrachys as entitled to family rank in
the Zatrachydide, Williston, 1910.

Euchirosaurus of Gaudry is probably a reptile.

Suborder—EMBOLOMERI, Cope, 1881.
FAMILY—CRICOTID2 Cope, 1884.

Cricotus gibsoni Cope, 1877. Carboniferous of Illinois.

“ erassidiseus Cope, 1884. Permian of Texas.

“ heteroclitus Cope, 1875. Carboniferous of Illinois.

“ hypantricus Cope, 1884. Permian of Texas.

“sp. indet. Case. Carboniferous of Illinois.
Diplospondylus punctatum Fritsch. Permian of Bohemia.
Spondylerpeton spinatum Moodie, 1910. Carboniferous of Illinois.
?*Discosaurus permianus Credner. Permian of Saxony.

FAMILY—DISSOROPHID Williston, 1910.

Dissorophus multicinctus Cope, 1895. Permian of Texas.
x mimeticus Cope, 1896. Permian of Texas.
testudineus Cope, 1896. Permian of Texas.
Cacops aspidephorus Will. 1910. Permian of Texas.
Aspidosaurus chiton Broili, 1904. Permian of Texas.

“

From the foregoing diagnosis is will be seen that about 55
species of the 350 which have been assigned to the fossil
Amphibia of the world can be located in the order Temno-

242 KANSAS UNIVERSITY SCIENCE BULLETIN.

spondylia. Without doubt some of these species will prove in-
valid and there will be others already described which will in
time prove to belong to this group, but the above statements
represent an approximation of our present knowledge.

Since the present essay is to deal especially with a new mem-
ber of the genus EHryops, only those groups will be here de-
fined to which Evryops and its allies pertain.

ORDER—TEMNOSPONDYLIA, Zittel, 1887.

Terrestrial or semiaquatic vertebrata; skull bones pitted and
grooved; lateral line canals present; pineal foramen some-
times absent; sclerotic plates present; vertebre rachitomous
or embolomerous; notochord usually persistent; one or two
sacral vertebra; tail present, short or long; limbs and girdles
all well developed; pectoral and pelvic girdles composed of the
usual stegocephalan elements; an osseous pubis’ present;
cleithrum present on the scapula in some forms; limb bones
well developed and bones of forearm and leg separate; carpus
and tarsus osseous, 12 elements present in tarsus of one form,
11 carpals; five digits in hand and foot; phalangeal formula
2, 8, 3, 4, 2 (?) for the hand and 2-3-3-3-2(?) for the foot;
venter covered with an armature of osseous scutes, sometimes
overlapping; skin of back bare or armored; ribs heavy, double-
‘headed, curved, and either moderately long or short. Body
short and heavy. As compared to skull, about two to one.
Temporal foramina present in a few forms.

Range. Pennsylvanian to (Triassic) Upper Permian.

Distribution. North America—lllinois, Kansas, Oklahoma
and Texas. Europe—Germany, Bohemia, France. Asia—
India.

The suborder Rhachitomi is characterized solely by the na-
ture of the vertebral column, which is cut into triangular seg-
ments.

FAMILY—ERYOPID4 Cope, 1882.

Large terrestrial or amphibious vertebrata; skull bones
deeply marked with pits and grooves which at times take the
form of lateral line canals; carpus and tarsus osseous; pubis
osseous ; pentadactyl fore and hind limbs; orbits in the typical
genus located far back on the skull and near the median line;
cleithrum present on the scapula; vertebre rachitomous, the
intercentrum supporting the arch in the dorsal region; para-

MOODIE: PERMIAN ANPHIBIA. 243

sphenoid well developed or reduced; teeth on pterygoids, pala-
tines, prevomers and parasphenoid.

Range. Upper Pennsylvanian to Permian.

Distribution. America, Europe, Asia.

Genus—Eryops Cope, 1877.

The following definition of the genus is given by Doctor
Branson: “Skull long, comparatively narrow; proportion of
length to breadth about nine to seven. Roof bones coarsely
sculptured posteriorly, finely sculptured anteriorly. Nasals
and premaxille very large; frontals excluded from orbits by
junction of pre- and post-frontals. Pterygoids not meeting in
the median line; parasphenoid dagger-shaped, tapering grad-
ually to a point just in front of the palatine foramina; pre-
vomers large. Orbits subcircular, situated in the posterior half
of the skull; nares subovate, remote, at a considerable distance
from the tip of the skull. Many minute denticles on pterygoids,
palatines, prevomers and parasphenoid. Teeth circular in
cross section, strongly ribbed near base, dentine strongly in-
folded. Three large teeth on each palatine. Mandible without
postcotyloid process. Vertebrz rachitomous. Ribs double-
headed. Pelvic bones coalesced.”

There have been up to the present five species of Hryops de-
scribed, with a doubtful sixth, ? Eryops platypus Cope, from
the Carboniferous of Ohio.

In 1889 Lydekker described a species of E'ryops, E. africanus,
from the Karoo beds of South Africa. Broom has subse-
quently shown that this species does not belong in this genus
and has placed that species with two others in his genus
Rhinesuchus, in which he includes R. (Macromerion) gumbeli
(v. Ammon), R. (Eryops) africanus (Lydekker), and R.
whaitsi Broom.

An additional species of Eryops is represented by remains
in the collection of the University of Kansas and it is here
described as new.

Eryops willistoni, new species.

Dr. S. W. Williston in 1899 described and partially figured
m the Kansas University Quarterly portions of a skeleton of
a species of Eryops, which he tentatively referred to the
species Eryops megacephalus Cope. He has several times
expressed to me the opinion that the form is distinct from
EF. megacephalus, and a recent careful study of the ma-

244 KANSAS UNIVERSITY SCIENCE BULLETIN.

terial has led me to regard the remains as a new species,
which may be very appropriately named Hryops willistoni as
a partial expression of gratitude to my honored preceptor in
paleontology. Since, also, several interesting points in
anatomy and morphology have come up for consideration the
following discussion is presented in the hope of adding some
knowledge on this peculiar Permian genus.

The material which represents the present species comprises
the following portions of the skeleton: Portions of the skull;
several teeth and a nearly entire right mandible; the scapula-
coracoid of both sides, incomplete; the right clavicle; portion
of the interclavicle; right humerus, radius, ulna and a
phalange; about a dozen vertebre and the left sacral rib.

There are various fragments of the skull preserved. On one
of these fragments is observed a portion of the right orbit and
a part of the prefontal and frontal bones clearly marked off
by sutures. The present form differs in this respect from the
known specimens of E'ryops megacephalus, in which the sutures
remained unknown until they were discovered by Doctor
Branson, who was able to trace them with the aid of a lens.
On the present specimen the sutures stand out with almost
startling distinctness. They do not have the sharply zigzag
form assumed in #. megacephalus and other of the larger

Amphibia, but the sutures occur in bold curves. I detect a.

portion of the supraorbital lateral line canal on the fragment

containing the orbit. The sculpture of the cranial bones in the

present form is quite different from that of E. megacephalus.
It is much more rugose and the rugosities have a larger form.

This holds true for the mandible also (plate XLIX, fig. 1), and

it is taken as one of the specific characters.

There are preserved in the collection also a single complete
tooth and fragments of others. The complete tooth. (plate
XLIX, fig. 2) is of the typical labyrinthodont form. It is oval
in cross section with sides flattened. It arises from a tumid
base of cancellated bone as in the mosasaurs. It is recurved
and sharply pointed.

In 1899 Ludwig Strickler (Paleontographica Bd. XLVL p.
85, plates XI, XII) investigated the minute structure of the
teeth of Eryops megacephalus.

The mandible of the present species is preserved in a muiéh
broken condition, but its form is nearly complete. The form
and ornamentation of the posterior portion were well shown in

Te a

MOODIE: PERMIAN ANPHIBIA. 245

Williston’s figure (Kan. Univ. Quart., Ser. A, vol. VIII, pl.
XXX, fig. 1), which is reproduced here with slight changes.
The sharp angulation and the posterior position of the angle is
a specific character which serves to distinguish this species
from Eryops megacephalus Cope, to which the present form is
most nearly allied.

The jaw (plate XLIX, fig. 1), is heavy and elongate. The
portion posterior to the coronoid angle is an acute triangle
which ends in the articular. Only a portion of the suture
bounding the articular can be distinguished. This is indicated
in the figure. The coronoid suture is evident for its entire
course. The suture starts from the articular suture, and run-
ning directly forwards turns upward just before reaching the
teeth, which do not occur on the coronoid. This element is a
thin bone superiorly but more thickened below. The sutures
separating the angular, surangular and dentary cannot be de-
termined. The arigular and surangular contain, on their exter-
‘nal surface, the peculiar nodose ornamentations which help to
characterize the present species.

The sculpture on these elements consists of scrobiculate
‘knobs some 5 to 7 mm. in diameter. This ornamentation ends
apparently at the dentary, since the tooth-bearing element,
which is quite large, is ornamented only with coarse elevations
and pits.

‘The operculo-mandibular lateral line canal can be traced as a
depression beginning on the articular and running forward as
far as the bone is clearly preserved. It occurs as a shallow
groove interrupted by the nodose and scrobiculate sculpturing
of the mandibular elements. The internal surface of the man-
dible is not preserved in sufficient detail for description.

Complete teeth are lacking from the mandible in its present
condition. Doctor Williston figured two of them as complete,
and it is quite probable that they have been broken off since he
studied the specimen, because the material is excessively
fragile. The hollow character holds for all of the tooth bases
which are preserved on the mandible. An additional portion
of the mandible is shown in figure 5, plate XLIX. Its char-
acters do not differ from those described above.

- In regard to the hollow nature of the teeth, Tomes has the
following explanation: ‘The laminz of pulp, with their several
systems of dentinal tubes, instead of passing out in straight

246 KANSAS UNIVERSITY SCIENCE BULLETIN.

lines like the spokes of a wheel, pursue a tortuous course as
they run from the central small pulp chamber toward the sur-
face. Not only do they undulate, but they give off lateral proc-
esses, and at their terminations near the surface of the tooth
the thin laminz of pulp (so thin that the radiating pulp cham-
bers are mere fissures) become dilated; so that on section cir-
cular canals are seen at these points, as is also the case at the
points where subsidiary processes branch off.” (Dental An-
atomy, 5th edition, 1898, p. 65.)
Measurements of the mandible of Eryops willistoni:

Greatest Tenet 66.88 se Pet A OS ee 48 cm.
Length of posterior portion (fig. 1)..............0.0- 33
GrORUGNE WHEEL, 5562555 epnrt Secaa bos Fach pines + Weeki naie eeee 10
Width of dentary midway from coronoid process...... 6.5
Width acréss artiteler cies. #65 SA csa ds ee oS 2.6

TV eRn tls OF LOOGE bs iia unin altcd wicin 4: cue aks Sea in Nie kar a 2.
Thickness "OF SAMI AC DASEs C0 cs sco ts Shee cae ss 5

_ The right arm of this species is preserved in an incomplete
condition, as may be seen by reference to plate LII. The form
of the humerus is well shown in the figures, plate LIII, fig. 7,
and plate LII. It suggests to a certain extent the humerus of
a mosasaur. The present element does not have the bizarre
appearance which exists with the specimen of that element of
Eryops megacephalus figured by Professor Cope in 1890
(Trans. Amer. Philos. Soc., vol. XVI, p. 367, fig. 3). The large
projections are wanting entirely from one side and on the other
they are not so highly developed, and yet the form of the bone
is such that it can be nothing else save a humerus. It is too
rugged for a femur.

The specimen is abraded to some extent, so that the exact
form of the element cannot be determined. It may have had
the form suggested by the broken lines in the figures. The
humerus is that of the right side. Viewed from in front the »
element is seen to have a peculiar form (plate LIII, fig. 7).
The lower portion of the inner surface is distinctly concave.
The inner side of the bone is comparatively smooth. There is
a low ridge which runs obliquely across the antero-superior
angle of the bone. There is a distinct twist in the inner sur-
face, which is continued to such an extent that the inner sur-
face is carried around almost parallel with a portion of the
outer surface. It then takes another and inward curve and
ends in a projection.

One of the projections from the outer surface of the humerus

MOODIE: PERMIAN ANPHIBIA. 247

is in the form of a hook and strongly suggests the hooklike
process which is developed in'the quadrate of the mosasaurs.
This hook and the other elevations on the edge of the bone
without doubt served as points of attachment of the deltoid and
pectoralis muscles, which were undoubtedly well developed in
the species of Eryops.

The outer surface of the bone is exceedingly rough and there ©
are no less than six distinct elevations on this surface. One
of these elevations is a ridge which runs at somewhat of an
angle across the posterior edge. There is an indication of a .
foramen opposite the pit which occurs on one side of the bone,
which at first was thought to represent the entepicondylar for-
amen. This was such an important matter that the bone was
cut through at this point by Mr. Martin and fully demonstrated
the total absence of this foramen in the element although there
is quite a distinct pit at this point. Cope described such a fora-
men in Acheloma, but Case, who has recently examined the
type, says the foramen is an accidental opening. Gaudry like-
wise described such an opening in Actinodon.

Measurements of the humerus:

Length of specimen as preserved.................. 96 mm.
Estimated length along same line................. 110
SEE EUG OF OUINIU, oc csc ect cen eadenevess 50
Width of top of humerus.....................005- 90
EN Mi ANION TONNE oi Mire ele trechinscn & v ¥ ove emis. dia: oe 72
Thickness of head as preserved................... 23
Diameter of entepicondylar pit.................... 3

The element which is here regarded as a radius is very sim-
ilar in form to that described by Cope for Eryops mega-
cephalus. Its form is well shown in the figure. The lower end
is somewhat larger than the upper, which is abraded and
broken. Viewed from the lower surface a cross section of the
bone shows a semicircular form. It is hour-glass shaped in its
entire length, with a shallow groove near the lower end. Both
ends of the bone near the articular surfaces are roughened by
vascular pits for muscular attachment or for the union of the
cartilaginous tips.

Measurement of radius:

Actual length of specimen......................... 91 mm.
Estimated length of specimen...................+-- 95
Width of upper end (estimated) ................... 40
Sesseeee miadie Of Shaft... ....5. 2.0. c. 6. le 24
ecw densiscsaseced 60

a, sv cece dcwstedosevaes 82

248 KANSAS UNIVERSITY SCIENCE BULLETIN.

The element here figured as an ulna (plate LII, U) is doubt-
fully referred to the arm, since the bones were found disasso-
‘ciated. It may belong to the leg. Its form is specifically differ-
ent from that which Cope figured as the ulna of Eryops mega-
cephalus, although its form is, in general, the same. It does
not depart widely, in point of form, from that of the radius,
which we should expect to be the case in a generalized verte-
brate. It is smaller, somewhat shorter, and the groove on the
ventral (?) surface is lacking. The radius and ulna are both
peculiarly amphibian in the way they have weathered. The
ends have assumed the concave form so characteristic of the
Branchiosauria and the modern Amphibia, indicating a higher
development of the perichondrium than of the endochondrium.

Measurements of the ulna:

Actual length of specimen..................000008- 80 mm.
Estimated length of specimen.................0200- 93
Width of shaft at midge. ois cast o0 asco aos Se Os 22
Width of upper end (estimated)................2.. 47
Width of lower end (estimated)................-5. 40

Associated with the limb bones there is a portion of a waa
lange of the hand (?) of the specimen. There are no carpals
preserved. The element figured (plate LIII, figs. 4 and 5) is
probably a metacarpal. Its form is very closely similar to that
figured by Cope for the metacarpal element of Hryops mega-
cephalus. Its upper end is very broad, with recurving edges
and with rugosities for muscular attachment. Its form is
crescent shaped and concave. The shaft narrows abruptly
from the upper end to the middle and then widens for the lower
end, which is lost.

Measurements of the metacarpal:

Actual length of speciMeti., <..« ¢cn% -55 5 seu oon ee tere 23 mm.
Estimated length of bone............... Me TS es Ee 4 am
Width of upper ends cu oi4 seas eek sere eee eeee ids Be
Thickness of upper end.......+.e+seseeceeereereeee 8

The following description of thes coraco-scapula (plates
L, LI) is taken from Doctor Williston’s sss ks The oe
tion of the cleithrum is appended:

“The bone is elongated, and nowhere very thick or massive.
The distal part of the scapula is much thinned and considerably
expanded; the immediate margin here, however, is wanting,
so that the precise outlines cannot be given.

“The position of the two bones must have been very obligge:
as is evident from the position of the glenoid cavity. The pos-

*
e)
a

re ne ee eT

MOODIE: PERMIAN ANPHIBIA. 249

tero-inferior border is moderately thickened, and rounded;
gently concave along the proximal part and convex distally.
The antero-superior border is much thinner than the opposite
one, and is concave throughout the extent of the shaft, except
distally, where it is codssified with the procoracoid. The union
of these two bones is very close in this region, the sutural line
being distinguished with difficulty if at all. The procoracoid
is narrower and more thickened below, reaching to the lower
part of the conjoined bone, and lying in close apposition though
not suturally united. Both the lateral surfaces of the scapula
are nearly flat. A little proximad to the narrowest part of the
shaft the bone is much thickened by a stout ridge on the inner
side, which includes between it and the remainder of the bone,
just back of or above the cotylus, a large, elongated foramen,
both of whose orifices can be seen from the inner side of the
bone only.

“The glenoid surface is elongated and deeply concave in its

long diameter. The scapular portion is much smaller than the

coracoid, and is partly separated from it by a constriction;
this surface is nearly fiat and is placed at right angles to the
plane of the coracoid surface, looking directly downward when

the bone is lying horizontally. The rest of the cotylus is mod-

erately concave and looks outward and somewhat backward
and upward with the bone in its former position. Just how
much of the conjoined bone is formed of each element it is im-
possible to say since the union is so close that no trace of a
junction is perceptible. Evidently, however, the coracoid
forms only a small part of the whole bone. Its anterior or in-
ferior border is gently convex, the posterior one slightly
concave.

“The anterior end of the superior border of the scapula is
gently convex, as is also the broad external surface. On the
inner side the foramen described above opens into a deep,
elongated, boat-shaped cavity at its distal end, the cavity
formed by the upper border of the ridge described. In the
lower or anterior end of the cavity there are apparently two
smaller foramina, the external orifices of which are just within
the anterior margin of the scapular face of the glenoid surface.
The proximal surface of the scapula is concave; that of the
coracoid, if the bone is limited by the ridge spoken of, is for the
most part gently convex and lies in a more mesial plane.”

2—Univ Sci. Bull., Vol. V, No. 12.

250 KANSAS UNIVERSITY SCIENCE BULLETIN.

The scapula is evidently that of the right side, the same as
the humerus described below.

On the antero-superior corner of the scapula is a part of a
peculiar element—the cleithrum. Its occurrence in the genus
Eryops has been noted by Case in two species. In EF. latus the
cleithrum takes a most peculiar form. The form assumed by
the element in the present species is impossible to determine,
since the greater part of it has weathered away.

The cleithrum is, apparently, simply laid on the upper edge
of the scapula, the union being hardly a sutural one. In the
present specimen it is represented only by the lower part, but
it could not have had the immense expanse superiorly as Case
and Williston have figured for the cleithrum in Eryops latus
Case. It was probably quite slender, as it is in the dicynodont
reptiles. The cleithrum is a thin’ plate of bone with a ridge
running along the interior border. The edge of the scapula
is depressed to receive the element.

The characters of the scapula which are at variance with the
scapule of the known species are the narrowness of the shaft,

the location and size of the cleithrum, and the position of the

cotylus.

In regard to the three foramina seen and described in the

scapula Williston has recently proposed the terms supra-, post-
and infra-glenoid or supra-coracoid to designate these open-
ings. The foramina occur in all known Temnospondylia, as
well as in the Cotylosauria (2) and in the Pelycosauria (2).
The writer has recently described a branchiosaurian from the
Pennsylvanian of Mazon Creek in which three foramina are
. visible in the scapula. This was thought to be of interest in
connection with the Temnospondylia, but investigation showed
that in the temnospondyles the foramina are in no case to be

regarded as occurring strictly in the scapula, but rather as be- |

longing to the coracoid which is united with the scapula.

An interesting question of homologies, or it may be parallel
development, arises in comparing the present scapula-coracoid
with the clavicular apparatus of Portheus.

Measurements of scapula-coracoid :

Actual length of ‘spect... eee ees 33.7 cm.
Estimated length of element..................... 36

Width of upper end of scapula................... 15.5
Thickness at Upper @M <os¢ 57ers +s -+++-+-.-.s. 3 mm.
Width at middle of saates. 23. an cue es ss sss. s 6.6 cm.
Width across coracoid portion.................... 17

= |

eae
*, a

MOODIE: PERMIAN ANPHIBIA. 251

Rength: of cotylus ..\. sci Paeeew owed cabes as Wiss ci 7:2
Widen Of cotylus ....0.. ses:04 saebineee se Sia opis kites 4 3.8
Diameter of infraglenoid foramen oe a aoe arpa 5 mm.
Thickness of bone at lower edge.............2+05- 7
Thickness of bone through cotylus................ 4.3 cm.
Length of portion of cleithrum preserved.......... 63 mm.
Greatest width of cleithrum .......:............. 22

The clavicle of the right side is preserved almost completely
(plate LIII, figs. 1 and 2). The element was figured by
Doctor Williston but not described. It is strongly curved and
spatulate. Its outer surface is ornamented with longitudinal
grooves, which arise near the center of the bone and run each
way, thus indicating its ossific center. The inner surface of
the bone is smooth excepting for two vascular pits near the
center. A prominent ridge runs along the dorsal inner edge
of the bone.

Measurements of the clavicle:

SR MARE CURVES or iacd i's ss fies deadeinig » 6b, ddig oe ond se 24.7 cm.
I NEI AU a 0 sy ap big boc a Gas Heb oe 8 ene ae 20

ES WERE s eg Pa si eee ek be SUS a ee es ola es 1.3
NE RRMEAR © isk faa. ove ga 509 Waa Ka ace ew Ds ve hie'e si 4.6
OS 6 5g W's. 0.6 echbaie vis Sev whe mae 1
TIEMIORMONE NG rep Oe ore Saks oboe evcveveves “2 | TAM,

There is a fragment of what I regard as the posterior

‘projection of the interclavicle. If it is such it is either a

pathological specimen or there were two projections from the
posterior edge of the interclavicle, for the fragment is highly
asymmetric. The spine, for such it is, has an irregular nodose
surface as if for the attachment of cartilage. It is long, slender
and somewhat flattened. (Plate LIII, fig. 3.)

Measurements of spine of interclavicle:

ET Oe SOOUMNONE oa owl cc ceicravavaercsvemned 13.7 cm.
EN Es be agin While « hae mees 2.2
I ng I re. a 5 uo you oc tse 0 to oes «15

There are remains of about a dozen vertebre preserved.
Portions of three of these are figured in plate LIV, figures 1
and 2. They all have the characteristic rachitomous form as-
sumed by the other species of Eryops. The divisions between
the pleurocentra, hypocentra and neurocentra are well marked.
The zygapophyses are well formed. The upper part of the
neural spines is much thickened, as is the case often in speci-
mens of Hryops megacephalus Cope. According to Broili the
spines may also be bifurcate, as Cope has described for LE.
erythrolithicus. The enlarged spine was undoubtedly for the
attachment of muscles to sustain the enormous head, and

252 KANSAS UNIVERSITY SCIENCE BULLETIN.

spines showing different characters might very easily occur
in various portions of the vertebral column.
Measurements of vertebre :

Length of centrum. ......csceseereececeseeegsanes 2.5 cm.
Width of centram (3208. i scuestu cee ere eee 3.2
Height of vertebre to tip of spine................ 13.4
Thickness of top of: spine. 7.0.5: su ¢ccee. cabs SO 3

Broili (Monatsb. d. Deutsch, Geol. Gesell., Bd. 60, Texttafel
p. 236, fig. 2) has figured in Hryops megacephalus a peculiar
element which he has identified as the sacral rib, having found
it in place with a sacral vertebra. There is no doubt that the
present element (plate LIV, figs. 3 and 4) is the same as he
has figured. Its interpretation as a sacral rib is somewhat
surprising on account of the extreme thinness of the bone. It
thins out to less than a millimeter on one edge. Its location
-any other place in the skeleton is, however, doubtful.

The element has a broad base as if for attachment suturally
with the vertebral transverse process. The base of the speci-
men has been broken transversely and only a portion of the
suture remains. The bone is flat immediately from the base
and is greatly expanded and thinned distally. It is, for the
most part, smooth, though there is a small roughened place on
the dorsal (?) surface. The ventral (?) surface is marked by
a depression which runs the entire length of the bone. The
distal end is lost.

Measurements of sacral rib:

Actual length of specime?i... § 66) 0tASs. .wavwee Oise He 90 mm.
Width -at distal end... is.seueip suk sk oe ee ee 62
Width at. middle of Blade... 32054 i reas. Fe ee 32
Width. at. inner endsics Sea eee ees oe eee 55

Eryops willistoni differs from EF. ferricolus Cope in the
absence of a raised rim to the orbits. From E. erythrolithicus
Cope it differs in the non-bifurcation of the neural spines.
This last is not a well-known species and may be synonymous
with E. megacephalus. E. reticulatus is a small species, from
which EL. willistoni differs in the character of the cranial sculp-
ture. It is distinct from E. latus Case on account of the di-
mensions of the scapula and the cleithrum. EF. megacephalus
also shows characters which are divergent from those of the
present species. These are the shape and structure of the
mandible and its ornamentation, the form of the humerus and

MOODIE: PERMIAN ANPHIBIA. 253

and structure of the scapula-coracoid, as well as
1 uctures which are characteristic.

N. Gould in 1896, in the red sandetiike southwest of Black-
, Okla. ~The specimen is No. 348 of the Kansas University

Vol. V, No. 14— March, 1910..

(Whole Series, Vol. XV, No. 14.)

a. =, CONTENTS:
‘DINOSAUR FROM THE UPPER CRETACEOUS OF WYOMING, _
ip gaa eee e Roy L. Moodie.

- PUBLISHED BY THE UNIVERSITY,

.

in Lawrence as second-class matter.

THE KANSAS UNIVERSITY
SCIENCE. BULLETIN.

* No. 14 | MARCH, 1910. Pens Santee

OF WYOMING.
BY ROY L. MOODIE.

* (Contribution from the Zotlogiesl Laboratory, No. 190.)
Plates LV to LIX, and one text figure.

ketch. of the horizon scons which the specimen was
naming the horizon Hailey shales, and suggesting

nearly every bed. They are often so
: (257)

958 KANSAS UNIVERSITY SCIENCE BULLETIN.

abundant as to number from five to ten on a surface of six
inches square. Owing to their hardness the beds give rise to
round-topped ridges of considerable prominence.”

Fic. 1. A typical exposure of the Hailey shales on Connant creek, Wyo-
ming. The Mowry beds may be seen on the left in the background.

In those parts of Wyoming where I worked, the Mowry beds
stand out as prominent ridges of hard whitish shales which
can be readily recognized from a long distance, and on that
account serve as excellent landmarks for the identification of a
region where the Hailey shales may occur. In thickness the
Hailey shales vary from a few feet north of Lander, Wyo., to
something over 200 feet in the Big Horn region. “Near Soap
Creek the dark shales overlying the Mowry beds are about 200
feet thick,” etc. (6). In most of their extent the Hailey beds
have numerous brown sandy clay concretions which very often
contain vertebrate remains. These concretions vary in size
from a few inches to many feet in diameter. In the Big Horn
region the concretions are stratified into distinct layers, but

such is not the case in other places where the horizon was |

examined.

So far as is at present known the Hailey shales outcrop along
the eastern flank of the Wind River mountains, in the Shoshone
anticline; in the anticline along Beaver creek near Hailey,
Wyo., the type locality; in an anticline north and west of the

* Paes +e;

-MOODIE: ARMORED CRETACEOUS DINOSAUR. 259

in: te mountains, on Connant creek; in a small anticline west
Rattlesnake mountains; along the northern and eastern
flanks of the Rattlesnake mountains. There are extensive ex-
Pe ures on the eastern slope of the Big Horns, and I do not
loubt that they occur also to the west of these mountains. In
_text-figure 1 is given a view of a typical exposure of the
Hailey shales, on Connant creek, north of the Granite moun-
ins, Wyoming. On the left of the photograph the fydeit
beds stand out in the characteristic ridge.
ie vertebrate fauna of the Hailey shales still remains to be
sately determined, but so far as known it is represented
large and small fishes, amphiccelian crocodiles with hollow
bones, turtles with completely ossified carapaces and
ra like Toxochelys and plesiosaurs of two types. The re-
of plesiosaurs are in places exceedingly abundant, evi-
of no less than fifty-five individuals having been dis-
ed during one summer. Occurring in these deposits, also,
been discovered the remains of three dinosaurs. The re-
described below represent one of these animals. The

"The Stegopelta remains herein described were once encased
1 one of the brown sandy clay concretions so abundant in this
nation, but many years ago it was exposed and by the
sthering processes of-a long space of time had become re-
hundreds of fragments. These fragments had washed

ing up some of the fragments came across some
aurian dinosaur teeth, the whole aspect of the

od the specimen weighed only 450 pounds, of
thirds was bone. It has been a long and
fit together the scattered and disassociated
rough the combined efforts of Doctor Willis-

260 KANSAS UNIVERSITY SCIENCE BULLETIN.

ton and myself a number of elements of the skeleton have been
reconstructed.

The specimen as it now stands consists of the following
parts: Fragments of the upper and lower jaws with a dozen
or more teeth, most of which are fragmentary; seven frag-
mentary dorsal vertebree; portions of the sacrum; two caudal
vertebree; many fragments of the ribs; a complete ulna be-
longing to the left side; portions of the right ulna; portions of
the radii; the left ilium nearly entire; a part of the right ilium,
showing the acetabulum; the right pubis nearly entire; a por-
tion of the left pubis; the lower portion of the right tibia; the
upper and lower ends of the right fibula; many large fragments
of the metatarsals; nearly a hundred small bony scutes; several
large bony plates, girdles and bony spines representing va-
rious portions of the dermal armor; together with a great
quantity of waterworn fragments which it is impossible to
determine.

The remains described as Stegopelta landerensis indicate a
stegosaurian dinosaur about two-thirds the size of Stegosaurus
ungulatus Marsh. The head was small, as is indicated by the
size of the teeth. The fore limbs were not relatively so large
as in Stegosaurus. The entire pelvic region was covered with a
dorsal shield of rather thick, scrobiculate plates with elevated
centers which are asuturally united over the pelvis. On each
plate there is found, either on one side or in the central emi-
nence, a pit, which possibly supported a more or less elongate
horny spine. Anterior to the pelvic shield there were many
dermal plates and scutes of various sizes scattered over the
dorsal region of the body. The anterior portion of the animal
appears not to have been armed. Back of the pelvis there were
massive dermal plates, huge bony spines and a bony girdle
probably encircling the tail in some such manner as in the
glyptodonts. The neural spines of the caudal vertebrze were
probably each surmounted by an elongate bony plate which
quite. possibly bore a horny excrescence of some kind. While
closely similar in many respects to Polacanthus, Stegopelta
shows wide structural differences from that form, although it
must have had much the same appearance as Nopsca (5) has
figured in Polacanthus. The present dinosaur, in life, may
have stood nearly seven feet high at the hips and attained a
length of sixteen to eighteen feet. Armed as it was with horns,

-

MOODIE: ARMORED CRETACEOUS DINOSAUR. 261

spines and excrescences it must have been a bizarre creature,
even in those Cretaceous days.

The teeth (plate LVI, figs. 4 and 5) of Stegopelta are typi-
cally stegosaurian. There are a dozen of them in more or less
perfect state of preservation. They are all of practically the
same size and character, showing but little variation among
themselves. The crowns of the teeth, which measure 7 mm. in
breadth, are subcompressed and narrowly palmate with eight
radiating coste, all of which end at the tip in sharp points.
The roots of the teeth measure 15 mm. from the base of the
crown to the tip of the fang. In point of size and number of
costz the present teeth are so closely similar to those described
by Leidy (5) as a supposed lacertilian form, Paleoscincus,
that had the teeth only been found they would undoubtedly
have been referred to that genus. It is possible that Stegopelta
may eventually prove to be a Paleoscincus form. This, how-
ever, is hardly probable, on account of the great interval of
time separating the strata which have yielded the two speci-
mens, Paleoscincus occurring as it does near the top of the Fort
Pierre Cretaceous and Stegopelta near the top of the Fort Ben-
ton. Were the form one of a stable group this might not be
sound reasoning, but in a group of animals which evolved so
rapidly as did the dinosaurs one would hardly expect a genus to
continue for so long atime. At all events, the name Stegopelta
may be retained until more evidence is forthcoming concerning
Paleoscincus. If, as Hatcher believed (23), Lambe’s Stereo-
cephalus is a Paleoscincus form, then Stegopelta is a good
genus, since it shows wide structural differences from the form
which Lambe has described. The tooth which Lambe (7) has
figured on page 57 certainly resembles those of Paleoscincus,
and if the tooth really is to be associated with the remains
described as Stereocephalus a question might arise as to
whether these remains were not in reality portions of the skele-
ton of Paleoscincus as Hatcher intimated. Lambe (8) says
the tooth provisionally associated with the Stereocephalus
remains “differs from those of the Red Deer river district, re-
ferred to the two species of Paleoscincus, and is about twice as
large as those of P. costatus.” Just how it differs he does not
Say, and one is at a loss in attempting to separate the forms
from his figures. Size alone is usually no safe criterion for the
separation of animal forms, and more particularly genera. I

262 KANSAS UNIVERSITY SCIENCE BULLETIN.

can but agree with Hatcher that the tooth figured as Steveo-
cephalus resembles the teeth of Paleoscincus.

The vertebral column is represented by posterior dorsal ver-
tebree, portions of the sacrum and two caudal vertebre. The
remains are more or less fragmentary, but enough is preserved
to give a fairly accurate idea of this portion of the animal’s
anatomy, although it is impossible to determine the vertebral
formula.

The dorsal vertebre (plate LVIII, figs. 5 and 6) obtained are
twelve in number, for the most part imperfectly preserved.
They are cylindroid in form, very slightly amphiccelous, nearly
amphiplatyan, and the arches are elevated as in Stegosaurus
and Polacanthus. The centra are slightly excavated below the
base of the neural arch as in Scelidosaurus. The neural canal
is small and is wider in the ventral than in the dorsal portion.
The centra measure 60 mm. in width by 70 mm. in length.

The sacrum (plate LIX, figs. 6 and 7) as preserved consists

of two sections. One includes three firmly codssified vertebre
of the posterior end and the other a part of the anterior end
representing two vertebre. The anterior section measures
120 mm. in length by 70 mm. in width at the widest part.
The fragment is very smooth and of a dense bony texture.
The neural canal is broad posteriorly but narrows somewhat
anteriorly. In plate LIX, figure 2, is represented a cross
section through this fragment to show the sacral enlargement
at the posterior end. The portion representing the posterior
end of the sacrum (plate LIX, figure 7) measures 203 mm. in
length by 64 mm. in width at the caudal end. The texture of
this part is very fine and compact externally, but internally
there is a tendency to a cancellated structure. An interesting
characteristic of this element of the skeleton is the high degree
of codssification of the various elements composing it. There
is not the slightest indication of a suture where the different
vertebree have become joined so that the sacral centra are con-
stituted in a solid bony rod tapering somewhat at the anterior
end, as do the sacra of Polacanthus (9) and Hylzxosaurus
(10). The attachments for the sacral ribs are large, and are
placed about midway dorso-ventrally of the centra of the sa-
crum. While the sacrum of Stegopelta resembles in a great
measure that of Polacanthus, yet it varies from it in that the
elements composing it are more highly codssified. This condi-

7

MOODIE: ARMORED CRETACEOUS DINOSAUR. 263

tion cannot be entirely attributed to the age of the individual
but is expressive rather of the higher specialization of the pres-
ent form.

Attached to the neural spines of the sacral elements was a
long, narrow dermal plate (plate LVI, fig. 3), evidently in firm
union with them. This plate occupied the mid-dorsal line of
the great pelvic shield and lay between the iliac armor of the
two sides. The plate, like all the rest of the dermal skeleton, is
scrobiculate and somewhat of the texture of woven cloth, as
Marsh has described in Nodosaurus textilis (11). It becomes
slightly expanded posteriorly, where it measures 30 mm. in
width.

The caudal vertebrx (plate LIX, fig. 1, and plate LVIII,
fig. 3) vary to a remarkable degree from those of the dorsal
series. They are greatly compressed dorso-ventrally; are very
short, with the sides much rounded and the centra plano-con-
vex. The vertebre have strong transverse processes. In
neither specimen is the neural spine preserved, so it is im-
possible to say as to its extent. This, however, may be inferred
to have been low and stout, if one may judge from its broken
bases. The neural canal is not large, measuring only 14 mm. in
diameter. The centra measure 82 mm. in width by 38 mm. in
length by 41 mm. in height. It is thus seen that the caudal
vertebre are more than twice as wide as long. This condition
is not a unique one, since Hulke (12) has described the same
condition in Polacanthus.

The ribs are all so fragmentary that no discussion of them
can be given. Many of the fragments are triangular in cross
section, agreeing in this respect with other armored dinosaurs.
Attached to some of the rib fragments are found small bony
seutes. The fact is of service in determining the position of
these dermal scutes on the body of the animal.

The ulna (plate LIX, figs. 8 and 9) of the left side is pre-

_ served complete. It shows considerable variation from any-

thing hitherto described for any other armored dinosaur. The
shaft of the bone is flattened from side to side and is twisted
from right to left. It ends in a rather pronounced expansion
distally. The coronoid process is quite prominent and is ex-
cavated to form a large sigmoid cavity. The olecranon is long,
measuring 70 mm. in length from the base of the coronoid
process to the proximal end of the bone, It is flattened from

264 KANSAS UNIVERSITY SCIENCE BULLETIN.

side to side and ends rather broadly, not in a point as does the
ulna of Stegosaurus. Viewed from in front the ulna curves
strongly outward, especially toward the distal end. It is not
so straight as in Stegosaurus. The articular surface for the
humerus is small. The radial articulations are, however, well
marked, with the upper one forming a distinct fossa below the
coronoid process. The lower articular surface of the ulna is
well rounded and is oval in outline. The ulna measures 385
mm. in total length, 106 mm. antero-posteriorly at the widest
part, 45 mm. near the middle of the bone and 60 mm. at the
proximal end.

The radius is represented by fragments only. The head of
one of the elements is nearly circular in outline and measures
60 mm. in cross section. The articular surface is so abraded
that it is impossible to determine its form.

The ilia (plate LV) are both represented in the collection.
The left one is preserved nearly entire. Enough is present to
determine the total length and breadth of the bone, and the
missing parts have been restored, as is evident from the photo-
graph. The portion of the right ilium preserved shows the
acetabulum. The description of the ilium will, however, be
confined to that of the left side. The bone is elongate and
slightly convex upwards; the convexity becoming strongly
pronounced on the outer side. It is 50 mm. thick at the acetab-
ulum, but thins to 5 mm. on the inner side and at the anterior
end. The thinning is gradual in these two directions. On
the outer edge, however, the bone remains thick for over
three-fourths of its length, when it gradually thins, but not so
much as on the inner edge. The entire dorsal surface of the
ilium is covered by a mosaie of thick, bony, scrobiculate plates
firmly united to it. These plates are more or less pentagonal
in shape and are not all of the same size. On nearly all there
occurs, on some in the center, on others near the edge of the
plate, a pit, which in life may have supported a more or less
elongate horny spine. When the pit is central it is located in
the top of the central eminence which occurs on all of the plates.
The plates were once, apparently, suturally connected, but
there is now- no indication of a suture. They form one con-
tinuous bony mass over the entire dorsal surface of the ilium.
The diameter of the individual plates varies from 56 mm. to 105

en — - ~s » ——
7 . Le
‘ of Ses :
.

MOODIE: ARMORED CRETACEOUS DINOSAUR. 265

mm., and they are all from 5 mm. to 15 mm. in thickness. A
single plate is shown natural size in plate LVI, figure 8.

The lower surface of the ilium is nearly smooth. The only
irregularities are the acetabulum and small ridges for muscular
attachments. It is gently concave, the concavity becoming
stronger towards the outer edge. The acetabulum is in the

_ shape of an elongate oval and is defined by a high ridge of

bone. The acetabular cavity is rather shallow and is deeper
anteriorly than posteriorly. Its length is 190 mm. and the
width, as preserved, is 82 mm. It may have had a somewhat
greater width, but not to exceed 90 mm. at the most. The
entire ilium measures 4814 inches in total length, 1014 inches
at the widest part, and 814 inches near the middle of the bone.

The ilium must have had a horizontal position during life,
for the position of the acetabulum is such that the femur could
not have articulated in it had the bone been in any other
position. This condition is not exceptional among dinosaurs,
since the ilia of Polacanthus had a similar position and the
ilia of Triceratops are horizontal to a much less extent. The
articular surfaces for the pubes and ischia have been broken
away and lost, so it is impossible to determine the relations
of these elements. However, they must have had much the
same relations as Seeley (13), Lydekker (17) and Hulke (14)
have described for Polacanthus.

The pubes (plate LVIII, figures 1 and 2) are both repre-
sented, the right one nearly complete but lacking the entire
post-pubic process. This element resembles more nearly that
of Claosaurus (15) than it does that of its nearest allies,
Polacanthus (13) and Stegosaurus (16). In form the bone is
elongate, flattened and spatulate. It curves gently inwards
and is only slightly twisted. In thickness the pubis is nearly
uniform, becoming thickened proximally to support the pec-
tineal process (ectopubis of Baur). The pubis measures 300
mm. exclusive of the pectineal process, which is 57 mm. in
length. In width the bone measures 112 mm. at the widest
part and 86 mm. at the distal end. The external surface of
the pubis is crossed by a diagonal ridge for muscular attach-
ments. The post-pubis is not represented in the collection
Save by a fragment which may represent the disal end. The
place where the process has been broken from the pubis is
large, so thé post-pubis may have been of some extent,

266 KANSAS UNIVERSITY SCIENCE BULLETIN.

The tibia (plate LVIII, figure 4) is represented by the lower
end of the bone belonging to the right side. Firmly coéssified
with it is the astragalus, which rises into a rather prominent
ascending process on the anterior surface. The union of the
astragalus with the tibia is so complete that were one to view
the specimen from the posterior side its presence would never
be suspected. On turning the bone over, however, the astraga-
lus is distinctly seen, as is shown in the figure. The spine of
the astragalus is broken off, but judging from the groove for
its reception it must have been nearly 80 mm. in length. It”
measures 50 mm. in width at the base. The tibia measures
145 mm. in width distally. The articular surface is broad and
pulley-shaped. There is a broad, shallow, trochlear groove.
which is bounded by two rounded projections. The groove
runs up prominently on the posterior surface of the bone but
is not so prominent on the anterior surface. From the nar-
rowness of the broken surface of the bone, only 50 mm. in
diameter, one would judge that the shaft of the tibia was very
slender as described by Hulke (18) in Hylzxosawrus and in
Polacanthus (19). It resembles these forms also in that the
shaft of the tibia is triangular in cross section, forming in
this case a nearly isosceles triangle. The section shows a
marked difference between the exterior of the bone, which is
smooth and compact, and the interior, which is coarsely can-
cellated.

The fibula (plate LIX, figures 4 and 5) is represented by the

proximal and distal portions of the bone belonging to the right —

side. The proximal end, which is subtriangular in cross sec-
tion, measures 85 mm. in width and is considerably expanded
as compared to the shaft. The upper end is so weathered that
no traces of the rugose surface for the attachment to the tibia
are left. However, one side is flattened as if for such a union.
The external side of the head is marked by a rather prominent
condyle, which projects some 5 mm. from the surface. The
shaft of the bone is very slender. At a distance of five inches
from its upper end the fibula measures 20 mm. by 30 mm. in
diameter being thicker antero-posteriorly than from side to
side. In structure the fibula is much like that of the tibia,
already described. The distal end, also, triangular in cross
section, is of the same proportions as the proximal end. ©

The metatarsals (plate LVIII, figure 7) are represented by

‘ =
| ne

MOODIE: ARMORED CRETACEOUS DINOSAUR. 267

several fragments. The proximal and distal ends of one of
them are preserved so that a reconstruction is possible. These
fragments represent a rather elongate, slender element with
the upper surface rounded and the lower surface somewhat flat-
tened. The proximal articular surface is widely concave and
the distal end is pulley-shaped. It is thus evident that the meta-
tarsal is not different from that of other herbivorous dinosaurs.
The length of the bone as restored is 150 mm., the breadth of
the proximal end is 80 mm., and the height 60 mm.

The dermal armor consists of small bony scutes, large dermal
spines and heavy plates, all of various shapes and sizes.
Mingled in with the other bones of the skeleton were more
than four score small bony scutes (plate LVII), varying in
diameter from 15 to 75 mm. and of various thicknesses from
4 to 20 mm. The scutes are all scrobiculate and are for the
most part rounded, though some are elongated, one scute be-
fore me measuring 60 mm. in length by 30 mm. in breadth.
The majority of them possess a ridge asymmetrically placed,
which runs nearly the entire length of the scute. Others run
up to a point in the center; others are flat.

In structure all of the scutes are of a coarsely cancellated
character, such as Hulke has described for the scutes occur-
ring in Polacanthus. Associated with them were many frag-
ments of heavy triangular ribs with which some of the scutes
were firmly united. From this fact it would appear probable
that the scutes were scattered along the back of the animal an-
terior to the great pelvic shield. The elongated scutes re-
ferred to above may possibly have been situated on the neural
spines of the vertebre, as in Scelidosaurus (20). The larger
‘scutes, some of which are as large as the palm of one’s hand,
were very probably located near and anterior to the great
pelvic shield.

Scutes very similar to these described for Stegopelta are
found in a number of different armored dinosaurs as well as
among the crocodiles. There are really no structural differ-
ences between the scutes of Stegopelta herein described and
those found in the common Mississippi alligators. The pittings
on the Stegopelta scutes are, however, smaller than in the latter.
Hulke (21) has described a few scutes belonging to Polacan-
thus which resemble the present specimens very much. Lambe
(22) has figured three scutes associated with the form Stereo-

268 KANSAS UNIVERSITY SCIENCE BULLETIN.

cephalus which are almost identical with the scutes of Stego-
pelta. Hatcher (23) was certainly in error in criticizing Lambe’s
association of these scutes with dinosaurian remains and in
claiming that they were of a crocodilian nature, since in the
present specimen there can be no doubt as to their dinosaurian
character. Nopsca (24) has also fallen into the same error
as did Hatcher when he says that crocodilian remains were
associated with the Polacanthus specimen. Seeley (25) has
described and figured scutes similar to the above in Crateomus.
Huxley (26) has described and figured identical scutes in Acan-
thopholis. Marsh (27) has figured some small bony scutes in
Triceratops. Owen has figured scutes of a like nature in
Scelidosaurus (20). Possibly other armored dinosaurs pos-
sessed such scutes.

The question of the morphological value of these scutes is
an interesting one. Judging from the scutes in Stegopelta, one
would incline to the opinion that they represent the rudiments
' of-a complete bony armor which in time, had the dinosaurs
survived, might have covered the whole dorsum of Stegopelta’s
descendants. Such a suggestion might apply equally well to
our modern crocodiles. It is pretty certain that the scutes in
the crocodiles are not vestigial, since it is not now conceded
that the modern Crocodilia are descended from the ancient
armored teleosaurs (28).

The dermal spines (plate LVI, fig. 1, plate LIX, fig. 3) are
all of a dense bony texture with the external surface pitted.
A portion of a small dermal spine is represented, natural size,
in plate LIX, fig. 3. Its base is broken away so it is impossible
to determine its extent, but this was, in all probability, not
considerable. One of the most peculiar portions of the dermal
armor is that represented in plate LVI, fig. 1, which is called
a large dermal spine, and which I have provisionally located
near the base of the tail. The spine is asymmetrical and is
lateral in position, having had a mate on the opposite side of
which nothing remains.

The spine preserved is much broken and the distal tip is
lost. As reconstructed it is of massive proportions in com-
parison to the rest of the exoskeleton. In cross section it is
for the most part triangular, with two sides of the triangle
concave and one convex. It is rounded near the distal end.
The spine, unlike most elements of a similar nature, is bifid,
as is represented in the figure (plate LVI, fig. 1). The short

MOODIE: ARMORED CRETACEOUS DINOSAUR. 269

spine measures 80 mm. from the point of its inception to the
tip, and the main spine measures 16 inches from tip to tip.
At its widest part it is 100 mm. in breadth. The surface of the
bone is pitted with irregularly placed cavities as though it were
encased with horny substance.

The exact position of this large spine on the body of the
animal is, of course, largely a matter of conjecture, since it was
found broken and disassociated and mingled with various
other parts of the skeleton, but it seems best to locate it back
of the pelvic shield, since the caudal vertebre are much stouter
than the dorsal ones and since there is no evidence of any armor
excepting the small bony scutes anterior to the pelvis. The
spine is totally unlike anything described for any other stego-
saurian dinosaur.

The dermal plates, represented in plate LVI, figs. 2 and 9,
are massive and of a compact texture. Their position on the
skeleton is a puzzle. The plate represented in plate LVI, fig.
9, is not entire. It is asymmetrical, one side being crossed by a
rather high ridge. Its upper surface has pittings and grooves.
The under surface is smooth and rounded. It is possible that
this is a portion of another girdle similar to the one described
below. That it is not a part of that girdle is evident from its
proportions. The plate measures 105mm. in diameter. The
other dermal plate (plate LVI, fig. 2) is apparently complete,
with a broad surface for articulation with some element which
is now lost. This plate is concavo-convex. It measures
140 mm. across in the longest diameter. It is without vascular
markings of any kind, but is smooth as though it were entirely
embedded in the flesh of the animal. ;

The part of the girdle represented in plate LVI, figs. 6 and 7,
is like the last described plate in that it is of a compact texture.
In the girdle the upper surface is pitted, and crossed at regular
intervals by rather high ridges the sides of which are slightly
concave. This bone at once reminds one of the similar ele-
ment figured and described by Lambe (29) as occurring in
the Stereocephalus form. Lambe (29) however, located the
girdle back of the head, and if his interpretation of the large
fragment which he calls the skull is correct such a location
would be quite plausible. The present girdle differs in a great
degree from that described by Lambe, since it is asymmetrical
and had a mate on the opposite side, of which there are frag-
ments in the collection. It differs further in being a solid mass

270 KANSAS UNIVERSITY SCIENCE BULLETIN.

of bone instead of a bony ring to which the scutes are united.
The girdle as preserved is in the form of an arch and is com-
posed of three plates, only two of which are ridged. The under
surface is smooth and longitudinally convex. The part of the
girdle preserved measures 320 mm. in length and 110 mm. in
width.

Stegopelta finds its nearest allies in Polacanthus, from the
Wealden of the Isle of Wight; in Paleoscincus, which must, I
think, include the Stereocephalus of Lambe, from the Belly
river deposits; and Stegosaurus, from the Lower Cretaceous of
Wyoming and Colorado. The characters which separate Steg-
opelta from Stegosaurus and Paleoscincus have already been
considered. From Polacanthus the present form differs in the
whole surface of the ilium being covered in the former animal
with a firm mosaic of very small, somewhat rounded, plates,
which are scarcely at all sculptured, as I learn from Doctor
Williston, who has examined the type specimen. These plates
_in Stegopelta are relatively quite large and strongly sculptured.
The present form differs from Polacanthus also in the diver-
gence of the points of the ilium anteriorly. In Polacanthus
the iliac region was entirely covered by a bony shield, but in
Stegopelta this shield was not continuous, but there was a space
between the anterior ends of the ilia which was filled, probably,
with small bony scutes. There have been many stegosaurian
dinosaurs described from various parts of the world, notably
North America and Europe. The group seems to have had a
very wide distribution geographically.

In conclusion, I wish to express my gratitude to Dr. S. W.
Williston, not only for his kindness in turning over the material
to me for description but for his kindly interest and help during
the entire investigation.

The specimen is in the Walker Museum, University of Chi-
cago.

Since the above was written there have appeared two papers
describing two new types of dinosaurian reptiles closely allied
to the above-described form. One of these papers,* from the
American Museum of Natural History, is of especial interest,
since the material supplements in an excellent way that which
has just been described. Mr. Brown’s restoration of the Anky-
losaurus magniventris, while based in large part on Marsh’s

* Brown, Barnum, 1908. Bulletin Amer. Mus. of Nat. Hist., vol. XXIV, p. 187.

MOODIE: ARMORED CRETACEOUS DINOSAUR. 271

drawing of Stegosaurus ungulatus, is not accurate as to the
form of the animal. The pelvic region was undoubtedly cov-
ered over with a heavy shield as in Stegopelta. Indeed,
Doctor Williston} says, “Ankylosaurus is either very closely
allied to or identical with Stegopelta Williston, a genus over-
looked by Mr. Brown.” ,

The remains described by Mr. Brown consisted of a skull,
some small dermal scutes, a few vertebre, a scapula, a few
ribs and other parts, most of which were lacking in the above-
described animal.

Dr. O. Abel has given (Verhandl. d. k. k. Zo6l.-bot. Gesell. in
Wien, Bd. LVIII, p. 215, fig. 3) a restoration of Ankylosaurus
magniventris Brown. It is to be doubted if he has added to our
knowledge of dinosaurian attitudes, since the reader is at once
struck with the grotesque and impossible attitude which he has
caused the animal to assume. It is unfortunate that Brown’s
ideas of the arrangement of the scutes on the back should have
been given by Doctor Abel, since there is not the slightest evi-
dence that they assumed anything like the arrangement he has
represented.

The other dinosaur is from the Niobrara Cretaceous of Kan-
sas and was collected by Mr. Charles Sternberg in the Hack-
berry creek region.

Doctor Wielandt describes these remains as Hierosaurus

- sternbergii, new genus and species. This form is also doubt-

fully distinct from Stegopelta, unless it be that the skeletal
characters were widely at variance to the characters of the der-

. mal plates. A comparison of figure 9, plate LVI, of the present
essay, with Doctor Wieland’s figure 6, will show how very

closely the dermal elements of the animals resemble each other.
Comparisons also of the figures on plate LVII of the present

_ paper with Doctor Wieland’s figures 2, 4 and 5 will reveal the
_____ @lose similarity of the small dermal elements. It would, how-
____ ever, be too previous to say certainly that Hierosaurus and -
_ Stegopelta are identical. The geological horizon of the two

_ specimens is certainly not far apart, or perhaps identical, since

the Hailey shales occupy relatively the same position in Wyom-
ing as the Niobrara chalk does in Kansas.

7 American Naturalist, vol. XLII, No. 501, p. 629.
¢ Wieland, G. R., 1909. Amer. Jour. Sci., March, p. 250.

2—Univ. Sci. Bull , Vol. V, No. 14.

}
5

272 KANSAS UNIVERSITY SCIENCE BULLETIN.

BIBLIOGRAPHY.

1. Wituiston, S. W. A New Armored Dinosaur from the Upper Creta-
‘ceous of Wyoming. Science, N. S., vol. XXII, No. 564, Oct. 20,

1905, p. 508. » E
2. News item, SCIENCE, N. S., vol. XXIV, No. 622, Nov. 30, 1906, p. 711. x 4
3. Darton, N. H. Geology of the Owl Creek Mountains. Senate docu- aM

ment No. 219. Washington, D. C., 1906. Page 22.
4. Lewy, JosePpH. Notice of Remains of Extinct Reptiles and Fishes ae al
Discovered by Dr. F. V. Hayden in the Bad Lands of the Judith a
River, Nebraska Territory. Proc. Acad. Nat’l Sci., Phila., vol. ; 5
VIII, pp. 72-78. ;
5. Lewy, JOSEPH. Extinct Vertebrata from the Judith River and Great 3 {
Lignite Formations of Nebraska. Trans. Amer. Philos. Soc., vol. \e
XI, pp. 139-154, with plates VIII-XI. 1860. See 4
6. Darton, N. H. Geology of the Big Horn Mountains. Professional
Paper No. 51, p. 55. Washington, D. C. 1906.
7. LAMBE; L. M. Contributions to Canadian Paleontology, vol. III,
pt. II: On Vertebrata of the Mid-Cretaceous of the Northwest 1
Territory. Art. No. 2: New Genera and Species from the Belly *
River Series. Geological Survey of Canada. 1902. ; “¥
8. Contribution to Canadian Paleontology, vol. III, p. 57. ae i \
9. HULKE, J. W. Phil. Trans., 1881, p. 654, pl. 71, fig. 1. : :
10. OWEN, RICHARD. Monograph of the Fossil Reptilia of the Wealden
Formation. Paleontographical Society, 1857, pl. V. peg
11. Marsu, O. C. The Dinosaurs of North America. 16th Ann. Rep. j
U.S. G. S., 1896, pl. LX XV, fig. 5.
12. HuLKE, J. W. Phil. Trans., 1881, p. 655. . ;
13. SrELEy, H. G. On the Os Pubis of Polacanthus foxii. Quart. Jour.
Geol. Soc., vol. 48, pp. 81-95, pl. I. :
14. HuLKE, J. W. Phil. Trans., 1887, p. 171, pls. 8 and 9. a:
15. Marsu. Dinosaurs of N. A., 1896, pl. LXXIX, fig. 5. Ki
16. Ibid, fig. 6.
17. LYDEKKER, RICHARD. On Part of the Pelvis of Polacanthus. Q. J. G.
S., vol. XLVIII, pp. 148-149, with 2 figures.
18. Hu.ks, J. W. Q. J. G. S., vol. 30, pp. 516-518, pl. XXXI, figs. 1, 2.
19. Hutxke. Phil. Trans., 1881, p. 657.
20. RICHARD OWEN. Monograph on Scelidosaurus, pt. II, Paleonto-
graphical Society, 1862, p. 20, pl. IX.
21. HuLKke. 1881. Phil. Trans., p. 657, pl. 75, figs. 5, 6.
22. LAMBE. 1902. Contribution to Canadian Paleontology, vol. III, p. 56,
pl. XXI, figs. 6, 7, 8.

| MOODIE: ARMORED CRETACEOUS DINOSAUR. 278

pom J.B. 1905. Vertebeeie Weunn: of the Judith River Beds.
"Bull. U. 8. G. S., pp. 67-108.

* 1896. Biases of North Sasi, ot LXX.
8. me ‘Aiperican ee Crocodiles, Jour. Geol.,

i

CE BULLETIN.

(Whole Series, Vol. XV, No. 15.)

CONTENTS:

To THE Sort ANATOMY OF CRETACEOUS FISHES
” PRIMITIVE HERRING-LIKE FISH FROM THE TEXAS
ae et Pens . . . . . Roy L. Moodie.

PUBLISHED BY THE UNIVERSITY,
LAWRENCE, KAN.

October, 1911.

in Lawrence as second-class matter.

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vou. V, No. 15.] MARCH, 1910. You XV, Ho. 16

>. A CONTRIBUTION TO THE SOFT ANATOMY OF CRE-
TACEOUS FISHES AND A NEW PRIMITIVE
HERRING-LIKE FISH FROM THE
TEXAS CRETACEOUS.
BY ROY L. MOODIE.
(Contribution from the Zodlogical Laboratory, No. 191.)
Plates LX to LXII.

rT HERE have been, during the past few years, several ad-
ditions to our knowledge of the soft parts of extinct
animals. This knowledge has to do, in large part, with

the firmer tissues, such as the cartilaginous portions of the
skeleton, the skin, the muscles, but in some cases the kidneys,
oviducts, nervous tissues, blood vessels and alimentary canal
are clearly preserved. Dean (1) has been especially fortunate
in the discovery of some of these structures in the sharks of
the Cleveland shales of Ohio. He has described very fully the
preservation of the kidneys, muscles, skin, the cartilaginous
elements of the fins and arches and portions of undigested food.
So perfectly are the remains preserved that the tissues, in
some cases, admit of histological differentiations into the
component elements. Eastman (2) has described the preser-
vation of the outline of some acanthodians from Mazon Creek.
Woodward (3) has contributed to our knowledge of the soft
-anatomy of fossil fishes in many ways and has added interest-
ing information on the anatomy of the lateral line system of
Cretaceous selachians. Jaekel (4), Dean (1, p. 267) and Gill
(5) have discussed the anatomy and the significance of Juras-
sic and Cretaceous chimzroid egg cases. Otto Reis (6) has
written much on the soft anatomy of various fossil fishes, more

(277)

278 KANSAS UNIVERSITY SCIENCE BULLETIN.

especially the Ccelacanthide, in which he has described the
form of the muscle fibers, the swim bladder and other struc-
tures. Eastman and Parker (21) have described the preserva-
tion of the brain, the internal ear and arterial vessels in
Rhadinichthys deani Eastman from the base of the Waverly
shales, Kentucky. Dean (20) has mentioned the preservation
of the lateral line sensory canals of the head, the auditory
organ and the rim of the nasal capsule in Acanthodes bronni
from the Permian of Lebach, preserved in the Berlin Museum.
Fritsch (7) has described very accurately the outlines of the
body and fin membranes of Pleuracanthus. Traquair (8),
Dean (1) and Sollas (9) have added to our knowledge of the
anatomy of Paleospondylus. Patten (10), Eastman, Traquair
and others have written on various structures of the Ostra-
cophores. Other authors have contributed from time to time
on the subject, until we have, in some instances, e. g., Bothrio-
lepis, Paleospondylus, Cladoselache, a fairly definite idea as
to the outward form and internal structure of the creature.
Among higher animals something has been done on the soft
anatomy of the extinct Amphibia, Ichthyosauria, and Dino-
sauria.

It is with some degree of pleasure that the writer is able to
add to the knowledge of the soft anatomy of extinct forms by
the discussion of the alimentary canal of two Cretaceous fishes.
One is a species of Empo, probably FE. nepaholica Cope, from
the Niobrara Cretaceous of western Kansas, and the other is
a new species of clupeoid fish from the Cretaceous of Texas.

The specimen which probably belongs to Empo nepaholica
Cope consists of the cast of a large stomach which, in all
probability, represents a fish of some ten or twelve feet in
length. It is No. 347 of the University of Kansas Paleontologi-
cal Collection. The remains were discovered in 1897 by Mr.
H. T. Martin in the Niobrara chalk four miles northwest of
Elkader, Kan. The specimen has recently been presented to
the Museum of the University by Mr. Martin.

The specimen consists of a cast of the larger portion of the
alimentary canal of a large species of fish. Attached to the
matrix of the cast on one side is the major portion of the right
pectoral fin, which is described and figured below. So far as
the writer is aware, the present specimen is the most perfect
example of the pectoral limb of an Empo which has been de-

MOODIE: CRETACEOUS FISHES. 279

scribed. Cope (18a) figured an incomplete one and Hay men-
tioned another. Hay (11) has figured a portion of the caudal
fin of an Empo showing the extreme character of segmenta-
tion of the rays. The same character is shown in the pectoral
fin.

The stomach is rounded, somewhat laterally compressed,
and elongate in a slightly U-shaped curve. There are eight
muscular constrictions on the ventral surface and four on the
dorsal. The constrictions, on the ventral surface, occur in
groups. Anteriorly there are two close together. This group
is separated by a space of an inch and a quarter from the next
group, in which there are three, which occur a little over one-
half inch apart. The last group, also of three, is separated from
the second group by one inch. The surface of the stomach
cast is covered with a dark, apparently carbonaceous, ma-
terial which may be carbonized muscle, together with a few
large scales of the typical Empo form. Running the entire
length of the specimen are longitudinal ridges and grooves
showing the cast of the muscular walls of the stomach.

The interior of the stomach, in cross section, shows no food
material, but only chalk. It is possible that the fish, like some
of its modern relatives, may have been a bottom feeder and its
stomach may have been partly filled with Niobrara mud at the
time of its death. There must, however, have been some sedi-
ment enter the stomach after death, for the full form of the
organ is preserved as though the entire stomach cavity had
been packed with mud. Furthermore, the form of the stomach
is that of a carnivorous fish, and recalls very strongly the
stomach of a mountain trout or of the sunfishes of our inland
streams, all carnivorous in habit.

The portion of the alimentary canal preserved is in two
lobes. The first lobe is undoubtedly the stomach proper, and
the constriction between the lobes is the pyloric region. The
other lobe is unlike anything among modern fishes with which
1 am at present acquainted. It is undoubtedly an enlargement
of the intestine and possibly served as a secondary stomach.
It lacks the muscular constriction and the longitudinal plice.
The plice are, however, continued well across the pyloric
region to the beginning of this second enlargement.

The pectoral fin, as preserved, is well characterized in the
photograph (plate LXII, figure 2). It is somewhat turned in-

280 KANSAS UNIVERSITY SCIENCE BULLETIN.

ward and bent, during interment, back against the stomach.
There are eleven rays preserved. The anterior rays are cross
segmented with long divisions, which measure 7 mm. in
length in the second ray. The square notches mentioned by
Hay (11, p. 87) as occurring on the specimen of E'mpo nepaho-
lica Cope in the United States National Museum, are entirely

lacking from that portion of the anterior ray which is pre- —

served in the present specimen. The teeth on the edge are
also absent, nor do I find that they are evident in Cope’s figure
referred to by Hay. The figure is very indistinct, and if the
notches were present they could not, in the nature of the case,
be normal, but would represent places where the segments had
dropped out. The first ray is not a spine. In other respects
the present specimen agrees well with: that figured by Cope
on plate LII, figure 1 (Cretaceous Vertebrata). The fifth and
succeeding rays are segmented like the anterior ones, but the
segments are smaller and measure, on the average, only about
2mm. The seventh ray is especially broad, equaling in its
proportions two and one-half of the other rays. All of the
rays are split distally. The seventh divides into four second-
ary rays and the divisions ascend more and more to the base of
the fin posteriorly. The fin supports are obscured by scales
and matrix so that their nature cannot be determined. On
the opposite side of the specimen, below the pectoral fin, there
are large scales and fragments of ribs.

The second intestinal enlargement is interesting, entomo-
logically, as showing the borings of some fossorial hymenop-
teron; possibly some one of the smaller species of. the
Andrenidz. There are fragments of pupa cases in the burrows,
so there is no doubt as to the recent origin of the holes.

The present specimen is so far the only remains known of
the soft anatomy of the Kansas Cretaceous fishes, and, so far
as I can learn, the first indication of the alimentary canal of
Cretaceous bony fishes of any region. Whether the stomach
and intestines in their various forms will ever be of any help
in determining the relationship of the various osseous fishes
remains to be determined. It is to be feared, however, that
the fishes have been so diversified according to food habits that
these structures will not be of any great phylogenetic value.
The remains are interesting, however, as indicating, in a
measure, the habits of life of at least one of the Cretaceous
fishes.

= ee ee

Tu
7)

aa

A:

MOODIE: CRETACEOUS FISHES. 281

Measurements of the specimen of Empo nepaholica Cope:

Entire length of the alimentary canal as preserved.. 53.2 cm.
Greatest diameter at anterior end................. 10
Least diameter, across pyloric region.............

Greatest diameter of posterior enlargement........ 7.4
Length of pectoral fin as preserved...,...........

Reeeee Wits Of TF (oc u cbse tak shabu ngs comes « 3.7
ammehs Of Diret TRY sick daisies ou Vides ks ses G's 8

Den OF LOC TOY oes a i awihe ye 8p einem 2 mm.
SeURMNCGE OF VATHO CORIO co's cnc ssc seees ascieeerss 15

Thrissopater intestinalis new species.

A species of clupeoid fish is represented in the University
Museum by the remains here described as a new species. The
form is located in Thrissopater of Giinther, described from the
Gault of Folkstone in 1872 (12). My thanks are due Dr. A.
Smith Woodward for the suggestion of a comparison of the
present form with that of Thrissopater. It was thought for a
time that the present form represented a genus distinct from
Thrissopater. The distinguishing character was thought to be
found in the position of the pelvic fins, which has served as a
generic character in other fishes. In Thrissopater salmoneus
the pelvic fin is opposite the dorsal and in the present form it is
distinctly posterior to it. There is, however, a great range of
variation in the position of the pelvic fin, especially among the
lower osseus fishes. My thanks are due Prof. E. C. Starks for

aid in reference to the characters of the modern bony fishes.

During the summer of 1909 the writer spent some weeks study-
ing with him the fishes of Puget Sound. He first called the
writer’s attention to the wide variation of the location of the
pelvic fin in the clupeoid fishes. This variation is easily un-
derstood when it is remembered that the pelvic fin lies free
from any firm attachment and hence its variation in location

would not mean as much as though it were attached to the

scapular arch. Further aid was rendered the writer in de-

_ termining the characters of the clupeoid fishes by Dr. W. G.

Ridewood, of London. An examination of the essays of this

gentleman has been of great service.

The absence of material for direct comparison with the
species of Thrissopater makes it best to locate the present
form temporarily in that genus. The systematic position of
Thrissopater has been the subject of a wide variance of
opinion. Dr. Giinther regarded Thrissopater as closely allied
to the modern Clupeidze and located it (13) in that family, in
which he also included such forms as Spaniodon, Albula, Elops

989 KANSAS UNIVERSITY SCIENCE BULLETIN.

and Engraulis, all of which have been regarded by different
authors as types of distinct families. Boulenger (14) regards
Thrissopater as a member of the subfamily Thrissopatrinz,
which is one of his four subfamilies of the Clupeide. Dr.
Jordan (15) located the form in the family Spaniodontide,
which is closely related to the Elopidze, between which and the
Clupeidze Boulenger regards Thrissopater as being interme-
diate (l.c., p. 564). Professor Starks writes me that Doctor
Jordan now regards the family Spaniodontidez as untenable.
Dr. A. Smith Woodward (16). regards Thrissopater as a mem-
ber of the Elopidz, which differ from the Clupeide in the pos-
session of a single supramaxillary, the degree of union of ‘the
parietals and the gape of the mouth and the presence of a
gular plate. The present form presents the characters of the
Elopide in so far as they are preserved. In recent forms, the
presence of a gular plate in the Elopide serves as a convenient
landmark for the distinction of the families of the Elopide and
Clupeide. Asa matter of fact, the families are so closely allied
that the characters used for their separation must in time be
broken down by the discovery of new material.

Herrings and herring-like fishes are not at all rare in the
Cretaceous deposits of the world. Davis (17) has described
many forms of clupeoid fishes from Mount Lebanon. Before
him Agassiz advanced the knowledge of these forms, and lat-
terly Woodward has described several interesting clupeoids.
Cope described several clupeoids from the Eocene of Green
River, Wyoming (18), and Jordan has cited the interesting
relations of these forms to forms now living in the rivers of
Australia and Chili. At the present time herrings form an
important item in the economic history of the world. Huxley
has dwelt (19) especially on the anatomy and relations of the
herring in this connection. The present form adds yet another
mite to our knowledge of these interesting fishes. It is believed
to be as early as or perhaps somewhat older than many of the
described clupeoids. The specimen comes from the Austin shales
or limestone, which is a probable equivalent of the Niobrara
Cretaceous. It is from near Baylor, Tex., and is No. 300 of
the University of Kansas Museum.

The remains preserved consist’of the atts complete fish,
as may be determined from an examination of the plate. The
caudal portion is, unfortunately, lacking. The outer surface
of the skull was badly broken and Mr. Martin very kindly ex-

MOODIE: CRETACEOUS FISHES. 283

tricated the skull for me from the matrix. It was a difficult
task and the results were hardly worth the efforts, for the
embedded portion was but little better preserved than the
outer. Enough of the skull is preserved, however, to show
many of the important characters. The head is naked; the
body compressed, but whether the ventral edge was drawn out
into a keel or not cannot be determined from the specimen.
The mandible is fully as long as the skull. The relations of the
articulation to the orbit cannot be determined, nor can the
position of these openings be definitely located. The parietal
bones are, apparently, small. Certainly the supraoccipital
projects forward as in Thrissopater magnus. Maxilla is
slender, with a single supramaxillary. The margin of the
jaws is provided with a single row of small, recurved, sharply
pointed teeth of uniform size throughout the length of the en-
tire mandible and maxilla. The quadrate is broadly V-shaped,
with a prominent articular surface. Nasal, ethmoid and pre-
maxillary bones ornamented with numerous small pits. The
same character occurs on the anterior end of the maxilla of
the right side. A single, squarish, punctate, thick, pharyngeal
bone is present. A very few branchiostegal rays are pre-
served; not over ten. From the relationships of the form we
would judge there were many in the complete fish. The oper-
cular apparatus is smooth. Posterior suborbital plate radiately
furrowed; its extent exceeding one-third of the length of the
skull; remaining elements indistinct. Greatest depth of the
body is slightly greater than the length of the skull from
premaxilla to supraoccipital. The length of the body is
possibly equal to four times the length of the skull. The fins
are relatively small. Dorsal fin median in position. The
pectoral fin has sixteen rays, which are cross-segmented but

are not divided longitudinally. The rays are supported by
five baseosts. The distance between the origins of the pectoral

and pelvic fins is equal to nearly four times the length of the

pectoral fin. The pelvic fins have nine rays, none of which

are cross-segmented. The pelvic bone is large and spatulate.
The body scales are small, cycloid, deeply imbricated and

marked with fine concentric lines. There is a large, elongated

and elegantly sculptured scale at the base of the pectoral fin,

as in Thrissopater salmoneus; though in the present instance

the scale is less than one-half the length of the pectoral fin

rays. The vertebre are preserved to the number of thirty-

2—Univ. Sei. Bull., Vol, V, No. 15.

284 KANSAS UNIVERSITY SCIENCE BULLETIN.

four. There may have been as many as forty-five or fifty in
the complete fish. They are fully ossified, slightly constricted
and marked with small longitudinal ridges. The length is
slightly greater than the depth. The neural spines are long
and interlock with the interneurals. Supernumerary ribs
present. Six of them occupy the space of a single vertebral
centrum.

The specimen as preserved is well characterized by the fig-
ures. The fish lacks the posterior end of the body back of the
anus. It is chiefly remarkable on account of the extraordinary
preservation of the casts of the rectum and intestine, of which
there are six coils or loops preserved. The remains are em-
bedded on the right side in a calcareous, arenaceous, shaley
limestone, which also contains remains of some species of
Inoceramus, small fish teeth and the base of a moderately
large shark’s tooth.

Perhaps the most interesting portion of the entire specimen
is the intestinal canal, from the presence of which is derived
the specific name. In general features the alimentary canal as
preserved recalls that of the common fresh-water buffalo fish,
Ictiobus bubalus Raf (plate LXII, figure 1). The similarity
in form is undoubtedly indicative of similarity of habit, and
since we know that the buffalo fishes are bottom feeders we
can easily predicate that our ancient Cretaceous fish had similar
habits and at the time of its death the alimentary canal was
filled with mud mixed with some organic substances; for the
fossil shows a different texture for the cast of the alimentary
canal from the matrix, indicating different materials. The
intestine as preserved consists of six coils or loops of the very
small intestine which immediately precedes the rectum, which
is likewise preserved. The rectum is elongate but no more so
than is the same structure in the buffalo fish. The essential
characters are shown in the illustrations.

The distinction of this species from the other three which
have been assigned to Thrissopater is to be found, first of all,
in the posterior location of the pelvic fin. Its base lies at a
distance posterior to the back edge of the dorsal fin, which is
equal to its own length. So far as I am aware the large axil-
lary scale in other species of Thrissopater is larger and un-
ornamented. From T. salmoneus Giinther the present form is
to be distinguished by the relative proportions of the head and

abs

=

MOODIE: CRETACEOUS FISHES. 285

body. The head and opercular apparatus is contained only
about twice or at most two and one-half times in the body.
From T. magnus it is to be differentiated by the relative pro-
portions of the opercular and the skull. The former is con-
tained twice in the latter in the present form. In T. magnus
it is contained three times. The present species is indeed very
closely allied to Thrissopater magnus Woodward from the
Lower Chalk of Hollingbourn, Kent. It is to be distinguished
by the relative dimensions of the vertebrz as well as by the
proportions existing between skull and opercular. The verte-
bre in 7. magnus are higher than long while in T. intestinalis
they are slightly longer than high, and the ends are occupied
by distinct rims, such as do not occur, apparently, in the
English form. The characters which the two species have in
common are striking. They both have the same notch in the
anterior end of the mandible; the same finely punctate ethmoid
and nasals; the same form and dimensions of mandible and
maxilla; the same divided posterior suborbital; and the same
relative shape of skull. Many of these are generic characters.

The present species can be distinguished from Thrissopater
(?) megalops Woodward by the proportions of the head. In
T. (?) megalops the height of skull from cotylus to supraoc-
cipital is equal to the length of the mandible, while in T. is-
testinalis the mandible exceeds the height of the skull from
cotylus to supraoccipital by 15 millimeters. It may be further
distinguished by the relative proportions of the pectoral arch
and skull as well as by the absence of the radiately furrowed
suborbital and the notch in the anterior end of the mandible
in T. (?) megalops.

Measurements of Thrissopater intestinalis Moodie:

I NTI os ons w cmeid db wade ale = 29 cm.
are ky o'iis 6 «v5.0 00d 0 eh oie hee 9
Length of skull (with opercular apparatus)....... 9.7
“jechet MeN) OG GUAOTALE,. 665. ov es ca ee we vce esis 4.8
I INNIS oon o as vin vic cc ccs vcesseces 6
Depth of mandible at cotylus..................... 1
Diameter of pharyngeal plate.................... 9 mm.
cue cess 2.5
Width of Opercular apparatus.................... 3 cm
ES ES I 2.5
SS 2.5
TS EG 1.2
RII TS ce ee ee cee 2.4
MI on ek ccc ewe cc eee 1.2
CS OI eee 3 mm
Length of dorsal fin as preserved.,...,....,....... 2.1 cm,

236

1

g

3a.

oo

10.

it.

KANSAS UNIVERSITY SCIENCE BULLETIN.

Width of dorsal fin as preserved................. 1.8
Length of caudal hemapophyses.................. 9 mm.
Length of vertebre:. .'.:< is» sis seen eee 6
Depth of vertebre......... ++ smucss sue eee 5
Width of stiall ‘intestine. 2.255300. 2, Fae 3
Length of. rectum. . ..04:64' 5. eh +. ee wee eats aa 12) cm.
Width of Tectum. . 6.5 aioe ve snes oye en 1.3
Length of ‘intetneural 0205... Se eee ee 1
BIBLIOGRAPHY.

. DEAN, BASHFORD, 1909. Studies on Fossil Fishes (Sharks, Chime-

roids and Arthrodires). Memoirs of the American Museum of Nat-
ural History, vol. IX, part V, pp. 211-287, with 65 text-figures and
' plates XX VI-XLI.

. HussaAkor, Louis, has given a very full bibliography of the forms

here discussed. Studies on Arthrodira, Memoirs of the American
Museum, 1906, vol. IX, part III.

EASTMAN, CHARLES R., 1902. The Carboniferous Fish Fauna of

_ Mazon Creek, Journal of Geology, vol. X, p. 536.

Woopwarp, A. S., 1888, Paleontological Contributions to Selachian
Morphology, I: On the Lateral Line of Cretaceous Species of
Scyllide. Proc. Zool. Society of London, p. 126.

1898. Preliminary Note on a New Specimen of Squatina
from the Lithographic Stone of Nusplingen, Wiirtemberg. Geo-
logical Magazine, December, IV, vol. V, No. 409.

COLLINE, W. E., 1895. The Morphology of the Sensory Canal System
in Some Fossil Fishes. Proc. Birmingham Phil. Soe., vol. IX.

JAEKEL, OTTO, 1901. Ueber jurassiche Zihne und Eier von Chimer-
iden. Beilage Band Neues Jahrbuch f. Mineralogie, etc., Bd.. 14,
pp. 540-564, fig. 3, Tafeln XXII-XXITI.

GILL, THEODORE, 1905. An Interesting Cretaceous Chimezroid Egg-
case. Science, N. S., vol. XXII, No. 567, p. 601, November 10.

REIS, OTTO, 1888. Die Celacanthiden. Paleontographica, Bd. XXXV,
pp. 1-96, Tafeln I-V. t

REIS, OTTO, 1889. Ueber ein Art Fossilization der Muskulatur.
Gesellschaft f. Morphologie u. Physiologie in Muenchen, pp. 1-6.

REIS, OTTO, 1894. Ueber Phosphoritisirung von Cutis der Testikel
und des Ruckenmarks bei fossilen Fischen. Archiv f. Mikro.
Anatomie, Bd. 44, pp. 87-119. :

Reis, OTTo, 1898. Neues ueber petrificirte Muskulatur. Archiv f.
Mikros. Anatomie Bd. 52, pp. 262-268.

. Fritscu, ANTON, 1895. Fauna der Gaskohle und der Kalksteine der

Permformation Bohmens, vol. III, p. 1.

. Traquair, R. H. See bibliography in reference No. 9.

SoLLAs, W. J. and I. B. J., 1904. An Account of the Devonian Fish,
Palzospondylus gunni Traquair. Phil. Trans. Royal Society,
London, vol. 196, B, pp. 267-294, plates 16-17.

EASTMAN, CHAS. The results of Patten, Traquair, Dean, Eastman
and others are very adequately discussed in Eastman’s two
memoirs:

Devonic Fishes of the New York Formations, Memoir No. 10 of
the New York State Museum, pp. 24-65;

Devonian Fishes of Iowa. Annual Report Iowa Geological Survey,
vol. XVIII, pp. 31-291; 41 text-figures, plates I-XVI. 1907.

Hay, O. P., 1903. On Certain Genera and Species of North American
Cretaceous Actinopterous Fishes. Bull. Amer. Mus. Nat. History,
vol. XIX, p. 88, plate I, fig. 4.

MOODIE: CRETACEOUS FISHES. : 287

, A., 1872. Figures and a of British Organic Re-
ins. "December. XIII, ik Geol. Survey No. 1, plate I.

rR, A., 1880. The Study of Fishes, p. 656.
RIDGE NATURAL History, vol. 7. Boulenger, Systematic Ac-
unt of the pe oP ap pp. 562-564. 3

1901. pope a of ctype Piches of the British
m, part IV., page 32, plate V, fig. 1, VII, fig. 4.
W., 1887. The Say Fishes of the Chalk of Mount Leb-
| Syr The Scientific Trans. Ro eee a ‘Society, vol.
ser. IN), art. XII, p. 567, plates XXXI-
E. D., 1884. Tertiary Vertebrata, book I, p. i plate IX.

E. D., Cretaceous Vertebrata, plate LII, fig. 1.
[ ai A, 1881. ca ——_ Scientific Memoirs, IV, pp.

R., 1908. ‘The Devonian Fishes of we: wal: XVIII,
Survey, pp. 264-272. Article by Dr. G. H. Parker on
Organs ont Other Soft Parts,” on p. 272.

Nowlin.

PUBLISHED BY THE UNIVERSITY,
LAWRENCE, KAN.

October, 1911.

— ee
Tas
4

ms

5

nba.

may

*
area
;
~~
i Hae
‘
5 as
he
to ?
»

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vou. V, No. 16.] MARCH, 1910. Pirong , Aarne

A NEW SPECIES OF HOLOTRICH.
BY NADINE NOWLIN.
(Contribution from the Zoédlogical Laboratory, No. 193.)

Two text figures.

N April of 1909 there was brought to my room at the
Naples Station a jar of the small barnacle Lepas
pectinata, and in examining microscopically one of the

appendages I observed a minute rotifer-like animal creeping
along its edge. This animal, after being studied for some
time, was found to be an infusorian—a Holotrich, resembling
in a general way Huxley’s Dysteria armata. Upon careful
study, however, it differed in so many ways that I concluded
it to be a distinct form.

Symbiosis is such an old and well known condition in the
organic world that any discussion of it for its own sake is un-
necessary here. As in other cases, these two animals live
together, deriving mutual benefit from the combination; the
ciliate finding a shelter in the appendages and the barnacle, no
doubt, getting a tasty morsel when it succeeds in dislodging
the little animal long enough to sweep it into the gullet. Itisa
question how the protozoén manages to thrive under these con-
ditions. We know that the water currents set up by the host
are swift and frequent, and life under such tempestuous condi-
tions would seem worse even than the chances in the open sea.
The structure of the smaller organism partially explains this.
It has first of all an armature, a tough siliceous skeleton, cover-
ing the body, except a narrow strip on the ventral side. With
this protection it is not easily crushed. Then caudally there
protrudes a hook-like tail, which not only helps the animal in

(291)

292 KANSAS UNIVERSITY SCIENCE BULLETIN.

creeping over the rough surface of the appendage, but also
serves as an efficient means of attachment. Near the oral
opening protrudes a long flagellum which lashes rapidly to and
fro and secures food for the animal.

The small visitor thus established on the appendage of a
barnacle obtains its food from the currents of water which are
constantly coming in to the host. When the dangers come and
the appendages are enclosed by the calcareous plates of the
barnacle, the ciliate still has moisture and some food. One
might think that in casting its lot with a sedentary animal the
free-swimming protozo6n had limited its opportunities. This
is not wholly true. Lepas pectinata grows on masses of lava,
and -during storm these masses become broken and widely
scattered. It is only occasionally that the Naples collectors are
able to find this barnacle—its appearance in the bay depend-
ing upon, first, a storm at sea, and then a strong south wind to
sweep it in. During April and part of May I was able to ob-
tain this material but twice. This was unfortunate as the
nucleus is an interesting one, and promises well for cytological
work. The barnacles live less than a week indoors and the
ciliates perish with them. Moreover, the ciliates are not nu-
merous. As a last resort for material, I took barnacles from
the storeroom, preserved in formalin, and succeeded in find-
ing the specimens, but they did not prove satisfactory for
nuclear study.

DESCRIPTION OF SPECIMEN.

The animal is oval in form, being slightly wider cephalically.

It is covered with siliceous skeleton, giving the appearance of

a bivalve shell, and possesses a caudal appendage. So differ-
ent is it from the usual infusor that it is little wonder most
zoologists at first glance pronounce it a rotifer, or a mollusk, or
a low crustacean. The animal is flattened dorso-ventrally, the
dorsal shell being more extensive and bending over the right
side. The dorsal is slightly convex; the ventral is concave and
attaches to the dorsal by a deep groove on the left. The right
edge of the dorsal shell and the left edge of the ventral do not
meet, thus leaving exposed a very narrow strip of protoplasm.
This uncovered protoplasm is ciliated and it is by means of
these cilia that the animal swims. The surface of both shells
is smooth.

The crevice between the ventral and dorsal plate lies then

cS —— lO a ee

NOWLIN: A NEW HOLOTRICH. 293

the right ventral area of the animal, and because the animal

the posterior end of the animal completely along the side and
rounds across the cephalic end. At its caudal extremity is an
: _ appendage, and at the cephalic, a long flagellum.

EXPLANATION OF TEXT FIGURES.

The figure to the left is a view of the right dorsal side of the ‘animal,
wing the flagellum anteriorly and the ventral cilia; the calcareous
orting rods of the mouth, the two contractile vacuoles and the
s. The caudal appendage is shown by a continuous line and its
of motion by dotted lines.
middle figure is a view of the animal on edge, showing a convex
side, concave ventral, left dorsal groove and caudal appendage.
The reg =r is an imaginary cross-section through the nucleus.

us lies at its distal end, as is claimed for Onychodactylus, |
unable to say. Its range of movement is nearly 180 degrees

ig strides by placing the end against a solid structure.
motion which resembles strongly that of a rotifer. In
: the tail is usually folded into the groove like a knife-
in its handle, and the animal is carried forward
tion of its cilia. The animal when creeping over a
e stands on the ciliated edge, much as a mussel when
rough the sand. In swimming it lies slightly on its
id has a swinging motion through the water.
ellum is located in a deep depression of the cephalic
On superficial examination this depression might

294 KANSAS UNIVERSITY SCIENCE BULLETIN.

be mistaken for the mouth, but the mouth is just ventral to it
and convenient to the currents set up by the flagellum.

The gullet is a long, narrow funnel running obliquely up-
wards and backwards until it nearly reaches the opposite side
of the body. Like other protozoa of this group, it is supported
by calcareous rods. These are plainly visible when picro-
sulphuric acid is first placed on the specimen, but very soon the
acid destroys them and the gullet shows only as a clear place
in the protoplasm.

Two contractile vacuoles are present, and a very large nu-
cleus. The nucleus is especially interesting in these forms be-
cause the meganucleus is heterogeneous, the anterior half re-
maining almost clear when stained with picro carmine, the
posterior half staining densely. The micronucleus may lie
quite far to the posterior end of the meganucleus. The fact
that the two halves of the nucleus react differently to the
same stain suggests a segregation of functions. Since in di-
vision the meganucleus breaks transversely, the resulting ani-
mals are dimorphic as to meganuclear protoplasm.

CLASSIFICATION.

According to Lankester’s classification the specimen de-
scribed above belongs to the sub-order Gymnostomata, that
division of the Holotrichs in which the mouth is closed in the
intervals between the acts of ingesting food. It belongs to the
family Dysterine (Clap. & Lach.), which corresponds to Calk-
ins’s sub-family Erviline: Cilia confined to ventral surface
or a portion of it; caudal end invariably possesses a movable
style arising from caudo-ventral surface.

Of the various genera grouped under this family, or sub-
family, our specimen corresponds most closely to the last of the
following list:

Avgyria (Clap. & Lach.), 1858.

Onychodactylus, Entz, 1884.

Trochilia, Duj., 1841.

Dysteria, Huxley, 1857.

Dysteropsis, Roux, 1902.

The only species of this genus, so far as I can learn, is the
small one found in the lakes of Geneva and called minuta by
Roux, because it measures only 28 micra long and 16 micra
wide. The animal described in this paper is marine; the av-

NOWLIN: A NEW HOLOTRICH. 295

nen is 83 micra aa and 50 wide. Its habitat in

BIBLIOGRAPHY.
i G. 1880. Klassen und a des Thier Reichs, vols. I,

1 and LACHMANN. 1850-60. ‘Etudes sur ae cike et les
_ Mem. Inst. Genevoise, 6.6%.

G. N. 1902. Marine Protozoa from Woods fidie Bull. U. S.
Comm. 21, pp. 415-468.

GN. 1909. Protozoology. Lea & Febriger.

THos. 1857. Mic. Trans. Jour., vol. 5.

‘Eray, Protozoology, part 1, fascicle 2.

Loigeags Infusoires cilies des environs de Geneve. Rev.

- Feces Infusoirienne des environs de Geneve. Vehr.

Vol. V, No. 17—Mareh, 1910.

_ (Whole Series, Vol. XV, No. 17.)

|< » CONTENTS:

FROM THE Santa Rita Mountains, ARIzoNA, CoL-
THE UNIVERSITY OF KANSAS EXPEDITION, James A. G. Rehn.

_ PUBLISHED BY THE UNIVERSITY,
LAWRENCE, KAN.

Ocroser, 1911.

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vou. V, No. 17.] MARCH, 1910. —_[W#oun, Sanems

ORTHOPTERA FROM THE SANTA RITA MOUNTAINS,
_ ARIZONA, COLLECTED BY THE UNIVERSITY
| OF KANSAS EXPEDITION.

BY JAMES A. G. REHN.

Academy of Natural Sciences of Philadelphia.
Plate LXIII.

NHE collection forming the basis of this paper was made in
_ the Santa Rita range during the summer of 1907 by
the University Museum expedition under the late Prof.

FORFICULIDA.
apicidentata Caudell.
Beate Rita mountains, 5000 to 8000 feet, July. F. H. Snow.
One female.
_ This species has previously been recorded from localities in
se ranging from near a hundred feet to about five
thousand feet above sea level.

BLATTIDA.
voptera notha Rehn and Hebard.

300 KANSAS UNIVERSITY SCIENCE BULLETIN.

In Arizona this species appears to be restricted to certain
elevations on mountains of the southern part of the territory,
viz., Santa Rita, Huachuca and Patagonia ranges. °

Blattella dilatata (Saussure).
Santa Rita mountains, 5000 to 8000 feet, July. F. H. Snow.
One female, one male.

These specimens agree with those recorded by the author* —

from the Huachuca mountains.

MANTIDA.
Bactromantis virga Scudder.
Santa Rita mountains, 5000 to 8000 feet, July. F. H. Snow.
One female.
This specimen is slightly larger than a Huachuca mountains
individual of the same sex.

PHASMIDE.
Pseudosermyle truncata Caudell.
Santa Rita mountains, 5000 to 8000 feet, July. F. H. Snow.
One female.

Parabacillus coloradus Scudder.
Santa Rita mountains, 5000 to 8000 feet, June. F. H. Snow.
Two males, one female.
ACRIDIDAS.
Telmatettix aztecus Saussure.
Santa Rita mountains, 5000 to 8000 feet, June. F. H. Snow.
One male.

Mermiria texana Bruner.

Santa Rita mountains, 5000 to 8000 feet, July. F. H. Snow.
One female.

Prorocorypha, n. gen.

Allied to Paropomala Scudder, but differing chiefly in the
great development of the cephalic portion of the head, the
greatly elongate supra-anal plate of both sexes and the sub-
genital plate of the male.

Body sub-bacilliform. Head greatly elongate; vertex and
pre-ocular region strongly produced into a rostrate process,
fastigium lanceolate, tectate, carinate; antenne, at least of
male, somewhat triquetrous, subensiform. Cephalic and
median limbs very small. Caudal limbs very slender. Pro-

* Proc. Acad. Nat. Sei. Phila., 1907, p. 26.

1 rpwa, prow; Kopydy, head.

Be a

ee

Fe Re a aot ae

ee

REHN: SANTA RITA ORTHOPTERA. 301

sternum with a median compressed protuberance; mesosternal
lobes contiguous. Supra-anal plate aculeate in both sexes; sub-
genital plate of male strongly produced, knife-like.

Type: P. snow, n. sp.

Prorocorypha snowi n. sp. (Plate 63.)

Types: ¢g¢ and ¢ (not mature); Santa Rita mountains,
Pima and Santa Cruz counties, Arizona, 5000 to 8000 feet, J une
(¢), July (9), 1907. F. H. Snow. (University of Kansas.)

Size medium; form very elongate. Head one and one-half
times the dorsal length of the pronotum, occiput with a hardly
perceptible longitudinal arcuation ; interocular region but little
narrower than the greatest width of the fastigium; rostrum
very long and deep, occupying nearly half the length of the
head and the greater part of the depth; outline when seen from
the side with fastigium slightly ascending from the interocular
region, face nearly horizontal, inter-antennal portion rounded
to the acute angulate fronto-fastigial angle; fastigium lanceo-
late, the greatest width contained twice ( ¢) or more than
twice (9) in the length, apex very narrowly rounded, lat-
eral margins distinctly carinate, surface tectate, median
carina as distinct as the lateral and extending to the
inter-antennal region; lateral foveole very elongate, lanceo-
late; frontal costa narrow, very slightly but regularly
expanding ventrad in the male, very narrow dorsad, expanding
somewhat in the inter-antennal region and subequal ventrad in
the female, in both sexes moderately suleate and becoming ob-
solete before reaching the clypeal margin; eyes flattened,
elongate ovate, not at all prominent; antennz, in the male at
least, slightly exceeding the head and entire thorax in length,
thick, three-sided, one narrower than the others, ensiform.
Pronotum about twice as long as its width, cephalic margin of
dorsum truncato-rotundate, caudal margin truncate, median
carina distinct; lateral lobes with the ventral margin truncate.
Wings and tegmina undeveloped. Prosternum with a very dis-

tinct longitudinal rounded tubercle; mesosternal foramina con-

siderably impressed. Abdomen with the segments considerably
elongate; supra-anal plate in both sexes strongly produced,
needle-like, its length being but very slightly less than the
dorsal length of the pronotum; cerci somewhat compressed,
simple, sub-styliform; subgenital plate of male produced into
a compressed, slightly arcuate knife-like structure, slightly
more than half again as long as the supra-anal plate; ovi-

302 KANSAS UNIVERSITY SCIENCE BULLETIN.

positor jaws of the female placed well under the projecting
supra-anal plate. Cephalic and median limbs very small;
caudal femora reaching to the fifth abdominal segment, very
slender, tapering; caudal tibize equal to the femora in length,
armed on the external margin with fourteen to sixteen small
spines.

General colors of female burnt carmine and apple green,
the venter of the body and head ochraceous, the carmine
strongest on the dorsum, the green only on the limbs; caudal
femora yellowish proximad. Color of male destroyed by im-
mersion in liquid preservative.

MEASUREMENTS.
Male. Female.

Length of body (including subgenital

and supra-anal plates) .......... .O mm. 44.0 mm
Length of head! 23 vs ia este co sweet 5.5 7.2
Length of fastigium ............... 2.7 3.8
Length of pronotum .............-. 3.2 5.0
Length of caudal femur ........... 10.5 13.6
Length of supra-anal plate ........ 3.4 4.4
Length of subgenital plate ......... 5.0

A single pair have been examined. Doctor McClung, of the
University of Kansas, who has studied the spermatogenesis
of this species and whose assistants collected specimens under
Professor Snow, informs me that ‘“‘specimens such as those you
have were constantly caught in copulo, and a microscopical ex-
amination of the testes shows the spermatozoa well developed.”
The specimens in hand have the tegmina and wing pads in the
reversed nymphal condition.

Eritettia variabilis (Bruner).

Santa Rita mountains, 5000 to 8000 feet, June and July. |

1907. F. H. Snow. One male, four females.

One of the females in this series is in a greenish phase of
coloration, while two of the remaining females have a pattern
of chiefly solid contrasting colors. The green phase has the
greater part of the head, dorsal half of the lateral lobes of the

pronotum, the pale stripe on the tegmina and the dorsal face-

of the caudal femora apple green, while the strongly contrasted
phase has broad seal brown postocular bars extending over the
head and pronotum and regularly expanding in width, the
dorsal aspect much lighter, either ochraceous or rust red, and
the usual dark portions of the tegmina also seal brown. The
supplementary carine of the pronotum are distinctly but not
strongly indicated in two of the females, represented by the

Sea ee e

Se ee

i == ———

REHN: SANTA RITA ORTHOPTERA. 303

merest traces in the female and completely absent in the two
contrasting colored females.

The only previous Arizona record of this species is from
Douglas, Cochise county.

Amphitornus ornatus McNeill. ;
Santa Rita mountains, 5000 to 8000 feet, June and July,
1907. F. H. Snow, One male, one female.

Cordillacris pima Rehn.

Santa Rita mountains, 50('' to 8000 feet, July, 1907. F. H.
Snow. One male.

This species was known only from the eat moun-
tains, Pima county, Arizona.

Psolessa texana Scudder.

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One female.

This specimen is in the blackish phase of coloration—the
typical texana form.

Aulocara femoratum Scudder.

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One male.

This species is now known in Arizona from the Huachuca
and Santa Rita mountains, and from Phenix.

Arphia aberrans Bruner.
Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One male, one female.
_ This species is now known from localities extending from
the Huachuca range to Nogales, Ariz.

Hippiscus corallipes (Haldeman).
Santa Rita mountains, 5000 to 8000 feet, June and July,
1907. F. H. Snow. One male, one female.

Scirtetica ritensis, n. sp. (Plate 63.)

Type: ¢; Santa Rita mountains, Pima and Santa Cruz
counties, Arizona. Elevation, 5000 to 8000 feet. July, 1907.
F. H. Snow. (University of Kansas.)

A very beautiful species of the genus, differing from its
allies in the absence of a median band on the wings, the dark
color being on the periphery, in the color of the disk and in the
heavy, robust build.

Size medium; form robust. Head with the extreme width
very slightly greater than the greatest width of the cephalic

304 KANSAS UNIVERSITY SCIENCE BULLETIN.

margin of the pronotum; occiput slightly arcuate longitu-
dinally, the caudal section of vertex with a pair of distinctly
impressed subpyriform areas; interocular region slightly wider
than the fastigium, broadly and deeply suleate; fastigium very
greatly declivent, sulcate as in the interocular space, the sul-
cate portion, however, being slightly broader than between the
eyes; frontal costa somewhat narrower than the fastigium
dorsad, very gradually expanding ventrad to very near the cly-
peal suture, where the margins curve strongly laterad, sulcate
ventrad to immediately dorsad of the ocellus, where the sulcus
terminates abruptly, and from the ocellus ventrad it is present
shallower and sub-obsolete as the clypeal suture is approached ;
lateral foveole lanceolate, trigonal, distinctly impressed; eyes
broad, somewhat depressed ovoid in outline, about equal in
length to the infra-ocular groove, moderately prominent when
viewed from the dorsum; antennz about equal to the dorsal
length of the head and pronotum; ventral portion of head dis-
tinctly broader than the dorsal. Pronotum with the length of
the disk about half again the dorsal length of the head, and
with the greatest width of the disk but slightly shorter than the
length; cephalic dorsal margin very slightly obtuse-angulate,
caudal margin decidedly obtuse-angulate; lateral angles of the
disk very distinct on the metazona and the cephalic portion of
the prozona, broken and sub-obsolete mesad; median carina
distinct, moderately elevated, straight, cut by the transverse
suleus very slightly cephalad of the middle; lateral lobes dis-
tinctly deeper than long, the cephalic and caudal margins
obliquely emarginato-sinuate, ventro-caudal angle broadly
rounded. Tegmina exceeding the tips of the caudal femora by
slightly more than the length of the pronotum, moderately
broad, slightly narrowing distad, the apex obliquely rotundato-
truncate; intercalary vein prominent, through the greater
part of its length closer to the median than to the ulnar
vein. Wing with the greatest width contained one and
one-thirds times in the length, apex sub-rectangulate with the
distal portion of the costal margin arcuate. Mesosternal lobes
with the interspace very strongly transverse; interspace be-
tween the metasternal lobes transverse. Caudal femora robust,
the ventro-lateral carina distinctly lamellate, reaching its
greatest expansion at two-fifths the distance from the apex,
pattern of the paging distinct; caudal tibiz slightly shorter
than the femora, armed laterad with seven spines.

General colors vandyke brown and pea green, marbled and

ie tal a ss

REHN: SANTA RITA ORTHOPTERA. 805

blotched one over the other, the tegmina, dorsal face of the
caudal femora and disk of the pronotum with the green pre-
dominating; ventral surface isabella color. Dorsum of the
head and pronotum with a regular figure of green which is
margined laterad with clove brown, this figure being con-
stricted at the caudal margin of the head and almost severed
mesad on the pronotum; antennz clove brown annulate with
pea green, their insertion bordered dorsad and ventrad by a
pair of transverse lines of clove brown; eye walnut brown:
Wings with the disk orange rufous, the distal and caudal
margins with a moderately broad band of vandyke brown which
fails to reach the proximal margin, ulnar tenia extending
about half way to the base of the wing; apex with the smaller
areas and a few of the short veins creamy white. Caudal
femora with the dorsal face crossed by a median and a disto-
median bar of clove brown, while the apex is touched with the
same color, ventral sulcus washed with pale blue; caudal tibiz
turquoise blue with the lateral face of the proximal third green-
ish white and the extreme distal portion of the internal face
blackish, leaving a pregenicular pale annulus, the apex dark
blue, and a dark blue annulus placed next to the pale one;
tarsal joints very pale ochraceous.

MEASUREMENTS.
SE OE kes Ga Wele a Eats sss Ses ye hess 23.8 mm.
BER ee A a a aaa a Per 4.5
Between. Of Caudal femur... 2.) deca oss cc te tse rscs 12.0
Me ONO Soe sees vue te a chic ads ou es 21.0

The type specimen is the only one seen by the author.
Tomonotus aztecus (Saussure).

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One male.
Mestobregma rubripenne (Bruner).

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One male, one female.
Conozoa carinata Rehn.

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One female.
Trimerotropis alliciens (Scudder).

Santa Rita mountains, 5000 to 8000 feet, June and July,
1907. F.H. Snow. Three males.

These specimens are somewhat smaller than a female from
the Huachuca mountains. -

2—Univ Sci. Bull., Vol. V, No. 17.

306 KANSAS UNIVERSITY SCIENCE BULLETIN.

Trimerotropis laticincta (Saussure).

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One male, one female.

This species has previously been recorded from Nogales,
Douglas and the Huachuca mountains in southern Arizona,
and Flagstaff in northern Arizona.

Trimerotropis cyaneipennis (Bruner).
Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One female.

Heliastus benjamini Caadell

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One female.

This specimen has pale interrupted decussate markings on
the pronotum.

Previous records of this species are from the Huachuca moun-
tains, the Baboquivari mountains, and Nogales, Ariz.

Heliastus aridus (Bruner).
Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One male.

Dactylotum variegatum (Scudder).
Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. Two females.

TETTIGONIDA.
Arethzxa sellata Rehn.

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One male.

This is the first record of the species outside of the Huachuea
mountains.

Scudderia furcifera (Scudder).

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. One male.

GRYLLIDA.

Gryllus alogus Rehn.

Santa Rita mountains, 5000 to 8000 feet, July, 1907. F. H.
Snow. Two females.

These specimens measure as follows:

Length of Pronotum. Tegmen. Caudal femur. Ovipositor.
4.0 mm. 9.0 mm. 12 mm. 14.5 mm.
3.3 5.5 fo... ae

This species has been recorded from the Huachuca mountains
and Phoenix, in addition to Albuquerque, the type locality.

Vol. V, No. 18— March, 1910.

(Whole Series, Vol. XV. No. 18.)

ints:

PUBLISHED BY THE UNIVERSITY,

LAWRENCE, KAN.

October, 1911.

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vou. V, No. 18.] _ MARCH, 1910. Fae Pee

OBSERVATIONS ON THE GRYLLIDZ: III. NOTES ON
THE CLASSIFICATION AND ON SOME HABITS
OF CERTAIN CRICKETS.

BY W. J. BAUMGARTNER.

(Contribution from the Zodlogical Laboratory, No. 195.)

CLASSIFICATION.

N the course of collecting material for some cytological
studies on the crickets, I was early confronted with the
‘question of what are the true species of Gryllus. The first
specimens collected I (2) called Gryllus assimilis, after care-
fully comparing them with the labeled specimens in the Uni-
versity of Kansas collection, and after reading such descriptions
as were then available in the library. Later I collected abou!
Chicago, Ill., and Woods Hole, Mass. My attempts at classify
ing these specimens led me to the conclusion that the species
of Gryllus, the common larger field crickets, are not fixed but
grade into each other. I found that in all of these places there
were two groups with different breeding seasons—one that
passed the winter in the nymph stage, and another that passed
it in the egg. The former matures and breeds around Law-
rance, Kan., during June and early July, and the other during

_the latter part of August and September.

This question of the true species in Gryllus was frequently

discussed with my fellow student, Dr. F. E. Lutz, now of Cold

Spring Harbor, N. Y., during our study at the University of

‘Chicago. I am glad to confirm his recent publication (14) in

which he says that the species of Gryllus as now named do not

differ in characters, “but merely in the degree of common

characters.” My study has not been especially along the line
(309)

310 KANSAS UNIVERSITY SCIENCE BULLETIN.

of taxonomic characters; but my attempts at classifying the
specimens collected in Douglas and Harvey counties, Kansas,
in Chicago, IIl., in Woods Hole, Mass., in the Santa Rita moun-
tains, Arizona, and in Tarpon, Tex., trying to follow the keys
given by Scudder, Blatchley, De Saussure and others, have led
me to believe that Lutz is right when he says: “Either we
simply name stages in a great continuous mass of variation and
call them species or there is but one species of Gryllus in east-
ern United States, and the names we give are not the names
of species at all, but simply inaccurate, shorthand expressions
for recording the approximate size, proportions and color of
individuals found.” This applies to our common field crickets;
and I do not think that it should be limited to the eastern Uni-
ted States, but should include the central portion as well.

On a collecting trip to Tarpon, Tex., last summer, I found
a color variation which confirms this opinion that all these so-
called species grade into each other. On the low sandy islands
I found that the crickets were straw yellow. Most of them were
in the last or next to the last nymph stage. This was about
June 12. At first I thought they were the imported Gryllus
domesticus; but later collecting disclosed some with a few, and
some with many dark markings. A few of the adults taken
subsequently were quite black. These were under the same
boards or stones with the straw-colored ones, and were mating
with them; and they probably came from the same mother.. A
number of nymphs were brought to the laboratory at Law-

rence and raised to maturity. All of them turned much darker

and some became jet black. As far as I could see these black
ones could not be distinguished from our native species.

One peculiarity of these crickets on the islands at Tarpon
offered an additional reason for thinking they were G. domes-

ticus, or a closely allied species; 7. e., the young nymphs varied

much in their stages of development, a peculiarity I had noticed
in the domestic species in the greenhouses in Chicago. This
fact must be due to the climatic conditions—both forms de-
veloping where there is a long-continued breeding season, with
even temperature.

In the laboratory I found that these Texas forms mated
very readily with our Kansas forms, both the spring-maturing
and autumn-maturing broods. Some of the adults I brought
with me paired with some tardy spring forms and some of the

———— eee

- a ae

Rat SS |
~ ree: ee

*

BAUMGARTNER: OBSERVATIONS ON THE GRYLLIDA. 311

smallest nymphs were not matured before autumn adults
appeared.

In trying to classify these southern forms I finally concluded
that they were our common black field crickets, which had lost
a little, very much, or nearly all of their pigment. Although
Blatchley, Scudder, and De Saussure use the color difference as
one of the prominent characters separating species, I do not
believe that it can properly be so used. “Black color” and
“straw color” do not stand for different species in the crickets
of the Texas coast. The black gradually shades off into the
straw color; and a black one and a light one may have the same
mother.

An examination of the germ cells reveals no differences in
cell structure between the southern light-colored specimens
and our native black ones. But both differ markedly from
that found in Gryllus domesticus, as has been and will be shown
by the papers dealing with spermatogenesis.

All of the collections made in the various localities show
dimorphism as to wing length. The short-winged forms are
very much more numerous in all places, but the long-winged
forms vary greatly in frequency in the several localities, as
Lutz (14) has found.

Blatchley (6) is correct when he suggests that the failure
of past monographers of this genus is in part due to the fact
that they have neglected the study of the animals in the field.
By using this method he has added some very useful hints on
the habits and structure, as they bear on the classification. He
has plainly shown that there are in many localities really two
broods, one maturing early and the other later. He considers
them as belonging to different species. Lutz denies this. In
whatever region I have observed the two broods, the autumn
specimens are larger and more robust than the earlier ones.
They also differ in proportions and color enough to represent
two species according to ordinary criteria for species; but the
intergrading of the forms from different localities would re-
move all distinctive characteristics. So while Blatchley is ap-
parently right, I feel confident that extensive careful collecting
will show that Lutz is correct.

The earlier brood lives “in burrows singly or in pairs,”
while the later ones “are more sociable,” and there is not much
“forsaking of burrows,” as Blatchley (6) thinks. My ob-
servations have led me to the following conclusions: Of the

312 KANSAS UNIVERSITY SCIENCE BULLETIN.

spring brood each individual has a separate well-made burrow,
early in the season, usually under some stone or board. The
male keeps his as long as he lives, or well through the breeding
period at least; while the female abandons hers when she be-
comes an adult, or even before. Thereafter she may be found
with the male in his burrow or in any convenient hiding place.

The young individuals of the autumn brood never make much
of a burrow, but live under bunches of loose, dry grass or old
rags, or whatever they find. I have frequently found more
than a dozen in an old newspaper in the grass. The adult
males sometimes have a sort of burrow, particularly late in the
season, but most of the time I find them in any kind of a hiding
place. I am quite sure, however, that these creatures, young
and old, especially the males, have a selected spot in the grass
or paper, which serves as their home, and so the difference
between the spring and autumn forms is really this: the former
dig a burrow for a home, while the latter simply select some
convenient place to stay.

Judging from my study of the germ cells of Gryllus domes-
ticus, given in a former paper (3), this species is quite different
from the other forms. The difference of chromosome number
and shape are such that I should expect the domestic species to
be very different in taxonomic characters; but such is not the
case.

In other genera of the family Gryllide the species are more
distinct and limited. In Gcanthus, following Hart (10), we
classify the species largely by the color markings on the basal
joints of the antennz, and this seems to be quite constant. I
have found that I can separate the nymphs quite readily by
means of these markings. Nemobius shows more variation,
and probably after large, widespread collecting the species may
prove to intergrade. In Gryllotalpa the species are quite
distinct.

FOOD HABITS.

Very many observers have written of the food habits. It
is known that the common black field crickets may eat almost
anything. In captivity they will sometimes devour each other,
the stronger ones feasting on the weaker ones even before they
are dead. I have seen a female chew the wing of a male, and
I have found a crippled female with her abdomen partly eaten
away. In their free life I think this rarely or never occurs.

_—

7 ote” >

BAUMGARTNER: OBSERVATIONS ON THE GRYLLIDZ. 313

The females will eat the empty spermatophores whenever they
find them.

Among the mole crickets I observed this peculiarity between
two species. In a box of specimens of Scapteriscus sent me
from Porto Rico I never discovered any partially devoured
ones among the few dead specimens; but while collecting our
own species Gryllotalpa borealis, in northern Indiana, I placed
one female adult and six nymphs in a bottle full of sand in
the field. When I returned to the laboratory I found but two
nymphs, the others having been devoured by the adult. This,
with some later experiences, led me to believe that adults will
eat the nymphs whenever they find them in their burrowings.
However, this cannot be true for the very young nymphs, as the
eggs are laid in a mass in a much frequented part of the
burrow, and the mother, no doubt, cares for the eggs and
young for a while.

; EGG-LAYING.

Blatchley (6) says: “The eggs of most crickets are laid
singly in the ground.” My observations confirm this as far
as Gryllus and Nemobius are concerned. The large black field
cricket selects usually a somewhat barren spot in a grassy
field, where she lays her eggs. She will force her ovipositor
_ into the ground and deposit a single egg, then removing the
ovipositor partly will put it down at a different angle and
plant another egg, and repeating the process will leave a third.
On no occasion did I see more than four eggs laid without the
ovipositor being completely removed and pushed into the
ground at a new place. Nemobius lays its eggs in a similar
manner. Two or three, rarely four, eggs are laid almost side
_ by side, and then the next batch are placed a quarter of an
inch or more away.

__ In but one instance did I find eggs laid by the mole cricket.
_ They were “in a heap on the floor in the enlarged part of a side
gallery,” just as Barrett (1) has described.

CHIRPING.

A peculiar habit of the mole crickets, of which I made brief
mention in an abstract (4), is the chirping of the female. A
hurried examination of the tegmina of the females will show
that the nerves are modified into a rasping and sounding organ,
which is not as large or as well developed as that of the male,
but well enough to have made thoughtful observers of the past

814 KANSAS UNIVERSITY SCIENCE BULLETIN.

suspect that it might function, and that female Gryllotalpa
might chirp. As far as I have been able to read the literature
no one has observed that they actually do so. Most of our
books say that only the males stridulate. LaCordiare (12) says,
“The chirping organs of the crickets are simple and limited to
the males.” Scudder, speaking of the crickets, says, “his
egotistic love song.” Comstock (7) writes, “the males of the
crickets have musical organs.” Lang (13) says: ‘In the Lo-
custide and Gryllide only the males stridulate, by rubbing the
rough basal portions of their wing cases against each other.”
Packard (15), after speaking of the organs in the males, says:
“The females are not invariably dumb, both sexes of the Eu-

ropean Hphippigera being able to faintly stridulate.” Henne- —

guy (11), in speaking of the musical organs, writes: ‘Where
they are found they are well developed in the males only; in

the females they are more or less rudimentary. Such is the

case in the Gryllide.” Barrett (1) describes the stridulating
organ in the male “Changa” or Porto Rican mole cricket,
Scapteriscus didactylus; but he has completely overlooked the
same but less well developed organ in the female.

The female mole cricket has quite a loud and distinct chirp.
It usually consists of a single note; but there may be several
at short intervals. This note is less shrill than the ordinary
call of the male. However, the male has a note very similar to
that of the female which it uses for the same purposes, namely,
as a means of recognition in the dark burrows. The call is al-
ways given when one individual is approaching another, es-
pecially when digging a new tunnel. Both genera, Gryllotalpa
and Scapteriscus, have the stridulating organ on the female
elytra, and so both must be able to chirp. I never isolated a
Porto Rican female to hear its chirp, but after hearing the call
of our native cricket I feel sure that I have heard the insular
female’s chirp also.

This unusual ability possessed by the female is an adaptation
to life in underground burrows. It enables the individuals to
recognize others which are approaching under conditions where
sight cannot be used. Thus enemies and friends can be dis-
tinguished; while if the female were dumb, as she is in all
other crickets as far as I know, they might often attack even
their mates.

|

i” Fee

a |. yale
EE . ; '
: FAN | STINE Sea”

BAUMGARTNER: OBSERVATIONS ON THE GRYLLIDAZ. 315

PROTECTIVE GLANDS IN THE MOLE CRICKETS.

As indicated in an abstract (4), I have found that the correct
interpretation of the function of the anal gland, which has so
long puzzled investigators, is protective. Leon Dufour (8), a
careful French investigator, first described in the mole crickets
in both sexes a pair of azure or skim-milk colored glands con-
nected with the rectum. Their secretion he compares in con-
sistency with the vitreous humor of the human eye. To this
secretion is added some excrement from the rectum, and when
this mixture is expelled it forms a brown liquid of nauseating
fetidity. He calls the gland “an organ of excremental se-
cretion.”

Berlese (5) describes the same structure and thinks it is a
prostatic gland analogous to that found in the locustids. Al-
though he found it in the female also, he does not seem to try
to explain it there.

Fenard quotes both of the above descriptions and adds a good
many observations of his own. He describes the gland from
sections and gives the action of certain fixatives and stains
upon the tissues of the gland and its contents. He states in
detail the macroscopic and microscopic structure. He concludes
as follows: “Judging from the position of this organ, from the
consistency of the liquid which it contains, and from its points
of similarity with the prostatic glands of the locustids, I think
that it ought to be considered also as a gland furnishing a
mucus destined to lubricate the copulating apparatus. This
organ exists in the female, it is true, but in this case it fur-
nishes without doubt still a lubricant for the vagina, or a
liquid to form the nest of these insects.” After describing the
details of this gland in the female he says: “I think that these
organs can only be some secreting agent of a mucus destined
to lubricate the genital organs; or perhaps they glue together
and hold the spermatophores; or perhaps again they secrete
the substance used to form the nests in which are found, as
we all know, two to three hundred eggs all massed together
and more or less united.” It is evident from this uncertainty
that Fenard did not know the function of the glands in ques-
tion, yet he was inclined to follow Berlese and called them
“prostatic glands.”

Packard, in his work on Entomology, places the anal odor-
iferous glands described by Dufour among the repugnatorial

2—Univ. Sci, Bull , Vol. V, No. 18.

316 KANSAS UNIVERSITY SCIENCE BULLETIN.

glands, apparently because he has concluded that all fetid and
anal glands are repulsive.

As far as the position, size and structure of these glands are
concerned, these earlier observers are on the whole correct.
They agree, too, in the main points. Dufour has the glands
attached to the rectum, while Fenard has them attached to the
genital duct. The explanation of this difference of observation
is partially suggested by Fenard, when he says: ‘En somme
ces organes paraissent d’eboucher dans une sorte de cloaque ou
arrive l’oviducte.” The mole crickets have but a single opening
at the posterior end of the abdomen; and a short common duct
carries the genital and excrementary products. This should
very properly be called a “‘cloaca.’”’ Into this cavity the short
ducts of the anal glands empty.

Fenard gives as the sizes of the glands.“‘about six millimeters
in length and three millimeters in thickness.”’ I found none
as large as that, but the size would depend in part upon the
amount of secretion in the gland. Both Dufour and Fenard
speak of two lobes and a median constriction. There is some
tendency for such a constriction to show, but it is not constant.
The shape and position of the organ would depend somewhat
on the amount of extension of the abdomen and the full-
ness of the rectum. The two lobes when present do not differ
in histological structure, and not in function, as Fenard has
shown by his. careful work by means of sections. The
walls are resistant, the cavity large, and the contents appear
homogeneous, granular, and they coagulate as a result of fix-
ation, and color strongly whenever stained. All these facts
Fenard has correctly described.

But Fenard must have worked with preserved specimens
only, or he would not have made the error concerning the func-
tion of the gland. Although he quotes Dufour, he cannot have
followed his suggestion when the latter says: “If one seizes
a mole cricket of either sex, it squirts from the anus a brown
liquid of nauseating fetidity. This liquid is formed in part by
excrement from the rectum and is in part the product of a
special secretion.”

I have studied Scapteriscus didactylus from Porto Rico and
Gryllotalpa borealis taken in northern Indiana and in eastern
Kansas. My observations and experiments show that the above
quotation is correct in most parts. If the insect is held or irri-
tated in the region of the head or thorax, there is no discharge.

ey eee

BAUMGARTNER: OBSERVATIONS ON THE GRYLLIDAS. 317

But if held or pinched or pricked or chemically irritated on
any side of the posterior part of the abdomen, or on the hind
legs, there is always ejected from the anus a bluish-white
liquid with some excrement. The discharge is directed as
nearly as possible to the point of attack, be it above, below,
behind, or on either side of the abdomen. It is driven with
considerable force, enough in some instances to carry it across
an aquarium eight inches in diameter. After several ejections
there is less excrement in the liquid, which becomes almost
colorless, losing its milkiness.

The ejected mass has a very fetid odor and is very sticky,
so sticky that a half-grown nymph can readily be suspended
by lightly touching a needle to some of the secretion and then
to its abdomen. An adult female, in spite of her strong legs,
was held for nearly a minute as a result of touching her be-
smeared body against the side of the jar.

In some breeding experiments reported elsewhere I was able
to study the effect of this ejection and the conditions under
which it is made. There was no discharge when the male was
carefully introduced into the jar with the female, but on one
occasion it happened that the male became excited and rushed
upon the female in his attempt to get away. He received a
discharge upon his head and into his face. He stood for a long
time trying to clean this off. He apparently could remove but
little of it, and died on the second day thereafter. At another
time a female received a lesser discharge from a male. She,
too, spent hours trying to scrape off the sticky stuff, but failed,
and died on the third day. The other pair lived for many weeks
longer. Perfectly calm individuals, when put into a jar in
which there had been a discharge a day or so before, became
very much agitated and tried hard to escape from the en-
closure. This behavior suggests that when these insects get
this odor it warns them that an enemy has been or is near, and
they try to escape. I repeated this test several times with the
same result. I introduced some affected sand into a jar con-
taining a calm individual. He became agitated. In every in-
stance the crickets became excited when they perceived the
odor.

The fetidity of the liquid must repel very ardent pursuers,
and the stickiness must retard them should they become en-
tangled in a discharge. It is, no doubt, for the purpose of
so entangling the enemy that the cricket directs its discharge
toward the point of attack.

318 KANSAS UNIVERSITY SCIENCE BULLETIN.

This defensive organ probably explains the fact that mole
crickets have so few natural enemies, as reported by Bar-
rett (1).

Since the Gryllotalpidze move most of the time in under-
ground burrows the discharge from the anus would protect
against attacks from the rear. Hence there is no discharge
when the irritation is on the anterior half of the body. The
head and thorax, besides being very hard, are further pro-
tected by the powerful fore legs. The abdomen is compara-
tively soft and without other protection than that described
above.

My observations and experiments prove conclusively that the
secretion of the anal glands, or “prostatic glands” of Berlese
and Fenard, is preéminently protective, as any one who will
take the trouble to secure a live specimen and repeat these tests
can see for himself. Neither Berlese nor Fenard can have
handled live individuals, or they should have seen the use of the
anal secretion.

As far as we know no other orthopteran has these protective
glands, nor has it the same peculiar habits. The mole crickets
running along the narrow underground tunnels have the soft
abdomens constantly exposed to the attacks of enemies which
they cannot see or perceive, so they have developed a special
organ which can instantly repel or retard a pursuer.

SUMMARY.

1. The species of Gryllus in eastern and central United
States are not distinct, but form one large intergrading series,
as Lutz has shown. This is true also for the supposed dis-
tinguishing straw and dark colors as shown by the specimens
collected in Texas. ;

2. The female mole cricket has a partially developed chirp-
ing organ on its elytra. With this instrument it produces a
single note used as means of recognition in the dark tunnels.

3. The anal gland of Dufour, the prostatic gland of Berlese
and Fenard, is protective in function. The secretion operates
as a repellant by its fetidity, and as a retardant by its
stickiness.

4. Both the female musical organ and the protective gland
are adaptations to life in underground tunnels.

UNIVERSITY OF KANSAS,
May 10, 1909.

‘BAUMGARTNER: OBSERVATIONS ON THE GRYLLIDA. 319

BIBLIOGRAPHY.

BARRETT, O. W., 1902. The Changa or Mole Cricket in Porto Rico.
Bulletin No. 2, Porto Rico Agricultural Experiment Station.
2. BAUMGARTNER, W. J., 1902. The Spermatid Transformation in
: Gryllus assimilis with ea ial Reference to the Nebenkern. Kan.
= Univ. Sci. Bul., vol. 1, |
3. BAUMGARTNER, W. J., i504. ‘ore New Evidences for me Indvidual-
ity of the Chromosomes. Biol. Bul., vol. VIII, No.
-BauMGaRTNER, W. J., 1905. Some caged Habits. of the Mole
_ Crickets. Science, N. S., vol. XXI, No. 544
BERLESE, A., 1882. Ricerche sugli_ organi genital degli Ortotteri.
Atti della R. Accad. dei Lincei. Ser. III, vol. XI. .
Buatcuiey, W. S., 1902. The Orthoptera of Indiana. Twenty-
~ seventh Annual Report « of the Department of Geology and Natural
Resources of Indiana.

omstock, J. H. and A. S., 1895. A Manual for the Study of the
Insects. Ithaca, N. Y.
[ LEON, 1841. Recherches anatomiques et physiologiques
‘sur Orthoptéres, les may onteres et les Néuroptéres. Memoire
des Savants étrangers,
ARD, A., 1897. Recherches sur les organes complémentaires in-
res de eopeeriel génital des Orthoptéres. Bull. Sc. France-
Belg., vol.

ary oo A., 1892. On the Species of Gicanthus. Ent. News,
vo: ° .
NEGUY, L. FELIX, 1904. Les Insectes, Morphologie, Reproduction,
Embryogénie. Paris.

ACORDIARE, M. TH., 1834. Introduction a L’ entomologié. Paris.
13. Lane, A. Textbook of t Comparative Anatomy. (Translated by H. M.
and M. Bernard.) London.
4. Lutz, F. E., 1908. The Variation and Correlations of Certain Tax-
-onomic Characters of Gryllus. Carnegie Institution of Washing-
ton, Publication No. 101.
15. PACKARD, A. S., 1893. Textbook of Entomology. London.

}. SAUSSURE, HENRI DE, 1872-78. Melanges Orthopterologiques. Geneva.
ie Scupper, S. H. Notes on the Stridulation of Some New England

Ss. go The species of Gryllus on the Pacific Coast.

DER, S. H., 1901. The Species of Gryllus Found in the United
aa East of sed Sierra Nevadas. Psyche, vol. IX.

8—Univ. St, Bull, Vol. V, No.8

pes ° CONTENTS:
THE Gryuupa: By, Corutation, W. J. Baumgartner.

ae

_ PUBLISHED BY THE UNIVERSITY, oe
7 LAWRENCE, KAN. |
; October, 1911.

34%.

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Yor. V, No. 19.] MARCH, 1910. ge Af ig

"OBSERVATIONS ON THE GRYLLIDZ: IV.
COPULATION.

_ BY W. J. BAUMGARTNER.
(Contribution from the Zoélogical Laboratory, No. 196.)
Plate LXIV.

HISTORICAL NOTE.

2S first scmpaabicns of the correct method of the transfer
_ of the spermatozoa from the male to the female in the

annexed to the ejaculatory duct and noted that their
nm coagulated very readily when exposed to the air, and

“Siebold believed with the older naturalists that the
Scanere i is the end of the penis broken off and often re-

é man, es Charis dines (11). But the work was little
known and sae questioned by Milne-Edwards (15); so that
tes. (828)

324 KANSAS UNIVERSITY SCIENCE BULLETIN.

textbooks of to-day still disregard it, or quote it with a ques-
tion mark. Since I shall use Lespes’s work very freely in my
own description, I shall not review it farther here. Suffice it
to say that the observations are strikingly keen and correct
for having been made so early, and they should long since
have been inserted in our texts.

Leydig (13) in his fine work, dealing more with the histo-
logical side of the subject, says that in many invertebrates the
accessory sex glands form a spermatophore.

Milne-Edwards (15), in 1870, said, speaking of the sper-
matophore mentioned by Stein and Siebold: “It is possible
that it may exist in certain cases, but in other cases it is evi-
dent to me that the appendage in question was certainly a
part of the penis.” Of Lespes’s statements he writes: “It
does not seem to me sufficiently demonstrated that the so-
called spermatophore is not the terminal part of the penis
which in copulation is broken off and remains inserted in the
female apparatus, as it does so often in the case of other
insects.”

Milne-Edwards’s prominent position, no doubt, gave his
work the preponderance, and so we find Lespes’s work hardly
ever quoted. To show how it was disregarded, let me quote
from Girard, who wrote three large volumes on entomology.
In 1873 (7) he said of the glands joined to the vas deferens,
“They pour in a liquid to delay or modify the sperm.” In the
third volume, in 1885 (8), he repeated the above statement
about delaying the sperm, and then adds, ‘‘or perhaps to give
more activity to the sperm, which are almost immotile in the
testis.”

Berlese (4) studied the external and internal genital organs
of various orthopteran species. In the body of his work he
does not seem to realize that the Gryllide differ from the other
Orthoptera, and so he describes in Gryllus a penis and all the
other accessory parts, without any thought of their correct
function. In an appended “Physiological Note” he comes
nearer the truth. He says, concerning copulation, that fertili-
zation is internal and that copulation in the Acridide and per-
haps in the Mantide is by means of a penis, the act lasting a
' relatively long time; while in the Locustide and Gryllide,
“the male mounts on the back of the female, the penis being
inserted in the vagina and the sperm injected in a short time.”

a a ee Sad eis

BAUMGARTNER: COPULATION IN GRYLLID. 825

He supposes that the mucus secretion of the annexed glands
serves to lubricate the organs. He notes the fact that in
Gryllus and Gryllotalpa the mucous secretion flows out through
the same duct with the sperm. He adds further: “In Gryllus,
I think, fecundation is more complex, occurring by means of a
special organ which is left in the vagina of the female and
which can be reformed up to three times a day, as I myself
have seen. This singular organ holds the sperm in its interior
and is a true spermatophore.” The only way I can explain
how Berlese came to overlook the discrepancy between his
description and his “Physiological Note’ is to think that in
the body of the paper he was concerned with the homology
of the parts in the various groups of Orthoptera, and so did not
think of the function.

Graber (9) shows that in many Orthoptera there is no true
copulation, but that the sperm are transferred by spermato-
phores. Palmen (17) studied by means of dissection and sec-
tions several species of Orthoptera, and he confirms Graber,
showing the absence of a penis and the correct method of
spermatozoon transfer.

Peytoureau (19) in his long thesis on the “Morphology of
the Genital Armature of Insects,’ does not even mention
Lespes’s work, again showing how little this early, accurate
investigation is known; or if known, how little it is appreciated.
Peytoureau worked on the mole crickets, some long-horned
and some short-horned grasshoppers. His special study was
the hard parts, and he showed that the penis is absent, ex-
cept in the Acridide, the Forficulide and the Ephemeride. But
his study was centered about the homology of the male and fe-
male hard parts. Questions of the number of abdominal seg-

ments, the relative position of the ovipositor and the penis and
___ to what segments these belonged, were the ones which he con-
sidered; and so the only observation of importance bearing
upon our topic is the absence of the penis in the Gryllide.

_ Fenard (6) in 1896 published a long paper on “Les Organes

Complimentaries Internes de L’appariel Genital des Orthop-
teres.” He studied both sections and dissections of the an-
nexed glands in the two sexes of several genera of most of the
families of Orthoptera. In his observations he is correct on
the whole. But in his interpretations he is not always so
fortunate. The errors made are due largely to the fact, I

326 KANSAS UNIVERSITY SCIENCE BULLETIN.

think, that he worked with preserved specimens only, and so
missed the function of parts as proven by the habits of the ani-
mals. The suggestion that the anal gland in the mole crickets
secretes a lubricant for the genital organs is entirely wrong, as
I have shown elsewhere (2). Concerning the glands annexed to
the ejaculatory duct he reasons correctly, as follows: Gryllids
and locustids form a spermatophore, as shown by Siebold
(21), Lespes (11) and Berlese (4). But these families are
the ones that have this mass of annexed glands; hence these
glands secrete the substance for the spermatophore. He
thinks there must be a spermatophore formed in Gryllotalpa,
although it has never been seen because of the habits of these
insects. In this surmising he was fortunate, as I show below.
But he is mistaken in the function of what he calls the ear-
shaped gland. He thinks that it is analogous to the anal
gland of Gryllotalpa. This is surely wrong. The position of
the pair of glands and its relation to the excretory duct is such
that it can have no such function as I have found for the anal
glands of the mole crickets. My many dissections and obser-
vations lead me to believe that there is no gland which secretes
a lubricant for the sexual organs.

Although the method of copulation by spermatophore and
the function of the annexed glands have been so well described,
it is rather disappointing to see our textbooks still quoting
Carus’s figures showing that the male generative apparatus of
Gryllus is among the simplest found in the insects. See
Henneguy (10) et al. Although this Frenchman cites Fen-
ard’s paper in his bibliography, he makes no reference to his
findings, but says of the annexed glands: “Elles ont été peu
étudiés jusqu’ icé.” But under the topic of “Copulation” he
gives the process in Gryllus in detail, quoting Lespes’s-descrip-
tion and some of his figures of the spermatophore. In this
case he does better than our own Packard (18), who, in speak-
ing of the spermatophore, says: “In Locusta, and perhaps
also in Gryllus, the sperm is enveloped by the secretion of the
accessory glands of the seminal duct.” This “perhaps” is
very striking when one considers how easy it is for any one to
observe copulation in our common field cricket. One need
only catch an individual of each sex during the height of the
breeding season, as indicated by their great chorus of chirping,
and place them with a little grass over night or longer into

a ee ., SP

BAUMGARTNER: COPULATION IN GRYLLIDZ. 327

separate jars. Then slowly turn the glass with the male over
that with the female. In a comparatively short time the male
will begin his chirping and one can see the whole series of
acts of courting and copulation. After observing the process
one can hardly prevent a smile when he reads Loeb’s (14)
quotation of Yersin’s experiment of having a pair of decapi-
tated crickets copulate, and then see the admonition: “Of
course it was necessary to place the male on the female.’”’ One
would hope that the observations in the experiment itself, and
any conclusions drawn therefrom, may be more to the point
than is the self-evident suggestion added at the end.

DESCRIPTION.

A free translation of Lespes’s paper of 1855 would describe
fairly correctly and with a good deal of detail the various steps
in the process of copulation. He notes at the beginning that
there is no true copulation, but that the male simply deposits
a spermatophore into the posterior end of the abdomen of the
female; and that this process may be repeated many times.

The Courting.

Lespes (11) studied first the field cricket. Catching speci-
mens of both sexes, he got them accustomed to their new sur-
roundings and then brought the two together. “The male
soon began to chirp, and to move around the female. Ap-
proaching nearer and continuing his chirping he turned his
abdomen, carried very low, toward the head of the female.
She remained quiet for ten minutes, not appearing to notice
the maneuvers of her mate; then she moved forward a little
and began to caress his abdomen with her mouth parts. Pres-
ently she mounted partly on his back; and he, stopping the
chirping, glided back under her. The female vulva now rested
above the end of the male abdomen.”’

This description of the courting is fairly exact. The caress-
ing by the female is not essential or even usual. Sometimes the
female hardly stops feeding during the whole process, simply
allowing the male to slide under her. When the male ap-
proaches the female, or when she approaches him, his chirp
becomes much softer and less shrill. On some occasions if the
female did not yield after a certain amount of chirping the
male became angry and fought his mate.

328 KANSAS UNIVERSITY SCIENCE BULLETIN.

The Transfer of the Spermatophore.

Lespes says: “The plate, the tenth tergite, which covers the
opening of the anus and the pieces of the genital armature,
was raised, and a hook of this apparatus penetrated into the
vulva. Immediately after this the end of a small brown body
directed by the hook was introduced into this opening. The
female then left the male, carrying the small brown body of
which I have spoken.’”’ Lespes then removed the small appa-
ratus from the female quite easily, as only its front end was in-
serted in the vulva. This apparatus he found composed of a
small hard ampulla and a thin, transparent plate curiously
twisted. Upon examining a jar in which he kept a male over
night he discovered that there was a spermatophore lying on
the bottom. He later proved that this was not exceptional, but
that males deprived of females rid themselves of their sper-
matophores.

The male which had copulated with the female seemed ex-
hausted and did not renew his courting for several hours, al-
though he was again put in with the female. Lespes observed
the transfer of the spermatophore very often, and each time
he got a spermatophore from the female, which he studied
later.

Taking up this transfer in detail, the Frenchman describes
the position of the spermatophore in the male. “It lies in the
posterior end of the abdomen covered by the supra-anal plate.
To see it in situ, gently raise the anal plate which covers the
anus; below this there is a horny plate with three hooks pro-
jecting backwards and upwards, and below this there is the
spermatophore lying in the concavity of the eighth sternite.
The ampulla extends backwards and is supported on either
side by fleshy pallets (plate is used by Lespes). ‘The plate and
the thread of the apparatus are still in the organs which formed
them. At the moment of copulation the supra-anal plate is
raised and the hooks are introduced into the vulva, holding the
two individuals firmly together. The end of the spermato-
phore plate from which projects the thread glides in a kind of
* hollow groove in behind the hook in a manner analogous to
that which surgeons use when they introduce a bistoury upon
a groove-director. By a rapid movement from before back-
wards the plate is fastened and the two pallets which hold the
ampulla move aside, and so the spermatophore is carried by the

BAUMGARTNER: COPULATION IN GRYLLIDA. 329

female, held by the plate only. Ten minutes after copulation
some males had formed a new spermatophore, but it was soft
and white, and attained its hardness and color only after an
hour. The spermatophore can be quite readily removed from
the male by means of a forceps. To form a new spermato-
phore the male goes through some motions similar to those
used in defecation. The female carries the spermatophore for
some time, in one case six hours, and it was not dropped until
she mounted a male again.”

My own observations show that this account is quite cor-
rect in most of the details. Our common black field cricket
differs in the following points: The female if undisturbed
will not leave the male immediately after the placing of the
spermatophore. She may remain perfectly quiet for several
minutes.

A third element of this sperm-bearer, which Lespes seems to
consider of little importance, is very essential, as I shall prove
below. It is the thread which projects beyond the plate. (See
fig. I.) This thread slides in the groove when the apparatus
is being introduced. The mechanical process can be better
understood by examining figure 5. The ampulla lies between the
two pallets, which are then extended backward very far. The
plate lies in the anterior part of the mold and the thread lies
in the groove below the hooks. In transferring the spermato-
phore the male turns the hooks upward by swinging them, or
rather the whole mold, at its attachment to the tergite as a
pivot. By this turning the hooks first extend upward and then
forward and are thus inserted into the vulva of the female,
holding her firmly. Then by means of the muscular mass be-
hind, the thread and plate of the spermatophore are pushed far
into the vulva. The hooks of the plate hold the apparatus and
the thread carries the sperm into the spermatheca. The
thread and the anterior end of the plate are quite flexible, and
so they are bent and directed by the “independent piece with
the short stylet” which guides the thread into the duct leading
to the spermatheca. The “forward and backwards” movements
are concerned with the proper placing of the thread.

The time suggested for a male to copulate a second time is
much too long. I have had a second pairing occur in fifteen
minutes. The spermatophore in this case was whiter and
softer than usual, but as far as I could see it was a normal

330 KANSAS UNIVERSITY SCIENCE BULLETIN.

copulation. This intervening time was unusually short,
but I am sure that it does not take an hour for a spermato-
phore to harden after it is formed.

I am sure that the males of owr field crickets rid themselves
of the spermatophore, if deprived of the females, only ea-
ceptionally. In all my many observations I have seen the thing
occur but three times, twice in Gryllus and once in Nemobius.
In every case it was rubbed off after an unsuccessful attempt
to introduce it into the female. In these trials the plate was,
no doubt, partially removed from its normal position in the
forming organs, because in most cases the males made no
effort to remove the spermatophore, although they had tried
to copulate. .

I have not seen that the males make any movements while
forming a spermatophore, but I would not positively deny that
they do so. I have frequently kept them under observation
for a few hours after copulation; some would form a new
sperm bearer and others not. One male copulated a second
time after a very short interval, but I did not watch him mean-
while. Some other specimens did not have the spermatophore
formed after hours of watching.

The female carries the vesicle very frequently until she is
about to mate again. If this comes soon after a previous copu-
lation, she will remove the vesicle. She does this with her
mouth parts, bending the abdomen ventrally. She may place
herself partly on her back and holding the abdomen against
the ground force her mouth parts back so as to reach the
ampulla of the spermatophore. If the abdomen is too much
distended by eggs, she frequently rubs it off by dragging the
abdomen on the ground. The longer a spermatophore has been
carried by the female the easier it is removed. In only one
instance did I see a female mount upon a male with the sperm
bearer still in place.. After a good many efforts the male suc-
ceeded in pushing the old spermatophore partially out of the
way and placing a new one. The female then carried both,
the old one apparently hanging on by one hook.

The Spermatophore.

Lespes says: “The spermatophore of the field cricket is com-
posed of.a vesicle almost round, of a brown, more or less dark,
color. It terminates at one extremity by a whitish papilla
and is continued on the other end by a transparent plate, al-

oy
4

BAUMGARTNER: COPULATION IN GRYLLIDAD. 331

most quadrilateral, formed by a thin membrane stretched over
three small cartilaginous pieces. One of these is median and
connects directly with the vesicle. It is tubular and contains
a horny thread which continues well beyond the plate. The
other two, situated to the right and left, are arch shaped, and
their two ends covered by the membrane, form on either side
two tooth-like hooks which serve to fix the apparatus in the
vagina. The size of the spermatophore varies a little but
ordinarily it attains almost 4 mm. from the papilla to the end
of the plate. The thread appears longer or shorter according
as it projects more or less from its tube. The spermatophore
is of different consistency in different parts. The vesicle is
extremely solid and its walls are very thick, while the plate
and its cartilaginous thread are soft at the moment of
copulation ; besides, it is covered by a white, thick liquid which
presents to me all the characteristics of a sperm fluid. The
thick, hard vesicle, composed apparently of double walls, to-
ward the extremities is hollowed out at the center. The round-
ish cavity is full of sperm. At one end a blind tube projects
from the cavity into the ampulla. At the other end the cavity
continues as a straight canal through the whole length of the
plate. This canal contains the horny thread.

“The white liquid which covers the plate when first removed
from the female shows, when examined with a high magnifier,
a large number of small, thread-like zodsperms. They are
about .04 mm. long and .002 mm. thick. These same bodies in
great number fill the cavity of the vesicle. They were never
united into feather-like bundles, nor did they ever show any
movement. The fluid taken from the testis of the male or the
copulating pouch of the female contains many very similar
zoosperms. These never exhibited movement, however
treated. Put into water, some of them twisted themselves
into knots.

“When the female lets the spermatophore fall the walls of
the vesicle are slightly ridged, but it still contains some sperm.
The horny thread is no longer found in the tube of the plate.”

I have quoted the above description at length because it
gives many facts as I find them in our American species.

The vesicle and the plate are similar to those described
above (see figs. 1 and 2). The thread is not contained in the
tube, but it is the tube itself continued, and it is much longer,
and bent back so as to run parallel with the plane of the plate

332 KANSAS UNIVERSITY SCIENCE BULLETIN.

and vesicle (see fig. 1). It is a hollow tube for its whole length
and the sperm comes out at the end of the thread, and not at
the end of the plate, as Lespes supposed. The thread varies
in length not because “it projects more or less from the tube”
but because it is broken off in removing it from the insect.
Many of those that I secured at first had more or less of the end —
broken off. It is easier to get a spermatophore with a complete
thread from the male than from the female, especially if the
latter has carried the apparatus for some time. This no doubt
accounts for the fact that the apparatus can be more easily re-
moved after it has been carried for a while. When the insect
drops it normally the thread has been broken off, or, more
probably, partly dissolved away; but it is not entirely gone,
as Lespes states.

The vesicle is apparently double-walled as the Frenchman
observes (figs. 1, 2,and 3). The inner one is thicker and more
granular, while the outer one is harder and quite transparent
(figs. 1 and 2). It is the inner wall that continues forward
and forms the thread, although this substance is less granular
than that around the vesicle. The substance forming the
plate is somewhat granular, and in some cases took the stain
just like the inner coat of the vesicle. In the lateral parts of
the plates I saw in several of my preparations many vacuoles.
The thread in many of the specimens fixed and mounted in
balsam shows a tendency to break away from the plate (fig. 3).

That the sperm pass out at the end of the thread was nicely
shown by several specimens which I placed into a normal salt
solution. The sperm flowed out very regularly and moved
about for a time. In one instance it took about fifteen minutes
for a vesicle to empty itself through the end of the thread.
After the sperm were all out a little fluid with small granules
flowed out. I do not agree with Lespes when he says that
some sperm remain in the vesicle normally; nor do I think
that the wrinkling is a characteristic of the empty vesicle. I
could not see any difference between an empty spermatophore
and a dark brown one which had been carried by a male for
some time. The wrinkling, which Lespes correctly describes,
is due to the longer hardening of the vesicle and does not seem
to affect the cavity on the inside.

The shape of the cavity differs from that described and
figured by Lespes in the total absence of the blind tube extend-
ing into the papilla. What he describes is only the second layer

BE ie” “Mata a

BAUMGARTNER: COPULATION IN GRYLLIDA. 333

of the wall of the vesicle as shown in figures 1 to 3. The dark
stain here shows the outer limits of the inner lining, and the
cavity is on the inside, as shown in the outline drawings. The
cavity is easily visible in the preparations but the photos do
not bring it out. At the posterior end the cavity is flattened
and at the other end it gradually tapers down to the size of the
thread; so that the shape of the opening is much like a flat-
bottomed flask. In this cavity all the sperm have their heads
directed toward the openings into the thread and their tails
toward the papilla. Figure 11 is a small group of sperm
photographed from a section of a spermatophore. In passing
out of the thread the sperm manifested some movement of
their own, but were not very active. Several may pass out
together but only two or three heads have room side by side.
The sperm are larger than Lespes thought, as can be seen by
comparing with my earlier published measurements (1).

_ The Spermatophore-forming Organ.

Since many of our textbooks on entomology—see Packard
(18) and Henneguy (10)—still cite Gryllus as showing one of
the simpler conditions for the excreting ducts of the testes
among the insects, I shall give a more detailed account of the
male generative organs, their ducts and the accessory glands,
than I should otherwise do.

The first description of these parts was given by Dufour (5),
in Gryllus campestris, then by Lespes (11), in G. domesticus,
later by Berlese (4), again in G. campestris, and then by Fen-
ard (6), in both of the above species.

The testes lie on either side of the abdomen above the ali-
mentary canal and extend from the second to the sixth or
seventh segments. The shape is much like that of an elongated
strawberry. On the outside is a thin membrane which encloses
the hundreds of straight or slightly curved tubules which
make up the organ (figs. 6 and 7). Their blind ends are
nearly all directed more toward the posterior end of the body of
the animal. At the center of the testes is a large tube, or
rather an irregular sinus, which receives the openings of the
hundreds of tubules (fig. 8). The small vas deferens leaves
this sinus on the anterior ventral side and comes to the ventral

side of the testes a little back of the middle. It passes back-

ward over the tubules, and leaving them at the outer posterior
edge passes back along the body wall muscles till it reaches the

334 KANSAS UNIVERSITY SCIENCE BULLETIN.

ninth segment, where it passes around the cercal nerve and a
few strands of muscle and ligament (figs. 7 and 8). Making
a sharp bend it goes forward and ventralward. It is very
much coiled upon itself from the bend on (figs. 7 and 8). The
loops of the coils may lie in planes passing dorso-ventrally or
laterally. The two vasa deferentia join and immediately empty
into the posterior ventral angle of the sinus, which receives
the openings of the annexed glands (fig. 9). Beginning at
the cercal nerve and sometimes a little in front of the bend ~
the duct is swollen so as to be 0.5 to 0.75 mm. in diameter
(figs. 7 and 8). Fenard says there are from four to five coils
and that the total length is 8 mm. In our common field cricket
I found no regularity about the position or number of coils
and the total length is more than 10 mm.

This swollen part of the vas deferens is always in the mature
adult filled with a mass of sperm, all of them with their heads
turned away from the testis.

Fenard (6), as well as the older investigators, says that “the
union of the two vasa deferentia forms:the ejaculatory duct.”
This is the statement generally made of the insects higher in
the scale of evolution; but it is true only in a certain sense—in
the sense that it serves as a common duct to carry out the
secretions of the testes. First, it is not a true ejaculatory
duct in the sense that it “during coition conducts the sperm
into the copulatory pouch of the female’—Packard (18).
This is true only in those insects with a penis; but in the
Gryllidz, where there is a spermatophore formed there can be
no such function. Second, nor is it true in the sense that it is
formed by a growing together of the two vasa deferentia. It
is formed by an invagination from the outside, as was first
shown by Palmen (17) from the anatomical standpoint and by
Nusbaum (16) from the embryological. This invagination
projects quite far beyond the point at which the two vasa def-
erentia join it (fig. 7). It extends dorsalward and forward,
where it receives the openings of the hundreds and hundreds of
tubules of the annexed glands (fig. 9). The lumen is very
much flattened laterally and widened vertically towards the
blind end. The more muscular tube leaves the sinus at the
posterior dorsal angle just above the point of entrance of the
vasa. This is well shown in figure 9. The openings of the vasa
cannot be recognized in this figure, but the serial sections leave
no doubt as to its location.

I

BAUMGARTNER: COPULATION IN GRYLLIDA. 335

The hundreds of glandular tubules which empty into the up-
per end of the common duct vary greatly in length and di-
ameter, apparently depending in part upon the room they find
for development (figs. 6,7 and 9). They take up all the avail-
able space, completely surrounding the seminal vesicles, ex-
tending far backward beside the rectum and the spermato-
phore pouch, or crowding forward and upward around the
intestinal tract. Many of them are short while others are long.
The whole group of tubules forms a large mass which takes
up quite a space (fig. 6).

The “prostatic gland,” first noticed by Berlese and later
described by Fenard, I found “only after much difficulty,” as
the latter says. From my dissections I had concluded that the
part so described was a part of the irregular pouch which
molds the plate of the spermatophore. But a careful study of
serial sections reveals the presence of a pair of oval glands or
pockets attached on either side to the common duct. They

‘have very short connecting tubes and they lie in the muscular

and connective tissue surrounding that part of the common
duct between the ganglion and the mold. The glands are not
surrounded by “yellowish fat,’ as Fenard suggested; but they
have “thin walls and are filled with a clear fluid.” This fluid
seems to harden into a clear yellowish chitin-like substance
when treated with fixatives. In this it resembles the secretion
from the annexed glands; but their secretion is darker and
more granular. Judging from the behavior of the secretion
and the position and connections of these glands I think they
do not furnish a lubricant, but probably have something to do
with the formation of the more flexible thread and the plate.

The common duct, after passing close to the floor of the ab-
domen, turns dorsalward and then empties into the cavity of
the spermatophore mold (fig. 9). This apparatus was care-
fully studied by Lespes (11). Many of his observations are
keen and to the point, but his description in some points is
not very clear. Berlese (4) studied this structure too, but he
could not get away from the idea that he was dealing with the
ordinary structures, so he applies the name “penis,” etc., to
the parts and so makes his description misleading. Fenard
(6) pays no attention to the hard structures and so he does
not make out the relations and functions of parts in this region.
Peytoureau (19) studied mostly only the morphological rela-

336 KANSAS UNIVERSITY SCIENCE BULLETIN.

tions of the hard parts and so he does not explain the functions
of the parts of this organ.

Lespes describes this apparatus “charged with the formation
of the spermatophore” as follows:

“It is situated at the posterior part of the ejaculatory duct
and may be considered as a part of the genital armature. In
the field cricket this armature is formed, first, by a curved plate
composed of many pieces extending to the right and left and
united by a tough membrane; and second, by a very remarkable
independent inferior part. The plate itself is composed of six
solid pieces. Above is a kind of shield-like piece which is
easily divided along the median line. Posteriorly it ends in
three hooks, of which the middle one plays an important part
in copulation. To the right and left are long slender rods upon
which are inserted the fleshy pallets which support the vesicle
of the spermatophore. Below this shield are two others, thin
and curved, with an irregular contour. Finally, on the inferior
face of the dorsal shield are two small, slender plates. During
copulation the median hook only enters the vagina and serves
to conduct the plate of the spermatophore. The independent
piece is formed in front by a sort of short stylet which is ap-
plied to the inferior face of the plate of which I have just
spoken. At the base this stylet enlarges, changes its nature
_and forms a white plate with cross striations. This plate is
bent along its length, forming an irregular circle. It is swollen
into a very thin vesicle situated under the genital armature
and in front of the two pallets which support the ampulla of
the spermatophore. This pocket is largely open in front and
under the genital armature. The ejaculatory duct, after
having passed around the lower part of the spermatophore
pouch, ends near this opening.

“This white plate is the organ for the production of the
spermatophore. On its convex surface is a transparent line ©
in which is formed the tube which holds the thread of the
spermatophore. The size of the plate is greater at the two
ends. At the terminal part of the enlarged inferior portion
ends the ejaculatory duct.”

Lespes claims to have seen the spermatophore while it was
forming. “It had a very thin membrane. When completely
formed the vesicle was removed out to the place between the
two fleshy pallets, and the plate and the thread remained in

i ala ;

ee

BAUMGARTNER: COPULATION IN GRYLLIDZ®. 3387

the forming organ, that is, in the stylet of the armature, until
given off to the female.”
This description is somewhat indefinite because of the un-

certain meaning of the phrase “genital armature.” Besides,

the application of the words “in front” and “behind” is not
certain. I do not believe that the differences between Lespes’s
description and my observations can be accounted for by sup-
posing such differences exist in the different species studied.
Both of us used Gryllus domesticus some, and as shown below
we found only small minor differences between the house
cricket and our own species of field crickets.

Before discussing the formation of this peculiar spermato-
phore farther let me state as clearly as possible the mechanical
problem involved. There is a mold made up of a system of
bracing rods of chitinous material and a thin fibro-muscular
membrane. Into this mold leads a single tube, which has to
bring the sperm and the secreted fluid for the covering. The
ampullar part of the finished spermatophore has apparently
a double wall, and so there may possibly be two kinds of secre-
tions, although a careful study of the gland shown in figure 10
makes me doubt it. The question is, then: How are the sperm
kept separate, and especially how are they held in the center
while the soft liquid is poured around them and hardens? If
there is a second layer in the wall the incomplete ampulla would
have to be held in the center while the second inflow and hard-
ening occurs. To place a semifiluid substance within another
fluid substance and hold the former at the center while the
latter hardens would be a difficult feat for any mechanic.
I do not see how the apparatus provided is sufficient to ac-
complish the result. Yet I have seen so many well-formed
spermatophores that there can be no doubt of the end product.
_ The way to solve the problem readily suggests itself. One
needs to get a lot of specimens showing the various stages in
the formation of the vesicle. Such a series of stages I have
been hoping to get, but have thus far not succeeded. In all
my many dissections I have never found a partially formed
vesicle. It is hardly feasible to follow the process in a living
specimen, as the attempts to make observations would interfere
with the normal formation.

Although I am not able to give a description of the exact
method of formation, I shall nevertheless make some sugges-

2—Univ. Sci. Bull., Vol. V, No. 19.

338 KANSAS UNIVERSITY SCIENCE BULLETIN.

tions. Lespes says that the ampulla is formed and then moved
out to its position between the two pallets. I think that this
cannot be correct, as it would mean that the plate and thread
would have to be joined onto the ampulla. In ali the specimens
I have seen the inside cavity and the walls are just as smooth
as they can be. I cannot imagine that if the apparatus is
formed in two parts they could be joined together so exactly
and smoothly. I would rather believe that the structure is
formed as a whole in situ. The thing that gave the idea that
it moved back is the fact that when the wall has partly hard-
ened the pallets retract a little, exposing the ampulla.

When there is no spermatophore in a male cricket the hooks
are drawn far down, the pallets are much contracted and
folded, and a part of the loose tissue on the floor of the cavity
is rolled up and fits into the mold as a sort of plug. The pallets
and the plug have deeply grooved and ridged surfaces vier
line the small lumen within.

Things occur about as follows when a spermatophore is
formed: The inner lumen becomes a little larger and assumes
the shape of the cavity of the vesicle. Then the sperm pass
down, filling it up. It is possible that the lining cells secrete a
small amount of substance to form a thin membrane, and
Lespes may have found an example in this stage. If a mem-
brane is secreted it cannot be discerned later. Now some secre-
tion comes down the common duct. This flows in and fills up
the grooves. The ridges withdraw, forming new grooves to
be filled up. The process continues till enough is laid down
to form the inner layer of the wall. Then the flow ceases for
a while, and the secretion hardens somewhat. After a longer
or shorter time a second inflow of secretion occurs, and it takes
the place of the withdrawn ridges till the surface of the lining
membrane becomes smooth, thus forming the outer wall of
the ampulla. Then the whole hardens into the completed sper-
matophore. I think the whole takes place in a short time rel-
atively. There may be movements accompanying the process,
although I have not observed them.

Since there is but one duct, the only other assumption which
could be made is that the secretion and the sperm come down
all together instead of in succession. If this be correct, then the
long sperm would have to wiggle their way to the center. This
is hardly possible. Besides, it would be difficult to explain the
second layer in the wall of the ampulla by this hypothesis.

BAUMGARTNER: COPULATION IN GRYLLIDA. 339

Another mechanical problem appears in the study of the
spermatophore, namely, What causes the sperm fluid to flow
out of the ampulla? Are they forced by contraction of the
walls, or is there some substance that swells and forces them
out? Does air penetrate the wall in some way and thus prevent
a vacuum?

I do not believe that there is any contraction; at least not
much. The walls do not collapse. Farther than this I could
find nothing that would throw any light on the problem.

Differences Found in the Domestic Cricket.

Lespes describes the spermatophore and the spermatophore-
forming organ of this species, but says that he has not seen the
act of copulation. I can readily believe the latter statement,
as the insects are much more shy and will not mate readily
when watched. Nevertheless I witnessed the process several
times. The behavior during courting and the movements dur-
ing the transfer of the spermatophore are very similar to
those of the field crickets. The differences are of no conse-
quence. :

The spermatophore is lighter in color, as Lespes says. But
this is in accord with the general lack of color in this species.
It is a little shorter and stouter, as the Frenchman indicated.
The plate is also narrower and more bent. But the cavity does
not extend back into the papilla, as I have already explained
about our field crickets.

There is one point to which I should like to call attention.
Lespes does not speak of it, yet shows a difference in the draw-
ings. In the house cricket the thread is a direct continuation
of the canal. This is correct for this species, and is precisely
what I find in our field cricket. I have not seen the spermato-

ihe. ‘ __ phore in G. campestris; but judging from our domestic and field
-_ erickets, and the conditions I find in Nemobius, I do not. be-

_ lieve that Lespes is correct when he indicates a break between
the thread and the plate, as he shows in his drawings, figure 2.
The canal could not be continuous if his drawing is correct. But
Lespes thought that the thread was the mass of sperm, and
so not thinking of the thread as a hollow conducting tube he
might have a break in it. -
The mold differs from that of our field cricket about as it.
differs from Gryllus campestris, judging from Lespes’s descrip-
tion. The hooks on the superior plate are rudimentary or

340 KANSAS UNIVERSITY SCIENCE BULLETIN.

absent, while the stylet is well developed and projects far
backward. The cavity is smaller, making it necessary that
the plate of the spermatophore be bent into a smaller curve.
There are other minor differences in size and relations.
Whether these differences are great enough to ‘prevent inter-
crossing of species is not positively settled. I tried some ex-
periments in cross-breeding, in which I was unable to get any
cross matings; but the tests were not extensive enough to be
very conclusive.
IN NEMOBIUS.

This genus resembles rather closely the form which Lespes
describes under the name Gryllus Sylvestris. It is much smaller
than our Gryllus and, like the European form, it likes the
woods or shady places; in the grass under the trees of a park
I have found the most specimens. In the months of August
and September they are very numerous.

In its courting it is, like the domestic cricket, much more
_ shy than the larger form, and it is rather tedious to watch till
one catches a pair copulating. In the main the whole series
of processes is quite similar to that in Gryllus. In the courting
the male is more active while the female less readily yields to
the courting. What Lespes (12) says of his forms holds true
for our two genera: “Except some details of form and size
all is similar. The spermatophore is smaller and more fragile
but is similarly composed. The genital armature presents the
same pieces, but they differ much in form.” Lespes again in
this species figures and describes the thread as continuous with
the canal of the plate. This strengthens my belief that there
is some mistake in the description of the spermatophore of
Gryllus campestris.

I have succeeded in getting a good photograph (fig. 4), that
shows the structure of the spermatophore of Nemobius fas-
ciatus, our most common species, very well. The ampulla is
almost a perfect sphere. Its wall is thick and its central cavity
rather small and nearly spherical. As in Gryllus sylvestris,
the plate is small and far from the ampulla, but the thread and
plate are much more curved in our species (fig. 4). There isa
fine, hollow, tube-like opening leading from the central cavity
to the tip of the thread. There can be1 no doubt that the —
fluid flows out through it.

BAUMGARTNER: COPULATION IN GRYLLIDA. 341

IN GRYLLOTALPA.

Fenard (6) cites the fact that the Locustide and the Gryl- .

lide produce spermatophores and that those two families have

the large mass of annexed glands. Continuing, he says: “Then

wherever we find the glandular tubes similar to those in the
above-mentioned groups we have a right to conclude, until
there is proof to the contrary, that they secrete a substance
destined for the formation of spermatophores more or less
complex.” He suggests that in Gryllotalpa, whose copulation
has never been seen, fertilization occurs by means of a sperma-
tophore, because they have the annexed glands similar to
those of Gryllus.

In the abstract (2) I made the statement that in the mole
crickets the sperm are transferred by means of a spermato-
phore. In order to observe copulation I kept four mole crickets,
two pairs, for several weeks each in a separate aquarium partly
filled with wet sand, grass roots and slices of potatoes. For
five nights I placed them by pairs into battery jars which con-

- tained just enough sand to allow the animals to make a burrow

around the edge. Thus I could watch them continuously. After
some twenty hours I finally observed copulation. As a result
of these vigils I am able to give the facts concerning chirping
and the protective gland published elsewhere (3), as well as
the following concerning copulation.

The courting is somewhat similar to that in Gryllus. The
male calls the female with loud, long chirps. As she ap-
proaches the chirps become short and much softer. He then
frequently turns his abdomen towards her. As the pair get
ready to copulate the position assumed is quite different from
that of any other animals of which I know. They turn poster-
ior end to posterior end, and ventral side to ventral side, so
that the cloacal openings are just opposite each other. The
female stands erect with her abdomen slightly raised, while
the male lies on his back. The abdomens are tightly held to-
gether by hooks, described with great detail by Peytoureau
(19). The sperm were carried to the female by a spermato-
phore. The time it takes for the transfer is not over a minute;
but the pair kept their relative position, the abdomens simply
touching each other, for more than ten minutes. After dis-
turbance the male followed the female and again assumed this

342 KANSAS UNIVERSITY SCIENCE BULLETIN.

relative position, but no further transfer of a spermatophore
occurred.

Right after the first separation of the couple the female be-
gan to chew at a part of the sperm vesicle, and only desisted
when disturbed by the movements of the male. She was also
perfectly quiet as soon as her abdomen was touched as in the
position of copulation. One.observation is not enough to estab-
lish the purpose of an act; yet, judging from the behavior of
the individuals of this pair, especially from the anxiety of the
male to touch the female with his abdomen, one could easily
conclude that the long lying in the position of copulation was
to prevent the female from chewing at the spermatophore too
soon and thus preventing the proper injection of the sperm.
After the pair had remained in the above-described position
for ten minutes or more the female left the male and mani-
fested no further desire to chew the spermatophore. She
carried it for about a half an hour and then dropped it.

As the vesicle was being transferred, or just after it had
been put in place, there was an outflow of some transparent
fluid on either side of the vesicle. This soon hardened. It
is this part of the apparatus that the female was chewing. The
spermatophore was found to consist of an oval ampulla which
contained the sperm in the cavity at the center. At one end
of the ampulla there is a projection by which the apparatus is
held in the vagina, and through which the sperm are carried
into the spermatheca. On either side of this projection is an
irregularly shaped mass formed by the above-mentioned out-
flowing fluid during the transfer. The two sides are unlike, as
part of one side was pulled and eaten away by the female, and
the other side was pressed out of shape by some falling sand
before it had time to harden. Having but this one imperfect
specimen of the spermatophore of these crickets I cannot give
as accurate a description of its detail as I should like. Suffice
it to say that it has the essential parts—a hollow ampulla to
contain the sperm and a projecting part to fasten it to the fe-
male and to carry in the sperm. |The structure of the latter is
obscured by the irregular mass.

The outflow of this fluid recalls an observation which I saute
repeatedly in the Chicago greenhouses. In the copulation of
a locustid the same phenomenon occurs, but i in a more marked
way. The species, an exotic one introduced from Japan; is

BAUMGARTNER: COPULATION IN GRYLLIDZ. — 348

very abundant in the greenhouses. The male has no chirping
organs and so the courting is all done by means of the long
antennze. The female mounts on the back of the male, when
he hooks an almost spherical ampulla full of sperm into her
vagina. During the latter act a viscid fluid flows out on either
side and forms two somewhat irregular roundish masses larger
than the original ampulla, which now lies between the two.
Sometimes the female began immediately after copulation to
eat this substance, doing it more readily if the couple was dis-
turbed.

As to a use or meaning for this outflow of fluid and its sub-
sequent hardening it is difficult to suggest a satisfactory one.
Of its regular occurrence there can be no doubt, as I observed
it many times in Diastremmena and once in Gryllotalpa. It
may function as an additional fastener to hold the apparatus in
the vagina. Yet this seems hardly necessary, as the spermato-
phore apparently remains in place before it is hardened. It
may have something to do with the emptying of the vesicle,
contracting around the ampulla, thus forcing out the sperm.
Or it may be concerned with the removal of the spermato-

phore from the vagina, acting as a bait for the female to re-

move it. aM MARY.

1. The reproductive apparatus of Gryllus is not of the
simplest type, as was reported by the earlier investigators and
as is still suggested by our textbooks. These organs are rather
of the most complex kind found among the insects. They were
studied and deseribed with a good deal of detail, and the cor-
rect method of the transfer of the spermatozoa was indicated
by Lespes in 1855; but the work was discredited and neglected.

2. The male carries the sperm to the female in a special
structure, the spermatophore, formed by the secretion of the
glands annexed to the common duct.

3. The essentials of the spermatophore, as shown by the
study of several genera of Gryllidz, are a hollow vesicle to act
as a retainer, and a hollow thread-like tube which carries the
sperm into the spermatheca of the female.. To the tube are
added some enlargements which serve to hold the structure in
the vagina.

4. The so-called genital armature of the male crickets con-
sists of a mold for forming the spermatophore and an appa-
ratus to transfer the same to the female.

344 KANSAS UNIVERSITY SCIENCE BULLETIN.

5. The differences between the field and domestic crickets
is probably great enough to prevent intermating, yet this is not
at all proven.

6. Gryllus has no gland which secretes a lubricant for the
copulatory organs. The so-called ‘“‘prostatic gland’ has to do
with the formation of the thread and plate of the spermato-
phore.

7. Mole crickets form a spermatophore. A transparent

viscid fluid flows out during the transfer. The male and the
female assume a very peculiar position relative to one another
during copulation.

UNIVERSITY OF KANSAS,
March 12, 1910.

BIBLIOGRAPHY.

1. BAUMGARTNER, W. J., 1902. Spermatid Transformations in Gryllus
assimilis. Kan. Univ. Sci. Bull., vol. I, No. 2.

2. BAUMGARTNER, W. J., 1905. Observations on Some Peculiar Habits
of the Mole Crickets. (Abstract.) Science, N. S., vol. XXI, No. 544.

3. BAUMGARTNER, W. J., 1910. Observations on the Gryllide; III.
Notes on the Classification and Some Habits of the Crickets. Kan.
Univ. Sci. Bull., vol. V, No. 18.

4, BERLESE, ANTONIA, 1882. Ricerche sugli organi genitali degli Ortot-
terri. Atti della Accademia dei Lincei, 1882.

5. Durour, LEON, 1841. Recherches anatomiques et physiologiques sur
les Orthoptéres, les Hymenoptéres et les Neuroptéres. Mem. d.
Savants etrangers, VII, Paris.

6. FENARD, A., 1896. Recherches sur les organes complementaires in-
ternes de l’appareil genital des Orthoptéres. Theses presentees a
la Faculte des Sciences de Paris. Lille.

7. GIRARD, MAURICE, 1873. Traite elementaire d’entomologie, vol. I,
Librairie J. B. Baillere et fils, Paris.

8. GIRARD, MAuRICE, 1885. Traite elementairie d’entomologie, vol. III,
Librairie J. B. Baillere et fils, Paris.

9. GRABER, 1877-’79. Die Insecten, vol. I-II, Mtinchen.

10. HENNEGUY, L. F., 1904. Les Insectes, Masson et Cie, Paris.

11. Lespes, CHAs., 1855. Memoire sur les Spermatophores des Grillons.
Ann. Se. Nat., T. ITI.

12. Lespes, CHAS., 1855. Deuxieme Note sur les Spermatophores du
Gryllus sylvestris. Ann. Sc. Nat., T. IV.

18. Leypic, FRANZ, 1867. Der Eierstock und die Samentasche der In-
sekten, zugleich einen Beitrag zur Lehre der Befruchtung. eve
Acta Acad. Leop. Car., T. XXXIII, Dresden.

14. Los, J., 1907. Comparative Physiology of the Brain and Compara-
tive Paycholors: New York.

15. MitNe-Epwarps, H., 1870. lLecons sur la Physiologie et Anatomie
comparee de Yhomme et des animaux, T. IX.

BAUMGARTNER: COPULATION ON GRYLLIDE. 345

, 1882. Zur Entwickelungsgeschichte der Ausfiihrungs-
: e der sexualdriisen bei den Insecten. Zool. Anz., Bd. V.
_PALMEN, J. A, 1884. Uber mires, 6 Ope aman des gesch-

Bicics, A. S., 1903. a Textbook of Entomology The Macmillan
_ Company, New York.

20, Sinnoto, Cc. T., 1887. Hecsacs Webbie ‘ec dio Bpecmeto-

_ zoen bei wirbellosen Thieren. Mueller’s Archiv.

Stepotp, C. Ty 1845. Uber die Spermatozoiden der Locustinen. Nova
te SAT

Sn Vergleichende Anatomie und Physiologie der In-
| _ Die Lbsscsios a Pereioineresns der Kafer, Berlin.

PUBLISHED BY THE UNIVERSITY,
LAWRENCE, KAN.

March, 1912:

sere
sali eee

Pes

The ae
Ss

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vo. V, No. 20.] MARCH, 1910. von av we

VoL. XV, No 20

AN EXAMINATION OF THE CHROMOSOMES OF
' ANASA TRISTIS.
BY C. E. M’CLUNG AND EDITH PINNEY.
I. Introductory statement.

II. The chromosomes during the spermatogenesis of Anasa, an inde-
pendent study: By the junior author.

Ill. A comparison of the results of studies upon the chromosomes of
Anasa: By the senior author.

1. Methods.

2. What is the number of spermatogonial chromosomes, and of
odgonial chromosomes?

3. What is the number of first spermatocyte chromosomes?

4. What is the behavior of the accessory chromosome and of the
plasmasome in the first spermatocyte prophase?

5. What is the behavior of the accessory chromosome in the first
spermatocyte mitosis?

6. What is the behavior of the chromosomes in interkinesis?

7. What is the behavior of the accessory chromosome in the second
spermatocyte mitosis?

8. What is the final chromosome constitution of the spermatozoa?

I. INTRODUCTORY STATEMENT.

The extensive studies now in progress upon the chromo-
somes of numerous species of plants and animals have led to
the formulation of several working hypotheses of considerable
general importance. In order that these may be of value and
serve as an aid to further progress, it is necessary that the
phenomena from which they take their origin be extensively
recognized and uniformly interpreted. Any conception of the
relations existing between germ-cell organization, as exhib-
ited by the chromosomes, and the structural characters of the
body is now unfortunately delayed by certain disagreements

(349)

350 KANSAS UNIVERSITY SCIENCE BULLETIN.

upon the actuality of the phenomena upon which such inter-
pretations are based.

Fundamentally, these disagreements concern more or less
intimately that view of chromosome organization which is
called the “theory of chromosome individuality.”” Modern
cytologists are either supporters or opponents of this theory,
and its adequacy or inadequacy will be determined by the
weight of evidence now accumulating. It is, of course, im-
possible to conceive that it will stand unmodified, for it is but
the expression of our present limited knowledge; but it must
now be determined whether these structures of the cell have
persistent individualities or whether they are merely incidental
and inconstant expressions of more fundamental phenomena.
Observed facts must support one or the other of these alterna-
tives, although in doing so they may readily enough modify
the established order.

A considerable body of evidence has accumulated within re-
cent years to support the theory of chromosome individuality
and to associate the development of sexual characters with
particular chromosomes. To this has been added the support
of experimental work on sex determination and the parallel-
ism between the segregation of Mendelian characters and the
behavior of the chromosomes in the maturation divisions of
the germ cells. Much of this cytological work has been done
upon the insects, particularly in the orders Orthoptera and
Hemiptera, and, in the large, there is a strong consensus of
opinion. More recently confirmatory evidence has been fur-
nished from studies upon nematodes, including Ascaris, by
Boveri, upon echinoderms by Baltzer, and upon birds and man
by Guyer. Here and there, however, there are workers who
question the fundamental facts and deny the correctness of the
theories founded upon them.

It is necessary, in order to proceed further with our inter-
pretations, that we reach an agreement upon the observable
facts in our science, and no honest question should remain
without answer.

At present one of the most serious diversities of opinion
exists regarding the character and the behavior of the acces-
sory chromosome in the bug Anasa tristis. Because of the ex-
tensive and painstaking studies of hemipteran germ cells by
Wilson, the contradictions and inaccuracies in the studies of

CHROMOSOMES OF ANASA TRISTIS. 351

Paulmier and of Montgomery seemed to be cleared up and the
phenomena in this order brought into agreement with those
of the other insects. To this was added the confirmatory
work of Stevens upon several orders of insects, the study
of Lefevre and McGill on Anasa, and more recently that of
Morgan on aphids and phylloxerans. Directly opposed to all
this are the observations of Foot and Strobell upon Anasa
tristis, the form studied in detail by Paulmier, Montgomery
and Wilson. Their contention, in brief, is that the number
of spermatogonial chromosomes is even instead of odd, that
there is no dimorphism of the spermatozoa due to the presence
in one-half their number of the accessory chromosome, and
that there is no evidence to show that the theory of chromo-
some individuality receives any support from observed mor-
phological continuity of any of the chromosomes. Aside from
these major questions there are others, such as the history
and fate of the plasmasome, upon which these authors are in
disagreement with other investigators.

The issues here defined are so clear-cut and definite and
their settlement of so much importance that the senior author
of the present paper was led to offer his services for a reéx-
amination of the whole question. This offer met a hearty
response from Professor Wilson, who sent not only live mate-
rial, but his entire series of mounted slides, including those
used by Paulmier. Professor Lefevre has been kind enough,
‘also, to permit the use of his slides. Miss Strobell, writing
for Miss Foot, expressed a lively interest in the proposed set-
tlement of this problem, but was unable to send the prepara-
tions used in their studies, because these had all been de-
stroyed on account of limited storage room. It is to be
regretted that this material is not available for comparison,
since it is upon it that the only divergent observations have
- been made; but since specimens from the same locality, pre-
pared by all methods, including the special ones employed
by the Misses Foot and Strobell, are at hand, there can be no
reasonable doubt that it is representative.

*To reduce the possibility of the influence of preconception
to the lowest possible degree, the junior author has worked
through the entire problem without knowledge of the work
done by others, and without promptings from the senior
author except such as were given by a series of questions for

352 KANSAS UNIVERSITY SCIENCE BULLETIN.

which answers were desired. Independently, the senior au-
thor went over the same ground, first upon material prepared
in this laboratory, and later upon the slides of Paulmier, Wil-
son and Lefevre. Whenever possible, photomicrographs were
made, and representations of them accompany the paper. The
two authors reached concordant results throughout, and their
common experience is embodied in what follows.

Since the matters of dispute are purely of facts and inter-
pretations, it will be necessary to take a definite position on
each of the points involved, but so far as possible the material
evidence for the decisions will be presented fully. It is hoped
that this may be complete enough so that others interested
may be able to form their own opinions directly from the ap-
pearances themselves. There will be presented first the per-
sonal work of the junior author, giving her independent
results, and this will be followed by a general discussion of the

problem by the senior author.

Il. THE CHROMOSOMES DURING THE SPERMATOGENESIS OF
ANASA TRISTIS. ‘

BY EDITH PINNEY.

Preparatory to an attempted settlement of various disputed
questions in the development of the germ cells of Anasa tristis
an independent reéxamination of the entire process was under-
taken. In order to avoid unconscious bias of opinion the pre-
vious reports on the spermatogenesis of this species were not
read. Without referring, therefore, to the accounts of other
observers for comparison or correlation it is purposed to record
the results of these observations.

Material.—Preparation and general description.

Material was taken from specimens of Anasa collected in
Lawrence, Kan., and in Woods Hole, Mass. The latter was
obtained through the courtesy of Dr. E. B. Wilson, of Columbia
University. All material was fixed in Flemming’s fluid, the
period of fixation varying from one-half to two hours. The
longer fixation was found to give the better results. Paraffin
sections were cut five micra in thickness and stained by
Heidenhain’s iron-hematoxylin, Flemming’s tricolor and
Auerbach’s methods. Slides prepared by the author from
nineteen individuals were studied. Most of the drawings
were made at a magnification of 3625 diameters. This is re-
duced one-half in reproduction.

PINNEY: CHROMOSOMES OF ANASA TRISTIS. 353

The testes of Anasa tristis are paired organs in the form of
two small fig-shaped bodies which lie on either side of the
median line in the cephalo-ventral part of the abdominal cav-
ity. They are composed of a number of long, slender follicles
lying parallel to each other and extending the entire
length of the testis. The follicles themselves taper slightly
from the broader base toward the sperm-duct. The follicu-
lar organization into cysts of germ cells of varying stages
of maturation is similar to that in the Orthoptera, and the
same method which was used to determine the sequence of
the observed changes in the Orthopteran species is applicable
here. Practically all of the cells of a cyst are in the same
phase, although exceptions to this regularity are frequent in
the older cysts, particularly those of the first and second sper-
matocyte generations. These variations are, however, to be
expected, and may reasonably be accounted for if one considers
the unavoidable differences which must occur in the rate of
metabolism in cells situated differently in the same cyst.

The longitudinal section of a follicle presents some noticeable
features of a structural nature caused by characteristic dif-
ferences in the cells of various regions. The spermatogonial
cells occupy the distal portion of the follicle, the extent of the
spermatogonial area being determined by the age of the in-.
dividual. These cells are small, and in the older cysts the nu-
cleus is accompanied by little cytoplasm. Hence the peculiar
staining quality of the nuclear material causes this region to
appear more intensely stained than the proximal portion of the
follicle. The cells, too, are crowded closely together. In con-
trast to this is the lightly stained spermatocyte region, due to
the fact that here the cells contain a greater amount of cyto-
plasm in proportion to the amount of nuclear material.: The
cells have increased so greatly in size in the later generations
that the cysts are larger. Usually a characteristic barrier ex-
ists between the two main regions of divisions. This is formed
by cysts of cells undergoing the process which has been desig-
nated as synizesis, a detailed description of which will come
later. In almost every instance, too, the line of separation is
emphasized by the presence of many cysts of degenerating cells
with their spherical masses of chromatin,

354 KANSAS UNIVERSITY SCIENCE BULLETIN.

OBSERVATIONS ON THE SPERMATOGONIA.

From a study of polar views of the equatorial plate the num-
ber of spermatogonial chromosomes was determined to be
twenty-one, and in no case was there observed a departure
from this characteristic number. The chromosomes in young
cells are larger than those in older cells and do not exhibit the |
tendency to coalesce so readily. Consequently cells for study-
ing the number, form and behavior of the chromosomes were
comparatively few, although enough were found to leave no
doubt in the observer’s mind as to the correctness of these
observations and their interpretation. Figures 1, 2, 3, 4, 5, 6,
7, 8, 9, 10, show polar views of the equatorial plate from
spermatogonial cells. Certain elements of the complex exhibit
constant and striking characteristics in size relationships. The
largest three chromosomes in each group are easily identified,
as are also the smallest two. The remaining members of the

group more nearly approximate each other in size, so no at-
- tempt was made to homologize these in the several cells.

The form of the chromosomes varies from spherical and el-
lipsoidal to kidney-shaped with variations in size and position.
The smallest two elements show the most striking differences
in these respects from the other members of the group. There
seems to be no fixed arrangement of the individual elements,
except that there is an evident pairing of twin chromosomes.
Sutton reported the presence of similar pairs in Brachystola
magna* and his observations have been verified by other work-
ers in this laboratory on other species of the Orthoptera.+ One
element of such a pair is considered to be of maternal origin
and the other of paternal origin. The tendency for equal ele-
ments to get together is very marked even in young cells, and
makes conspicuous the presence of an unpaired element, one of
the largest three chromosomes. Should its mate exist in the
cell it could not possibly be overlooked when even the smalies*
elements are so prominent. Owing to uniformity in size, pre-
cisely which of the three is the unpaired element may be de-
cided upon only when two of them are unmistakably paired.
The smallest two elements show this tendency but rarely, their
behavior at this stage being consistent with subsequent irregti-
larities.

*“S pernane divisions of Brachystola magna.’”’ Kansas University Quarterly,
vol. 9, No. 2, 1900.

+ The Chromosome complex of Syrbula ee W. R. B. Robertson. Kansas Uni-
versity Science Bulletin, vol, IV, No, 13, 1908

PINNEY: CHROMOSOMES OF ANASA TRISTIS. 355

All of the chromosomes lie with their long axis in the plane
of the equatorial plate, and from a study of lateral views of
this stage it is evident that the plane of the split in the chromo-
somes coincides with the same plane. Ideal conditions for equal
division of chromatin are therefore present. Division is simul-
taneous throughout the group. The separated halves move
toward opposite poles, maintaining their relative positions un-
til the pole is nearly reached. Thus the shape of the spindle
during anaphase is altered to accommodate the varying condi-
tions. See figures 12, 13 and 14. There is no evidence as to
the method of spindle-fiber attachment, and it is doubtful
whether the matter can be decided in this material. Figure 15
is a drawing of a section through an entire cyst. Most of the
cells are in metaphase. A few are in anaphase. The polarity
exhibited in this section is typical of all cells of this genera-
tion.

In lateral views of anaphases in young cells the chromosomes
are distinct. In the older, more crowded cysts the tendency of
the individual elements to coalesce produces the appearance of
a solid band of chromatin in the equatorial plate. Later stages
of the same show two separate bands with spindle fibers be-
tween. The plate arrangement is lost when the pole is reached.
Figure 17 shows a polar view of an early telophase. The
chromosomes have not yet lost their identity, although they
have apparently begun to disintegrate. . Between the telophase
and the succeeding metaphase there is an increase in the amount
of chromatin. The formation of a nuclear membrane and the
further dissolution of the chromosomes follow. A typical telo-
phase from a young cyst is shown in figure 18. The daughter
nuclei are flattened, but assume a spherical outline as the diffu-
sion of chromatin proceeds. Figures 19 and 20 show the cul-
mination of this process.

_ The inauguration of the next mitotic cycle is marked by the
appearance of an exceedingly long and much coiled spireme,
whether continuous or not it is impossible to say. At any rate,
after the thread has thickened and shortened considerably it is
plainly seen to exist in segments which finally condense to form
the homogeneous bodies of the chromosomes. Figures 21 to
28, inclusive, show the successive steps in this transformation.
The shortening segments show various twists, bends and con-
tortions. See figures 22, 23, 24 and 25. Not until very late in
the prophase does the longitudinal split in the threads become

356 KANSAS UNIVERSITY SCIENCE BULLETIN.

visible. (Figs. 26 and 28.) At this stage pairing of equal
elements is very common throughout the spermatogonial gen-
erations. Note the marked pairs in figures 1, 2 and 4, plate
LXV. It would seem from this that throughout the history of
the spermatogonial cells there is a consistent provision for the
final processes which mark the transition from spermatogonia
to spermatocyte. This process is called synapsis. The cells of
a mature spermatogonial cyst have not been counted, but judg-
ing from the number of cells in a cyst of developing sperma-
tids, eight mitoses complete the activity of the germ cells as
spermatogonia, that is, that after the last division there are
256 cells in a cyst. In such cysts the cells are poor in cyto-
plasm.

After the final division the typical telophase is replaced by
synizesis. During this process the chromatin is contracted
into a deeply staining granular mass whose position within
the nucleus is either central or slightly eccentric, leaving the
peripheral area clear. It is quite certain that the chromatin
mass is composed of intimately associated granular threads,
for occasionally ends protrude from the knot, although entire
threads can not be traced. (Figs. 30, 33.) It is true that the
chromatin enters synizesis in the form of individual threads,
and the presynizetic stage is characterized by the same ele-
ments. It is reasonable to suppose that the threads continue
as such during the period of synizesis. That some change is
taking place within the knot is shown later, but there is no
evidence to show that this change involves the disorganization
of the chromatin threads and its reorganization into new ele-
ments. Nothing can be said in regard to the duration of
synizesis. That an end to end conjugation of the paired ele-
ments of the spermatogonial nucleus occurs during this period
seems established beyond a doubt.

At one side of the conjugating threads lies a definite chro-
matin body of homogeneous consistency. Occasionally this
body shows an uneven ‘outline suggesting a slight disintegra-
tion of the constituent chromatin. Its position is constant and
it is invariably present. From a study of subsequent matura-
tion processes it is shown to be the accessory chromosome
(McClung, Orthoptera) and can be identified as the homo-
logue of the unpaired element appearing in the spermatogonial
metaphases. In studying the history of this odd member of
the complex, drawings of it in that stage were made, as well as

PINNEY: CHROMOSOMES OF ANASA TRISTIS. 857

in numerous subsequent stages, for size comparisons. Fig. 14,
plate LX VI, shows several drawings from different individuals.
The spermatogonial cells from which drawings of the accessory
were made vary in age, and as there is undoubtedly a decrease
in the chromatin content of these nuclei as the cells multiply we
would naturally expect to find homologous chromosomes vary-
ing in size in the same individual. The position of the chromo-
somes with respect to the equatorial plane varies previous to
division, that is to say, in the formation of the equatorial plate
the chromosomes may lie obliquely to the plane of the equator,
and this would cause apparent differences in the size of camera
drawings. Conceding that these possible causes of size varia-
tion are sufficient to account for the slight variations in size
of the accessory as shown in the drawings, we are forced to con-
clude that the size of this element in the spermatogonia is
strikingly constant. A similar study of the homogeneous body
observed in the nucleus during synizesis shows a like con-
stancy in size. Although these two elements which appear at.
different times in the nucleus are so unlike in form we cannot
but consider them as equal chromatin masses. Moreover, if
synapsis is a conjugation of twin chromosomes, obviously the
unpaired element of the earlier stages would take no part in
this process. This evidence is sufficient to incline us to believe
that the two masses are identical. Further evidence will be
given to strengthen this conclusion. An account of the presyn-
izetic stages or growth period will be given in the observations
on the first spermatocyte generation.

The First Spermatocyte.

The earliest condition of the spermatocyte cell is the result
of the disentanglement of the chromatin knot formed during
synizesis. Even before the chromatin has spread to the nuclear
membrane, and while the peripheral area is comparatively free
from it, it is easy to determine that the chromatin exists in the
form of granular threads or rods. See figures 1, 2, 3 and 4,
plate LXVI. This diffusion of the chromatin threads makes
possible the detection of a second body similar in size and
staining qualities to the chromatin body described as present
during synizesis. This body is not present before synizesis, is
not apparent during the synizetic period, and is present after-
ward in the interior of the nucleus. The natural conclusion
then is that this is a product of synizesis. Its staining qual-

358 KANSAS UNIVERSITY SCIENCE BULLETIN.

ities indicate that it is composed of chromatin. The distinction
between this and the accessory is only possible when the ac-
cessory lies close to the nuclear membrane, which is its usual
position. In this respect it resembles the accessory chromo-
some of the first spermatocyte prophase of the Orthoptera.
The history of the three distinct nuclear components, active
chromatin, accessory chromosome and plasmasome, will be fol-
lowed separately up to the spermatocyte metaphase, beginning
with the chromatin. The reorganization of the nuclear chro-
matin into typical spermatocyte chromosomes after diffusion is
completed gives no hint of a reversal of the mitotic cycle.
That is, the appearances which characterize chromosome re-
construction do not suggest a reversal of the changes which
were observed to take place in the diffusing chromatin. The
straight threads which typify the early prophase (figs. 1-4,
plate LX VI) gradually elongate and finally are disposed into a
fine reticulum at the intersections of which are very slight ac-
cumulations of chromatin. See figure 4, plate LXVI, and
figure 4, plate LXVII. This might be considered as the climax
of diffusion, for in this stage the chromatin is most evenly dis-
tributed throughout the nucleus. Subsequently there are more
extensive accumulations of chromatin at certain points within
the nucleus. These amorphous aggregations of granular
chromatin gradually assume the form of loosely organized
tetrads resembling a few of the types of tetrads found in the
Orthoptera. Figure 16, plate LXVI, shows four types of tet-
rads taken from one cyst all of the cells of which are in
the same stage of development; (a) shows the largest form.
By analysis it is seen to be composed of four equal chromatids
and represents the pair of the largest spermatogonial chromo-
somes conjugated. The disposition of maternal and paternal
chromatids is indeterminable in this as well as in the forms
shown by (b) and (ce), which are different types of the cross.
In each case two opposite arms of the cross are composed of
portions of maternal and paternal threads lying parallel to
each other. The ends of these arms are the synaptic ends of
the chromatids, but whether the synaptic end of a chromatid
represents the polar end, as is certainly the case in the
Orthoptera,* it is impossible to determine. One of the other
pair of opposite arms is composed of parallel maternal threads
and the remaining arm consists of parallel paternal threads;

* Pinney, Edith, Organization of the Chromosomes in Phrynotettix magnus, Kansas
University Science Bulletin, vol. IV, No. 14, 1908.

“

PINNEY: CHROMOSOMES OF ANASA TRISTIS. 359

(a) shows a rod form in which the division at right angles to
the long axis may or may not separate elements of opposite
parental origin. It is impossible to ascertain whether these rod
forms have passed through the cross stage, since some of the
earliest recognizable tetrads have this form and keep it as they
condense.

The manner in which these forms are evolved from two con-
jugated chromosomes of opposite sex, each split longitudinally,
is similar to that described for the Orthoptera by McClung ;*
(d) is an unusual condition at this stage of the smallest sper-
matogonial chromosomes. Even in stages as late as those
shown in figures 22 and 24, plate LXVI, the separate halves
of this pair often lie on opposite sides of the nucleus. Just
previous to the metaphase they approach end to end. The
constriction between the two halves is never obliterated by a
close union. Evidently synapsis has had no effect on the
smallest two members of the spermatogonial complex, al-
though they undoubtedly entered into the synaptic knot. Sub-
sequent changes consist in the condensation of these filiform
tetrads to form the homogeneous elements of the metaphase.
As the chromosomes approximate the homogeneous condition
they arrange themselves at the periphery of the nucleus. This
tendency may account for the characteristic position of the
accessory during synizesis, the growth period and the sper-
matocyte prophases. The same nuclear membrane which is
formed previous to synizesis continues to function up to the
time of a late spermatocyte prophase. At the latter stage
there are present within the nucleus thirteen distinct bodies.
Figures 10 and 11, plate LX VI, show two parts of the same
nucleus appearing in adjacent sections. The thirteen bodies
are drawn. These must represent the twenty-one sperma-
_ togonial chromosomes plus the plasmasome created during
synizesis. |

Before stating the homologies between these spermatocyte
structures and those of the previous generation the history of
the accessory will be followed, beginning with synapsis. In
this connection the behavior of the new spermatocyte element,
the plasmasome, will be reported. The time of appearance, the |
nature and source of these two bodies, have already been spoken
of. It has also been mentioned that immediately after synizesis
the two bodies are practically equal in size. In the succeeding

* McClung, C. E., ’00, The Spermatocyte Divisions of the Acridide, Kansas University
Science Bulletin, IX, L 02 ; The Spermatocyte Divisions of the Locustidx, ibid., XI, 8.

360 KANSAS UNIVERSITY SCIENCE BULLETIN.

diffusion stages the plasmasome increases in size and the dis-
tinction between it and the accessory becomes very marked.
In these stages the accessory remains closely appressed to the

nuclear wall while the plasmasome is located internally. See -

figures 3, 5 and 6, plate LXVI. This period during the dif-
fusion of chromatin and the growth of the plasmasome can
not be considered as belonging to the prophase of the first sper-
matocyte. The nucleus increases steadily in size, reaching its
maximum when diffusion is complete, so we may consistently
call this interval after synizesis the growth period. Its cul-
mination marks the installation of important changes, one of
which, the re-formation of the chromosomes, has already been
described. The accessory meanwhile maintains its former
position, size and staining qualities. During the prophase the
plasmasome gradually diminishes in size, and finally disappears
- just before the metaphase. During this time, too, it does not
stain so readily with chromatin stains. In the Heidenhain’s
iron-hematoxylin preparations the staining quality of the
plasmasome varied with the amount of extraction in the iron-
alum solution. Occasionally the plasmasome presents a vacuo-
lated appearance. This may occur at any time during its
history but is more characteristic of the beginning of its dis-

solution. Occasionally in the prophase there occur two small

homogeneous bodies staining like the plasmasome. As the
typical plasmasome is absent and the accessory present these
are considered as fragments of the plasmasome. There is no
regularity in their occurrence. The cysts in this condition are
not frequent and the majority of their cells contain typical
plasmasomes.

In later stages the plasmasome is identified by its shape, be-
coming spherical as it decreases in size. Figure 7, plate LX VI,
shows a nucleus containing the plasmasome and the accessory.
The difference in staining qualities is apparent. The accessory
remains in its former position, its form to the observer depend-
ing on his point of view. Plate LXVII consists of a series of
drawings showing the size relations of the accessory and the
plasmasome at the time of synizesis, the growth period and the
succeeding prophases.-: The drawing of nuclei at the top of the
page show the condition of the active chromatin. They are
numbered in the order of the changes which they present.
For instance, figure 1 shows a nucleus during synizesis. The
perpendicular column below figure 1, designated la, shows a

a

PINNEY: CHROMOSOMES OF ANASA TRISTIS. 361

number of drawings of the accessory taken from cells in the
same stage as figure 1. Figure 2 shows a nucleus of the stage
immediately following synizesis. The perpendicular column,
2a, below figure 2 shows a number of drawings of the accessory
taken from nuclei in the same stage as figure 2. The perpen-
dicular column, 2b, beneath figure 2 shows a number of draw-
ings of the plasmasome taken from cells in the same stage as
that shown in figure 2. In the same way columns 3a and 3b,
4a and 4), etc., are drawings of the accessory and plasmasome
from nuclei such as are shown in figures 3, 4, etc., respectively.
The horizontal columns A, B, C and D indicate the different in-
dividuals from which the drawings were made. The differ-
ences in size which are shown in these monoplane drawings
depend upon the point of view of the observer. That the ac-
cessory is constant in size is easily determined by different
focusings. From these drawings there is also found to be a
regularity in the relation between the size of the plasmasome
and the condition of the active chromatin of the nucleus. From
the foregoing observations it is now clear that the thirteen
nuclear structures present in the late spermatocyte prophase
comprise one accessory, one diminished plasmasome, two small
chromosomal elements of opposite parental origin, and nine
ordinary tetrads.
_ Figures 1 to 10, inclusive, plate LXVIII, show typical polar
views of the first spermatocyte metaphase. All of the groups
shown are from different individuals, with the exception of
figures 2, 3, 4, plate LX VIII, which were in the same section of
the same cyst. As many as six of these views, all typical, have
been observed in one section in one cyst. There are eleven
chromosomes present, each of which, with the exception of the
accessory, is composed of four chromatids. The plasmasome
has disappeared, also the nuclear membrane. The chromo-
somes lie in a clear area surrounded by the cytoplasm. Nine
of the complex, including the largest, form a rather compact
_ ring, in the center of which appears the smallest member of
the group. The relative positions of the chromosomes forming
the ring vary. One chromosome lies outside of the ring. Fig-
ure 11, plate LX VIII, shows six of these outer chromosomes at
this stage taken from the same individual from which figure 6
was made. So, also, figures 12, 13, 14 and 15 are drawings of
the similarly located chromosome found in the individuals from
which figures 7, 8, 9 and 10, respectively, were made. This

362 KANSAS UNIVERSITY SCIENCE BULLETIN.

eccentric member shows remarkable constancy in size. Its
size and its position suggest the accessory during synapsis.
See figures 29, 30, 33, plate LXV. Figure 5, plate LXVIII,
shows a polar view of an atypical metaphase. This condition
was observed only twice in this material, both times occurring
in different individuals. There appear to be twelve chromo-
somes present. In addition to the small chromosome within
the ring, there are two large equal elements which are not
large enough to correspond to any elements present in the
typical metaphase. Only eight chromosomes form the ring. It
must be that the missing one of the nine which usually form
the ring is represented by the two small and unusual bodies
which lie within the ring, one on either side of the small ele-
ment, and that this is merely an accidental case in which
synapsis did not occur. In connection with this supposition we
recall the fact that synapsis did not occur in the small chromo-
some at the usual time. This similarity of action during the
synaptic period may be significantly related to their similarity
of position. 3

In a lateral view of a spermatocyte anaphase the same sur-
face of the chromosomal elements is presented to the observer
as is seen in the equatorial plate of the spermatogonial cells.
By referring to figures 1 to 10, plate LXV, it will be seen that
the chromatids have become slightly shorter and correspond-
ingly thicker. Here again the smallest chromosome asserts its
independence. In the spermatogonial cells these small chromo-
somes present spherical outlines. In the first spermatocyte
generation, instead of shortening, as their fellows, they have
elongated. Only in this element and in the accessory was the
plane of division actually determined. Since the two halves
of the smallest chromosome are never so closely united that we
lose sight of their identity, it can be stated positively that
division in this case is transverse and qualitative, causing the
separation of maternal and paternal elements. Owing prob-
ably to the slight attraction exhibited between the halves of
this chromosome division is effected while the other chromo-
somes are still intact. Lateral views of early anaphases in- —
variably show this condition. Similar views of later stages
show its separated halves preceding the larger diads to the
poles. In the ordinary tetrads the formation of the crosses
might be interpreted to have the same significance as is at-
tributed to the like form in the Orthoptera, in which McClung

PINNEY: CHROMOSOMES OF ANASA TRISTIS. 363

has shown conclusively that the first division plane is longi-
tudinal. However, in this case there is no positive evidence to
offer. The chromosomes enter the equatorial plate with their
longitudinal axes parallel to the spindle fibers. Figures 16 and
17, plate LX. VIII, show that the polar ends of the chromosomes
converge toward the poles of the spindle. That both methods
of division may be employed by different members of the group
is shown by a study of the accessory at this stage. This ele-
ment, which is composed of two chromatids, was last observed
in the metaphase lying outside of the ring. The element (2)
in figure 16, plate LXVIII, is undoubtedly the accessory chro-
mosome. This drawing is not limited to one plane. The small
chromosome is shown dividing in its usual manner. This ele-
ment, it will be remembered, is located at the center of the
ring and no element of the ring could lie so far from this center
as the one marked (x). This, then, must lie on the outside of
the circle. From the arrangement found in polar views of the
metaphase one would expect in the lateral views to find sec-
tions which would show four chromosomes in the same plane.
These would include the small central member, two chromo-
somes of the ring, one on either side, and, at one side, the
accessory. Such a section is shown in figure 17, plate LXVIII.
- The small chromosome, it will be seen, occupies the axis of
the spindle, which in this view appears asymmetrical, the
asymmetry being due to the extra divergence of the fibers at-
tached to the accessory.

In these lateral views of anaphases the structure of the ac-
cessory is easily determined. It is composed of two chroma-
tids, which are shown in figures 19, 20 and 21, plate LXVIII.
All of the other chromosomes contain four chromatids,
although in some only two of the four appear in the same
plane. The accessory is easily identified by its form in these
stages, where it is shown to divide tardily, division not be-
ginning until the ordinary chromosomes have entirely divided.
In their journey to the poles the accessory chromatids lag be-
hind the others. Figures 17 to 26, plate LXVIII, show this
behavior at different stages. All of the chromosomes of the
complex keep their relative positions until the pole is reached.
Figures 1, 2, 3 and 9, plate LXIX, show polar views of the
dividing groups with the inner and outer members in their
characteristic positions. As soon as the pole is reached the

2—Univ. Sci. Bull., Vol. V, No. 20.

364 KANSAS UNIVERSITY SCIENCE BULLETIN.

chromosomes crowd together and, on account of the lagging of
the accessory, lateral views of this stage show the condition re-
produced in figures 25 and 26, plate LXVIII; (2) is the ac-
cessory. It is easily identified in both daughter groups. Fig-
ure 24 shows the only case observed which might be inter-
preted as the failure of the accessory to divide. The spherical
form of the structure (7%) may indicate a degenerate form of
the accessory or an unusually persistent plasmasome. The
crowding together of the diads at the poles completes the cycle
of changes occurring during the first spermatocyte generation.
No telophase involving the disintegration of the chromosomes
and the appearance of a new nuclear membrane occurs.

The Second Spermatocyte.

The second spermatocyte generation begins with the forma-
tion of an equatorial plate by the members of the new sperma-
tocyte nucleus. These second spermatocyte chromosomes con-
sist of ten diads and one monad or single chromatid, the acces-
sory. In figures 1, 2, plate LXIX, of typical equatorial plates
of this generation will be seen a tendency toward the same
chromosome arrangement which prevailed during the preced-
ing mitotic cycle. From the size relations shown in figure 4,
plate LXIX, which is a group of accessory chromosomes from
cells of the same stage and animal as figure 3, plate LXIX; re-
spectively, we are justified in our conclusions as to the identity
of this element. Figures 5, 6 and 7, plate LXIX, are polar
views of beginning anaphases. The chromosomes are viewed
here in cross-section. Figures 10, 11 and 12, plate LXIX, show
lateral views of the beginning of this ultimate division. Figure
10 suggests what transpired between the stage shown in figures
1, 2 and 3 and that shown in figures 5, 6 and 7. The transi-
tion is extremely brief, for evidence of it is comparatively in-
frequent. The chromatids which, during the first spermatocyte
division, lay parallel to each other have separated at one ex-
tremity and have swung around until they lie end to end, the
separate ends pointing toward opposite poles. The chromatids
maintain this position during a relatively long period of
time, for this condition and the following telophase are most
commonly noted in cysts of second spermatocytes. A typical
spindle is formed and the fibers are attached at the separated
ends of the chromatids. Separation of chromatids occurs
simultaneously, the smallest member here acting in unison

PINNEY: CHROMOSOMES OF ANASA TRISTIS. 365

with the others. This may be regarded as due to the fact that
the same conditions are present in all of the chromosomes pre-
vious to division; that is, the degree of union between the chro-
matids is the same in each diad, while it will be remembered
that this relation between the diads which composed the tetrads
‘of the first spermatocyte varied.

When the chromosomes separate sufficiently to expose a pias
zone of fibers in the region occupied by the equatorial plate
we find the accessory chromatid still in its place, and it keeps
this position until the remaining chromatids have almost
reached the opposite poles. Figures 13 to 18, inclusive, plate
LXIX, show the successive steps in this division. Well coa-
lesced daughter groups of this division are observed at oppo-
site ends of the spindle, while the accessory lingers half way be-
tween. The accessory finally approaches one of the groups and
almost immediately the dividing cell wall is formed. This
leaves the accessory near the newly formed cell wall, entirely
apart from the group to which it evidently belongs. Figure 19,
plate LXIX, is a group of such bodies drawn from cells in this
stage. The size of this lagging body was thus found to be
constant. The spindle fibers between the two groups persist
- until after the nuclear membranes of the daughter cells are
formed. Before this occurs, however, the changes shown in fig-
ures 20 to 25, inclusive, plate LXIX, take place. This consists
in the withdrawal of the accessory from its position near the
dividing wall into the now shapeless chromatin mass of the new
nucleus. In studying these processes it was very easy to obtain
sections which showed the two daughter cells in the same plane.
The connecting band of spindle fibers leaves no doubt as to
their relation even after the new cell wall has been formed.
The accessory occurs in only one of these cells. A further
significant fact is that in every such case where both daughter
cells can be identified beyond doubt one always shows the ac-
cessory at some stage of its characteristic activity, while in the
other no corresponding body is to be found.

As soon as the membrane forms around the spermatid nu-
cleus which contains the accessory this element becomes de-
tached from the mass of which it forms a part and moves again
to the nuclear wall. Its individuality again becomes prominent.
Figures 26 to 31, plate LXIX, show this condition. Here also
the same precautions in determining its identity were observed.

866 KANSAS UNIVERSITY SCIENCE. BULLETIN.

Care was taken to study both daughter groups, in one of which
it invariably occurred while the other offered no evidence of
the presence of a similar element. The constancy in size of
the accessory and the relation of its size here to the size of the
accessory as determined by -figures 4 and 19, plate LXIX,
help to establish its identity. This condition of the chromatin .
in the spermatid nucleus lasts for only a short time. The
large chromatin mass soon begins to break up into smaller
masses which eventually distribute themselves against the
nuclear wall. In an early stage of this disintegration the ac-
cessory is still easily recognized, as it lies on one side of the
nucleus apart from the rest of the chromatin. It still keeps
the elongated form it displayed im the second spermatocyte
anaphase.

A whole cyst of young spermatid nuclei was counted in order
to determine in just what proportion of the spermatids of one
cyst the accessory occurred. The importance of this observation
is obvious, although it seems that the result could be predicted
safely from the evidence obtained from previous observations.
An account of the method used in counting may help to verify
the results. Serial sections of an entire cyst were obtained.
The cyst chosen for counting ran through twenty-four sections.
The sections were five micra in thickness. The exact condition
of the chromatin was found to vary in almost every cyst of this
generation. In the cyst counted were found the. stages shown
in figures 28, 29 and 31, plate LXIX. From the conditions ex-
isting in such nuclei we would expect to be able to identify the
accessory from its shape and its position with respect to the
rest of the chromatin in the nucleus. Evidently the number of
nuclei in such a cyst and the number of spermatids formed
would be equal. All whole nuclei or almost whole nuclei and
fragments containing the accessory were counted. It was im-
possible to avoid the error of counting one nucleus as two, where
a nucleus happened to be cut exactly in half; one half appearing
in two adjacent sections, The number of nuclei containing the
accessory was 484. The number of fragments of nuclei which
contained the accessory was 40. Thus the entire number of
spermatids, in which the accessory occurred was 524. The
number of nuclei counted which contained no accessory was
577. These, however, would include the nuclei from which
small fragments containing the accessory had been severed.

M’CLUNG: CHROMOSOMES OF ANASA TRISTIS. 337

That would make the number of nuclei with no accessory 537,
which is slightly larger than the number with the accessory. It
has been pointed out that the nuclei which were cut in two
équal parts might be counted as two: This would add to the
number of cells which did not contain an accessory. Consider-
ing this source of error the number of cells of each kind may
practically be considered as equal. There can be no doubt
as to the occurrence of two kinds of spermatids which develop
_ into spermatozoa, one containing one more chromatid, the ac-
cessory, than the other; and from the evidence presentéd we
feel justified in concluding that the two kinds of spermatozoa
occur in equal numbers.

lil. A COMPARISON OF THE RESULTS OF STUDIES UPON
CHROMOSOMES OF ANASA.

BY C. E. M’CLUNG.
1. Methods.

The methods ét an investigator are a very important factor
in arriving at conclusions, and it is quite possible to secure
diametrically opposite results upon the same material by
varying the technical processes. It will, therefore, be neces-
sary to consider the methods of preparation and study em-
ployed upon Anasa. Paulmier, Montgomery, Wilson, Lefevre,
McGill and Pinney fixed their material with Flemming’s fluid,
corrosive-acetic mixtures, Bouin’s fluid and various other
cytological reagents. This material was then sectioned in
paraffin, and stained, principally with the iron alum—hema-
toxylin stain of Heidenhain. The slides thus prepared by
- Paulmier, Wilson, Lefevre and Pinney are beautiful examples
of cytological technic and leave little to be desired in delicacy
and precision. A considerable number of the slides sent me
by Professor Wilson and Professor Lefevre are prepared ac-
cording to the methods of Foot and Strobell, so that I have
been able to study the material under their conditions, al-
though, unfortunately, their own preparations have not been
accessible to me. Since the entire case of these investigators,
as opposed to all their fellow workers, rests upon their own
peculiar methods, it seems desirable to consider in detail the
technie upon which they rely.

The usual methods of cytologists are dismissed with prac-
tically no consideration, and instead of fixing the material, it

3868 KANSAS UNIVERSITY SCIENCE BULLETIN.

is spread upon glass slips and allowed to dry. (Foot and
Strobell, ’07, p. 282.) In this condition it is stained with
Bismark brown and mounted in balsam. Entire reliance is
placed in this method, and little attempt is made to correlate
the results thus obtained with the appearances presented by
other systems of technic. That the position of these authors
on this point may be clearly understood, I quote their state-
ment of the case.

“Professor Wilson, in his recent paper in Science, February,
07, replying to our preliminary note, says that he thinks the
contradiction in our results is probably due to the difference
of method employed, we having placed our faith in smear
preparations, while he has relied on sections. We are glad
of an opportunity to emphasize this faith, believing that for
demonstration of the structure and count of chromosomes our
modified smear preparations are more reliable than sections;
and it is for this reason we have abandoned the use of sections
in studying chromosomes, except for comparative work and
for studying the topographical relation of the cells. In cells
fixed and sectioned nearly all the delicate details shown in
the chromosomes of our smear preparations are completely
lost, and it ought to be too obvious to mention that a method
which presents clearly each individual chromosome in its in-
tegrity offers decided advantages when the question of accu-
rate counting assumes the importance and develops the con-
tradictions familiar in recent literature.”

The authors state that they have used sections for com-
parison and topographic work, but none of their photographs
are made from sections, and their entire argument, as pub-
lished, is based upon material prepared according to their
own method.

It is my judgment that this method, used alone, is entirely
inadequate for accurate results, and in this particular case is
responsible for the discrepancy between these investigators
and Paulmier, Wilson and Montgomery. From the work of
Miss Pinney it will be clear, I think, that in the prophases
figured by Foot and Strobell there are two darkly staining
bodies, the accessory chromosome and the plasmasome, instead
of the one shown in their photomicrographs. Owing to the
technic employed by them the plasmasome is. practically de-
stroyed, and instead of using other methods to determine the

M’CLUNG: CHROMOSOMES OF ANASA TRISTIS. 369

nature of the nuclear body, they assume that it is a plasma-
some. “As the presence of a plasmasome in the nucleus at
these stages is the typical phenomenon familiar in all known
forms (its absence being most exceptional), we feel justified
in interpreting the structure we find in the resting nucleus as
a true plasmasome, and not an odd persisting spermatogonial
chromosome, i. e., chromosome nucleolus.”’ (Foot and Strobell,
07, p. 284.) Further, the smear method does not give precise
and accurate pictures of the chromosomes. The variation in
size, owing to the different thicknesses of the smear and amount
of spreading, is very extensive, as may be seen by a com-
parison of the photomicrographs of Foot and Strobell. Com-
pare, also, figures 21-26, plate LXX, which represent smears
of cells in the same stage. Not only may the entire groups of
chromosomes in similar cells vary as three or four to one, but
the members of one cell may suffer unlike expansion or con-
traction to a similar extent, if the conditions of drying are
variable, due to the thickness of the smear film. There is but
one advantage of the smear method over sections, and that is
that it presents all the elements of the cell in one plane, so that
their entire outline is visible for studying and photographing.
It is a method that should be used in connection with sections,
and from the beginning of my work I have utilized it con-
sistently and with great profit in this way. I therefore speak
of it as a warm friend and advocate, but at the same time I
realize its limitations and am convinced that it should not be
used to the exclusion of all other methods.

Foot and Strobell have employed photography alone as a
means of presenting illustrations of their material, and it is
assumed by them that if a thing can be photographed it must
necessarily be a true picture of normal conditions. This I
consider to be a decided fallacy. A photograph is an inter-
pretation by the observer, just as is a drawing. The personal
factor is no more absent from one method of illustration than
it is from the other. Photographs may present with greater
fidelity the details of structure in an object, but the choice of
the object and the nature of details are at the command of
the photographer. It is possible, especially in smear prepara-
tions, to select cells that will illustrate almost any condition
and to photograph them. The highly interpretative character
of the photograph is illustrated admirably by the present con-

370 KANSAS UNIVERSITY SCIENCE BULLETIN.

troversy over the chromosomes of Anasa. Foot and Strobell
publish photographs that seem to show that the accessory
chromosome divides in the second spermatocyte division. This
is their interpretation of the behavior of this element, and
they illustrate it by photography. Not a picture which they
publish would indicate that there is ever a case where the ac-
cessory chromosome positively fails to divide at this time,
and yet literally thousands- of unquestionable cases of this
occur in every good preparation. See the group of cells
shown in figure 15, plate LX XI. By selecting a series of early
anaphases, and a very few obscure and distorted individual
cells, they are able to secure pictures which appear to bear
out their contention. In speaking thus of the methods of
Misses Foot and Strobell, I wish it clearly to be understood
that I am not reflecting upon their motives. It is my desire
only to show that photographs are interpretations of observed
phenomena and have no claim to infallibility. Again, as in
the case of the smear method, I would state that I speak as a
friend of the method and not as one who deprecates it. I
believe that scientific papers, particularly cytological, should,
whenever possible, be illustrated by photographs, and in my
own publications I have followed this plan so far as the
circumstances would permit.

2. What is the number of spermatogonial chromosomes and
, of odgonial chromosomes?

Confessedly, the accurate estimation of the numbers of chro-
mosomes in these generations of cells is a difficult matter, and
if it were the only means of determining the chromosomal
differences of the sexes there might be occasion for doubt with
regard to theories founded upon such enumerations alone.
Fortunately, it is only one of several criteria of sexual differ-
entiation in the germ cells, and so is not of commanding
importance. Accurate counts of the chromosomes, however,
are entirely possible with care, and concordant results are ob-
tainable by independent observers. In the case of Anasa this
is quite possible, and both Miss Pinney and myself have found
the number twenty-one uniformly present in the sperma-
togonia. Not a single instance of twenty-two chromosomes
was found in any spermatogonium, and we must therefore con-
clude with Wilson that thirty-one is the normal spermatogonial

M’CLUNG: CHROMOSOMES OF ANASA TRISTIS. 871

number. In plate LXX, figures 1-3, may be seen sperma-
togonial complexes showing this number clearly.

It is claimed by Foot and Strobell that the number is twenty-
two, and they publish four photographs in support of this
view. Only one of these represents a stage where the chromo-
somes are compact and homogeneous, and this shows twenty-
one chromosomes, although in the explanation of the figure
the authors state that one of the apparent chromosomes is
really two joined at right angles. Since there is no case in the
figure where one chromosome is bent at almost right angles,
the explanation is not very convincing. The remaining three
are of earlier stages, in which the chromosomes have a loose
structure that might readily be disturbed in the process of
smearing, thus separating one of the larger chromosomes. In
two of the photographs such an interpretation would seem to be
entirely justified. Photo 49 is the only one in which there
would seem to be any certainty with regard to the number
twenty-two. Here it appears as if conditions were normal,
and it might be recorded as an instance of twenty-two chro-
mosomes in the spermatogonia.

The conditions regarding the spermatogonial number may be
summarized thus: Paulmier, Montgomery at the time agreeing
with him, reported twenty-two spermatogonial chromosomes.
Later, Wilson, working upon Paulmier’s original preparations
in part, finds twenty-one spermatogonial chromosomes, and
Montgomery, restudying his own material, agrees with this
enumeration. Lefevre and McGill report twenty-one sperma-
togonial chromosomes from their own preparations. Foot and
Strobell, working upon smear preparations, consider twenty-
two to be the normal spermatogonial number. Miss Pinney,
upoh examining material from Kansas, Massachusetts and
Pennsylvania, and working without a knowledge of others’ re-
sults,. finds twenty-one to be the spermatogonial number.
Finally, from a study of the original material of Paulmier,
Wilson, Lefevre and McGill, and Pinney, I myself find in all
clear cases, where there is no doubt of the presence of all the
chromosomes, that the number is twenty-one. Foot and Stro-
bell object to the testimony of Montgomery, and of Lefevre and
McGill, on the ground that they reported at one time an even
number of spermatogonial chromosomes and at another an odd
number. It must of course be admitted that this weakens their

372 KANSAS UNIVERSITY SCIENCE BULLETIN.

testimony, for it is an evidence either of careless work or of
the influence of preconceptions. I can recall with what diffi-
culty I persuaded myself, against the current conceptions of
chromosome numbers, that the spermatogonial number of
chromosomes is odd in the Orthoptera, and I can easily under-
stand the temptation, in an early stage of an investigation, to
seek concordance with accepted opinions. The investigators
who reported even numbers for their material did so when all
the evidence from other forms seemed to demonstrate the uni- —
versality of this phenomenon. In a sense their testimony is
more valuable after a restudy of their material, because it in-
dicates the desire to arrive at facts regardless of consequences.
It is a very unusual man, or a very stubborn one, who finds no
occasion to change his opinions.

This is the present status of the question concerning the
spermatogonial number of chromosomes in Anasa. If it is im-
possible for the interested cytologist to come to a conclusion re-
garding the facts of the case from the studies so far made, he is
quite at liberty to undertake an investigation for. himself and
I have no doubt but that all the material that has been ayail-
able for my investigation will be placed at his disposal. -

There appears to be no dispute regarding the odgonial num-
_ ber, although the counts in these cells have been less numerous
than in those of the male. So far as the opportunity has
offered I have gone. over the female complex of chromosomes,
and I have no reason to doubt that the number is twenty-two,
as determined by Wilson.

3. What is the number of first spermatocyte chromosomes?

Regarding the number of first spermatocyte chromosomes,
there seems to be no difference of opinion, since every observer
has recorded eleven as typical for this generation. The clear-
ness with which the chromosomes of the first maturation mito-
sis appear in the equatorial plate almost precludes the possi-
bility of error. Here also may be found a close adherence to a
type of arrangement which presents the small chromosome in
the center, surrounded by a more or less regular ring of nine
chromosomes upon the outside of which is placed the accessory
chromosome. This is not an invariable arrangement, for it is
a culmination of the movements of the chromosomes in the pro-
phase where the elements bear somewhat similar relations to
each other, but it appears in a large proportion of the cells.

M’CLUNG: CHROMOSOMES OF ANASA TRISTIS. 373

The position of the accessory chromosome upon the outside of
the group has led Foot and Strobell to coin another new name
for it, and in their terminology it is the “eccentric chromo-
some.”

4. What is the behavior of the accessory chromosome and of
the plasmasome in the first spermatocyte prophase?

While there is no dispute regarding the number of chromo-
somes in the first spermatocyte, there is serious divergence of
opinion with reference to the structure of the accessory chro-
mosome. Upon its nature depends the number of spermato-
gonial chromosomes and the character of the four spermatids
derived from each first spermatocyte, so its determination is
most important. If it be a tetrad then the spermatogonial
number must be even and the spermatids all alike in the pos-
session of eleven chromatids; but if, on the contrary, it be a
diad the spermatogonial number of chromosomes is necessarily
odd and the spermatids of two types, one of which possesses
eleven chromosomes and the other ten.

With the exception of Foot and Strobell all observers are
agreed that there are but two chromatids in the accessory
chromosome, but these observers claim that it possesses four
as do the other chromosomes. As proof for this contention,
they publish a series of chromosome groups from the late pro-
phase in which the various elements are clearly shown. The
identification of the accessory chromosome is correctly made in
each case, I believe, but it is entirely clear to my mind that
there is nothing in the appearance of these photographs to
justify the interpretation of the accessory chromosome as a
tetrad. There is in each case a characteristic difference be-
tween the -tetrads and the accessory chromosome, for in the
former there always appears a diamond-shaped opening, the
angles of which point to the two lines of cleavage, while in the
case of the accessory chromosome this is lacking. Every ac-
cessory chromosome, on the contrary, shows a clear, straight
line of division along its entire length. In but a single instance
is there an exception, and that is in their. figure 4, plate II,
where the plane of longitudinal cleavage is interrupted by

transverse markings. A careful examination of this chromo-

some will demonstrate, however, that there is not only one of
these apparent cross divisions, but two of them. I am con-
vinced that the excellent photomicrographs of Foot and Stro-

374 KANSAS UNIVERSITY SCIENCE BULLETIN.

bell of this stage are sufficiently clear in their evidence of the
nature of the accessory chromosome, and I have not included
any number of these stages in the photomicrographs accom-
panying this paper. Moreover, in none of the material that I
have examined is there any indication whatever of a tetrad
structure in the accessory chromosome, and I am thoroughly
convinced that it is a univalent chromosome.

If the earlier prophase of the first spermatocyte be exam-
ined, the chief error of Foot and Strobell becomes evident. It
is claimed by these investigators that the accessory chromo-
some behaves during the prophase as do the other chromo- —
somes and that there is but a single nucleolus-like body present
—a plasmasome. The work of Miss Pinney showing the tran-
sition in form and size of the accessory chromosome and of the
plasmasome, both of which appear simultaneously in the nu-
cleus, makes it entirely clear that Foot and Strobell were mis-
taken on this point. As I have elsewhere pointed out, the
technic of these investigators is responsible for their failure
to find the plasmasome, for it is this structure, and not the
accessry chromosome, that is missing from the smear prepa-
rations.

These investigators have pointed out the large difference
between the appearance of the accessory chromosome in draw-
ings made by Wilson and the photographs illustrating their
own paper. Such variations in the form of this element will
receive explanation in part from the series of drawings made
by Miss Pinney to illustrate the transition in form of the pro-
phase elements. Both the stage of development and the method
of preparation influence the form of this accessory chromo-
some. As a rule, sections show this element as a.spheroidal,
apparently homogeneous body during most of the prophase,
but occasionally, just succeeding synizesis, it appears as a short,
thick thread, with a tendency to be bent at the middle.. This
is the nearest approach to a spireme condition that it attains,
and in this corresponds to similar conditions which I have de-
scribed for certain Locustids. So far as my experience goes it
would seem to be the rule that the accessory chromosome, to a
certain degree, undergoes a granular diffusion of the chromatin
similar to that of the other chromosomes, but that it is less ex-
tensive and is frequently marked by the close approximation of
divisions of the thread.

It will be observed from plate LXX, figures 4-6 and 11, that

i ae es

a

M’CLUNG: CHROMOSOMES OF ANASA TRISTIS. 375

the accessory chromosome in this spireme condition is accom-
panied by a large distinct plasmasome which lies toward the
center of the nucleus. Such a structure is absent from the
photographs presented by Foot and Strobell when the accessory
chromosome (their plasmasome) is elongated. As will be noted
in Miss Pinney’s description, the plasmasome does not appear
until after synizesis is established. Some of the nuclei figured
by Foot and Strobell would seem to be presynizetic in develop-
ment and could not therefore contain plasmasomes.

I would conclude, therefore, that after synizesis there are
present in the first spermatocyte nuclei two nucleolus-like
bodies, one the accessory chromosome and the other a plasma--
some. Further, that the accessory chromosome exists through-
out the whole period, for a time, as a short, heavy thread, at
other times concentrated into a mass, and finally in the late
prophases as a straight longitudinally split rod. The plasma-
some, on the other hand, is an inconstant structure, absent be-
fore synizesis, and again in the late prophase, and is not to be
mistaken for the accessory chromosome.

5. What is the behavior of the accessory chromosome in the
first spermatocyte mitosis?

Reference has already been made to the condition of the ac-
cessory chromosome in the late prophase and to its position in
the equatorial plate. It appears in the metaphase as a uni-
valent, longitudinally split rod lying outside the ring of ordi-
nary chromosomes. In division its halves separate and during
the anaphase move to the two poles of the spindle as do those
of the other chromosomes. Examples of this may be seen in
figures 2, 3, 4, 5, 6, 7, 8, 9, 10, plate LXXI. Both series of
daughter chromosomes may be seen in plate LXXI, figure 3, .
which shows a mid-anaphase with the halves of the accessory
chomosome accompanying each group. There appears to be no
question on the part of anyone regarding the movements of the
chromosomes in the first spermatocyte, so that an extended dis-
‘cussion on this point is not called for.

From the fact, however, that Foot and Strobell have used the
tardy division of the accessory chromosome in the first sperma-
tocyte as an argument to prove its division in the second
spermatocyte, because here also it lingers for some time near
the equatorial plate after the other chromosomes have moved
toward the poles of the spindle, it will be necessary to call at-

376 KANSAS UNIVERSITY SCIENCE BULLETIN.

tention to figures 6, 7, 8, 9, 10, plate LXXI. These show that
while the time of division is later than that of the other chro-
mosomes the actual separation occurs and the halves of the
divided accessory chromosome may, in each case, be seen in the
daughter cells. No instance could be observed where there
could be doubt of this fact. The clearness of these phenomena
is due to the fact that the accessory chromosome remains, as in
the prophase, distinct from the other chromosomes, so that it
can be followed even into the telophase. As will be seen later,
the same aloofness on the part of the accessory chromosome
is encountered in the second spermatocyte and is so marked
that there is no difficulty in tracing its behavior.

6. What is the behavior of the chromosomes in interkinesis ?

The interval between the two spermatocyte divisions is com-
paratively brief and the changes in the relative positions of
the chromosomes slight. While the chromosomes mass together,
the outlines of the individual elements may be seen clearly in
favorable preparations of the telophase. Since the whole
daughter chromosome complex travels to the pole at about the
same rate, its members at the end of the movement lie at
approximately the same level. When therefore the second
spermatocyte spindle is formed there is little or no change in the
positions of the chromosomes, and the equatorial plate of the
metaphase corresponds in general with that of the first sperma-
tocyte, from which it was derived. While this is true, the slight
movements of all the chromosomes during their poleward mi-
gration usually results in a destruction of the typical ring
arrangement so characteristic of the first spermatocyte. Be-
cause of its isolated position and somewhat greater variation
* of movement in the anaphases, the accessory shows more
marked differences in position in the equatorial plate than the
other chromosomes. If it has moved toward the pole at the
same rate as the rest of the complex and maintains its position
during the telophase it will then be found upon the periphery
of the equatorial plate in the second spermatocyte. If, on the
contrary, as often happens, the movement is slightly slower
than that of the other chromosomes and it swings around un-
der the mass during the anaphase movements it will later ap-
pear within the group of chromosomes in the resulting equa-
torial plate. It is not possible, therefore, to recognize the ac-

M’CLUNG: CHROMOSOMES OF ANASA TRISTIS. 377

cessory in the second spermatocyte by its position, as might
frequently be done in the first spermatocyte. I am of the
opinion accordingly that any identification based upon position
alone is not warranted in the case of the accessory chromosome
in the second spermatocyte. By this I do not mean that there
may not be similar arrangements of the chromosomes in both
spermatocyte metaphases, but rather that the opportunities for
movements of the chromosomes during the first spermatocyte
anaphase and telophase are so great that no rule may be laid
down to mark a type of second spermatocyte. (Plate LXXI,
figures 11, 12.)

7. What is the behavior of the accessory chromosome in the
second spermatocyte mitosis?

At this point in the history of the accessory chromosome of
Anasa there occurs the wide diversity of opinion between Foot
and Strobell and the other investigators who have studied it.
The issue is clearly drawn. Foot and Strobell claim that the
accessory chromosome is divided in the second spermatocyte
as in the first spermatocyte, but Wilson and others are just
as positive that it passes undivided into but one of the pair
of spermatids formed by the division of each second sperma-
tocyte. Issues are joined here on a question of fact, and it be-
comes necessary to determine, therefore, whether the accessory
chromosome divides in the second spermatocyte metaphase or
whether it passes undivided into one of the two daughter cells.

I must confess that it is a matter of no little astonishment
to me that there should be any difference of opinion on this
point, because the conditions are so clear and unmistakable as
almost to preclude error. A number of photomicrographs of
the second spermatocyte anaphase have been prepared from
both sections and smears and are reproduced in plate LX XI, fig-
ures 13-27. In figure 15 there are shown ten pairs of spermatids
as they appeared spread upon the glass slip. This slide was
prepared according to the method of Foot and Strobell and
shows the cells practically separated so that there can be no
question of a later division of the accessory chromosome. In
each one of these ten pairs it is clearly seen that one member
contains an accessory chromosome while its mate lacks it. The
same statement holds true of all similar cells drawn and photo-
graphed, and of the thousands of others examined. In no case

378 KANSAS UNIVERSITY SCIENCE BULLETIN.

was a divided accessory chromosome found in any second
spermatocyte. For a short time the accessory chromosome
lies against the mass of chromosomes, but when the spermatids
are separated and the nuclei re-form it moves away again and
lies at one side. Such conditions are shown in plate LXXI, fig-
ures 20-21, 25-27. No clearer or more precise demonstration of
a cytological fact could be asked than is here afforded by the
unilateral movement of the accessory chromosome.

The question naturally arises, why, if this be true, should
two observers decide that. the accessory chromosome does here
divide? It is of course, impossible to explain just why this has
occurred, but it would seem to me that Foot and Strobell were
impressed with the necessity of showing every chromosome in
both daughter cells and that they chose stages which exhibited
such arrangements. Since it is impossible to find these con-
ditions after a mid-anaphase, owing to the massing together of
the chromosomes, all their pictures were made in the early
anaphase while the accessory chromosome lies in or near the
equatorial plate. Not a single photograph of an unmistakably
completely divided second spermatocyte is shown among all
their figures. Since it is only at this time that the accessory
chromosome may definitely be demonstrated to oceur in only
one daughter cell their failure to figure it may be understood.

Since Foot and Strobell base their case upon the evidence
afforded by their photographs, it is proper that these should
be considered. They show in plate III twenty-three figures of
second spermatocyte anaphases and telophases. In all but
figure 45 the accessory chromosome lies undivided in the equa-
torial plate, and in this two independent nuclei with no clearly
marked accessory chromosomes are advanced as evidence that
the accessory chromosome has divided. Figure 46 shows also
two nuclei with no indication that they are dawghter deriva-
tives. Figures 29, 30, 31 and 44 are introduced to show the
actual division of the accessory chromosome. Figure 29 shows.
the accessory chromosome of a mid-anaphase with a slight
constriction at the center, figure 30 a similar cell with two
distorted and ill-defined chromosome masses, figure 31 two
other similar daughter groups with an apparently divided lag-:
ging chromosome, while figure 44 exhibits a mid-anaphase with
the daughter (?) groups widely separated and turned by spread-

M’CLUNG: CHROMOSOMES OF ANASA TRISTIS. 379

ing and a broken accessory chromosome midway between them.
In all of these cells the chromosomes are thin and of indefinite
outline and have apparently suffered much in the process of
preparation. There is not a clear case of a divided accessory
chromosome in the series, although individual cells, doubtless the
best that could be found, are selected and photographed: That
later stages show conclusively the undivided accessory chromo-
some, not alone in selected cases but in every cell in a micro-
scopical field, is demonstrated in plate LX XI, figure 15, of this
paper.

The objection may be raised that such a stage does not show
all the chromosomes; but it is not a valid objection because it
is impossible to photograph all the chromosomes at this stage
on account of their apparent fusion into a mass. Moreover.
the essential facts are admitted by all. It is agreed that all
the chromosomes, with the exception of the accessory chromo-
some, divide so that each group contains ten ordinary chromo-
somes; it is agreed that the lagging chromosome is the acces-
sory chromosome. The question therefore to be answered.
concerns the behavior of the lagging chromosome. Are its
derivatives found in each of the separated daughter cells of the
second spermatocyte; or is it, undivided, included in only one
of the two? The evidence on both sides has been presented in
statement, drawing, and photograph and the judgment of those
interested must, in the absence of personal study, be based
upon the presentations. The crux of the situation lies here,
for if the accessory chromosome divides there is no dimorphism
of the spermatozoa and no visible sexual differentiation of the
paternal germ cells, but if it remains undivided then there is a
dimorphism of the spermatozoa and a consequent visible chro-
mosome differential between two numerically equivalent
classes.

It is pertinent to consider also whether in evaluating the
evidence in this case, any considerable number of genuine
instances of unusual conditions can be held to invalidate the
common conclusions reached by numerous independent stu-
dents of these phenomena. There can be no question of the
occurrence of abnormal conditions in the development of the
germ cells. Entire cysts, or even testes, become degenerate
from imperfect coérdination in development ; certain daughter

%—Univ. Sci. Bull., Vol, V, No. 20,

380 -—»-« KANSAS* UNIVERSITY

groups. of cells: fail’ fail'to separate and tl
with two or four ‘tails and as many s
- somes; minor aberrations occur
these: be held to prove that there is no con

THE

KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vol. V, No. 21— March, 1911.

(Whole Series, Vol. XV, No. 21.)

CONTENTS:

FOUNDATIONS OF ARITHMETIC........:............ Arthur Bowes Frizell.

PUBLISHED BY THE UNIVERSITY,

LAWRENCE, KAN.

March, 1912.

Entered at the post-office in Lawrence as second-class matter.

3-3310

Me 27's

asa es

ie

THE KANSAS UNIVERSITY
SCIENCE BULLETIN.

Vou. V, No. 21] MARCH, 1911. you XV. No. tt

THE FOUNDATIONS OF ARITHMETIC.

DISSERTATION FOR THE DEGREE OF DOCTOR OF PHILOSOPHY,
SUBMITTED TO THE FACULTY OF THE GRADUATE
SCHOOL OF THE UNIVERSITY OF KANSAS.

By ARTHUR BOWES FRIZELL, OF BOSTON.

This thesis seeks a foundation for arithmetic in the
ideas underlying Cantor’s formulation of his system of
ordinal types.

It proceeds by postulating, but follows D. Hilbert
and G. Peano rather than E. V. Huntington. The
search is for postulates possessing heuristic and didac-
tic, not merely subsumptive value.

A motif is found in the notion of an abstract group,
_ which secures the development step by step of all num-
ber systems so far studied without further postulates
than those needed for the transfinite ordinals.

As axioms are to be avoided, it is necessary to state
carefully the definitions and theorems used even when
they are well known to the mathematicians, but in this
case the proofs are omitted.

1. Definition. A set of symbols a, b, . . . will be
said to form a K-class if we possess a test which enables
us to assert in every case either that a = dD or that a is
not = b subject only to the restrictions that

(383)

884 KANSAS UNIVERSITY SCIENCE BULLETIN.

a) the result of comparison be uniquely determined

b) the statements =a and a=6 shall be interchange-
able

c) from a=b and b=c must follow a=c

d) a=a if K contains no other symbol equal to a.

2. Definition. An infinite set of symbols is one
which contains parts that can be put into one corre-
spondence with the whole. A finite set is one that is
not infinite.

3. Definition. An ordered set is one in which we
are always able to say either that a precedes b (a<b)
or that a follows b (a >b) subject only to the restric-
tions that
a) the result of comparison is uniquely determined
b) a<6b shall exclude a=b but involve b>a
c) ifa<band b<c, thena<c
d) if a’ =a and b’ =, then a’ <D’ or a’ >b! according

asa<bora>b.

4. Definition. An ordered set is said to be well
ordered when it has a first element and every subset
beginning with the first element has an immediays suc-
cessor in the given set.

5. Defination. A procedure whereby to every a, b
of a K-class is assigned a definite symbol e= aob will
be called a C-rule or rule of combination provided that
by this assignment equals with equals give equals.

6. Definition. A K-class is said to possess the
fundamental group property with respect to a C-rule
when ao b also belongs to K.

if Definition. Modulus of a K-class with repeal to

a C-rule is a symbol wu of K such that aou=a=uoa
for every a in K.

8. Proposition I. An ordered set defined by the
requirements: a) it shall contain a given symbol e;
b) it shall possess the group property for every combi-

ae

FRIZELL: FOUNDATIONS OF ARITHMETIC. 3885

nation e ow where a denotes a previously defined mem-
ber of the set is also well ordered, infinite and forms a
K-class with respect to the given rule. Proof: By
hypothesis the set is to containeoe=e’, ece’ =e",
eoce’=e'”’,. . . Since it is to be an ordered set no two
of its members can be equal .*. to every element of the
whole set can be assigned some one of the subset e’,
ee”, . . . That is, the set is infinite. And by b)
every element has an immediate successor. Finally we
have a test for equality which satisfies the require-
ments of § 1.

9. Scholium. A set defined as in § 8 contains no
modulus. In what follows it will be referred to as an
e-set. | |

10. Postulates. We postulate an e-set for a lower
rule of combination which shall contain a lower symbol
u and e-sets for both this and a higher rule which shall
both contain a higher symbol w, the definitions of the
rules of combination to be completed in §§ 17, 26,
28, 30.

11. The different sets K[wo], K[wo], K[wal] are
to be so ordered that the lower shall precede the higher.
Thus they form together a well ordered set

2 uou=—w',uow=u",uou’=w"”,..

WwW, wow=wu’, woww'=wu", wow! =wu!”, ae

wow=w", wow" =w"", BO ee gs so

12. Postulates. We postulate e-sets for the lower

rule which shall contain respectively each of the sym-

om. we we, 6}. SC. Cin. succession. Thus
corresponding to every a of the set Flaps! we shall
have an e-set K[ao]:a, aoa =a’ aoau'’=au",
Now taking e.g. a= w"” by the principle of
11 every member of K[ wo] precedes ww" = wr
and therefore falls between w“” and w«*. That is,
all these new e-sets are interpolated between successive

386 KANSAS UNIVERSITY SCIENCE BULLETIN.

elements of K[wo]. Hence the totality of symbols
forms a well ordered set.

18. Proposition II. A rule of combination is asso-
ciative throughout a given K-class if

ao(boc)=aoboce
for every a, 6, c in K.

14. Definition. <A K-class is said to form a semi-
group with regard to a C-rule associative throughout K
if by this rule unequals with equals always give une-
quals or if from the relation a’0b =aob, resp. aob/=
aob we can always infer a’=a resp. b’=b.

15. Proposition III. No semigroup contains more
than one modulus for its defining C-rule. For if wu and
u’' not =u were both moduli, we should have aow/ =
a@=aow contrary to § 14.

16. Definition. An abelian K-class with reference |
to a C-rule is one for which the rule is without excep-
tion commutative.

17. Proposition IV. Sufficient conditions of an
abelian class are the relations

ao(boc)=aoboe and boa=aob
for every a, b, c in the class.

18. Proposition V. AC-rule is distributive over
another in a given K-class if the relations

(aob)c=acobe and a(boc)=aboac
are satisfied by every a, b, ¢ in K.

19. Proposition VI. A necessary and Gaficere con-
dition of the generating rule of an e-set being associ-
ative for the set is the inductive formula of definition.

eoaob=eo(aob).

Proof. If therelation of ao(boc) =ao0boe hasbeen
established for a certain a and every ), c of the set,
then by hypothesis eoao(boc) = eojao(boc)t =
eofaoboc} =eo(aob)oc = eoaoboe.

FRIZELL: FOUNDATIONS OF ARITHMETIC. 387

But eoboc=eo (boc) by definition.
Therefore e/oboc=e'o(boc) and so on.
Hence the theorem by strict induction.

20. Corollary. The rule is also commutative. For
if eoa=aoe, then by the associative law eoaoce =
€0o(a0e)=eo(eoa). But eoce’=e’oe by $19, there-
foreeoe’” =e”oe, andsoon. Andif boa=aob for
a certain a and every b then bo(eoa)=bo(aoe) =
boaoe=aoboe=eo0(a0b)=eoaob.

Hence the proposition follows by strict induction.

21. Proposition VII. The formula of § 19 is also a
sufficient condition of the e-set constituting an abelian
semigroup for its generating C-rule. Proof. If aob
_ belongs to the given e-set, then eoaob=eo (aob)
also belongs to it. Thus the first group property is es-
tablished. We have proved the associative law and by
definition no two numbers of the set are equal. There-
fore a’ ob is not = aob when a’ is not =a.

22. Postulates. Using capital letters M, N,....
to denote elements of the lowest class K[wo], and
italic letters a, ), . . . for those of the class K[wo]
where « shall precede ), we postulate for every MB =
BMaset MBoWNa for every Na, then using letters
a, b, . . . . where a shall precede b to denote the re-
- sulting symbols we postulate inductively new sets boa
where a= Naand b denotes successively the higher
“polynomial” or composite symbols. Every b-set is to
be simply ordered and possess the first group property
for every b o a. |

23. Proposition VIII. Every b-set is well ordered,
infinite and forms a K-class for the rule denoted by o.
For the b-set consists of the combinations b, boa,
boaw,boaw’,boaw’”, ... . andthe reasoning
of Prop. I holds good.

24, Proposition IX. The b-sets belonging to a

888 KANSAS UNIVERSITY SCIENCE BULLETIN.

given Mb form together an infinite, well ordered

K-class as regards the rule o. For the set of b’s:
Mb, Mbow™, Mbow™?o™, ‘

forms a well ordered K-class between consecutive mem-

bers of which the b-sets are interpolated.

25. Proposition X. The totality of the b-sets forms
an infinite, well ordered K-class as regards the rule o.
For the set of Mb belonging to a given 6 is a well or-
dered K-class and so is the whole set of 6’s, the latter
being simply K[ wo].

26. New combinations of symbols already defined
are defined inductively in accordance with the formulas
boa=aob and

ao(bocy= aod ove.
This secures the group property for the whole set and
the reasoning of VJ and VJI establishes

Proposition XI. The set of symbols generated ac-
cording to two rules, a higher and a lower, from a single
arbitrary symbol w by aid of the postulate of order,
the restricted form of the group property and the above
inducted formulas of definition, is a well ordered, infi-
nite set constituting an abelian a with regard
to the lower rule.

27. Scholium. The higher rule remains unre-
stricted and has not been defined beyond the set
K[waol], therefore is not yet a C-rule for the K-class
of § 26. .

28. Definition. The higher rule shall be defined
inductively throughout K[ wo ] by the formula

| wab=waab,
where a, b are any members of K [ wa] for which ab
has already been defined.

29. Proposition XI. The class K[ wa] forms an
abelian semigroup with regard to its generating rule.
Proof by Propositions VI and VII.

i a al I ll

FRIZELL: FOUNDATIONS OF ARITHMETIC. 389

30. Definition. The higher rule shall be defined
inductively for the remaining symbols of our abelian
semigroup on the lower rule according to the formulas

(aob)c=acobcanda(boc)=aboac,

where the combinations ab, ac, be shall have been
already defined.

31. Proposition XIII. The higher rule is distribu-
tive over the lower throughout the class of symbols
defined in § 22. Proof. First let a=b=c=Il where
lis any symbol of K[wao]. Then ll is defined by § 28
and /(lol)=lloll=(lol)l by $380. Now suppose
that the formulas of § 30 have been established for all
symbols of K[/o ] up to and including a certain a and
for every ) and c.

Then (Loaob)c={lo(aob) }c by associative law,
=lceo(aob)e_ by hypothesis,
=leo(acobc) by hypothesis,
=lcoacobe by associative law,

: =(loa)cobe.
Similarly in every other case when a, }, ¢ are replaced
byloa, lob, loe respectively. But by definition

. l(loNa)=lloaoNaand soon. Therefore by strict

induction the formulas hold for every a, b, c belonging
to the same K[/o]. Hence by Prop. V the theorem
is true in this case. If h, k, | denote different mem-
bers of K[ wo | the definition gives
(hok)l=hlokland h(kol)=hkohkl.

Then the proof is completed inductively by the same
reasoning as above.

32. Corollary 1. The higher rule is a C-rule and
by it unequals with equals give unequals. For since
our symbols form a well ordered set according to the
lower rule, we have a = box for any two unequal ele-
ments a, b and therefore ac =bceoxc.. not =be.

33. Corollary 2, The higher rule is associative,

390 KANSAS UNIVERSITY SCIENCE BULLETIN.

For this property has been established for the ele-
ments a, b, . . . . as members of K[ wa]. Therefore
by the distributive law it holds generally.

34, Corollary 3. The higher rule is commutative.
For this also has been proved in the class K[ wo ].

35. Corollary 4. The set of symbols defined in § 22
forms an abelian semigroup with reference to the
higher rule.

36. Scholiwm. Beginning with any a ot w we have
abelian semigroups with regard to both rules, and so
beginning with w for one or the other rule, bat no part
beginning with w forms a semigroup on both rules.

37. Proposition XIV. Necessary and sufficient con-
ditions that a set M constitute an abelian semigroup
with respect to each of two rules of combination, one
distributive over the other, that M contain an arbitrary
wand that M be as generated an ordered set, are the
rules of combination thus far defined.

38. Corollary. The properties in question may be
expressed by the following postulates:

A. There shall be a higher and a lower rule of com- _
bination. |

B. Combinations of symbols shall be so ordered that
those made by the lower rule precede combinations of
the same symbols by the higher. .

C. The higher rule shall be distributive over —
lower.

D. There shall be a set M containing the symbol w.

E. M shall form an abelian semigroup for the lower
rule.

F. M shall form an abelian semigroup for the higher
rule.

G. Every successive set generated according to re-
quirements shall be an ordered set. .

ee

FRIZELL: FOUNDATIONS OF ARITHMETIC. 391

39. Scholiwm. M is a well ordered set with no
modulus.

40. We postulated two rules of combination for the
symbol w, but only one for the lower symbol wu. If
we should postulate a higher rule for u by the same
requirements as for w the resulting set of symbols
would not differ as regards any group property from

the w-set; the new abelian semigroups would be re-

spectively holoedrically isomorphic with the former.
If we postulate no higher rule the set K[uo] still
forms an abelian semigroup holoedrically Asomorphic
with K[wo].

41. Instead let us set up postulates for the wu class
of symbols differing from those of § 38 only in adding
another postulate H. The symbol ~ shall be a modulus
for the higher rule, with the restriction on A-G that
they shall not conflict with H. Then the class K[wo]
reduces to the single element u, the abelian semigroup
on the two rules reduces to the set K[wo] and postu-
late B drops out, since every combination by the higher
rule is found among those made according to the lower
rule. This new set of postulates may be replaced by
the equivalent set.

42. a. There shall be a higher rule defined from a
lower by the inductive formulas (woa)b = uboab and
a(uoh)=auoab.

b. There shall be a lower rule defined inductively by
the formula woaoh=uo(aob).

-¢e. There shall be a set M containing the symbol w.

d. The set M shall possess the group property for

every combination woa.

e. M shall be an ordered set.

f. There shall be a set M’ built up by postulates ec, d
and e on the symbol ww.

g. The sets M’ and M shall be identical.

892 KANSAS UNIVERSITY SCIENCE BULLETIN.

43. Proposition XV. The set of symbols defined
by the above seven postulates constitutes an abelian
semigroup for each rule and has a modulus for the
higher rule, which is distributive over the lower.

Proof. The abelian semigroup on the lower rule is
established by Prop. VII, the distributive law follows
from XIII and the other semigroup by $35. It re-
mains tu prove the existence of a modulus.

Let e=wuu. By postulate g, e must occur in the set
GG ET oM Epes Ber) Aly Meee ee |

Suppose e=wu’’. Then every member of K[eo] will
be found in K[uwo] but beyond wu’. That is, w and w’
are not in K[eo], contrary tog. Therefore the only
possibility of satisfying g is e=wu. Since, then,
uu=u, it follows by the distributive law that wu is a
modulus.

44, Postulatesa . . . g define the natural num-
bers as ordinal symbols and by XV contain all the laws
of their arithmetic. The cardinal numbers may be de-
fined in a manner now familiar as names of classes,
e.g., ‘‘five” is the name given to the class of all well
ordered sets which are ordinally similar to the set w,
a, Pes 7) ed A uu”.

45. Definition. A group is a semigroup which con-
tains, corresponding to every a, /, in it, symbols p
and q such that aop=b=qoa.

46. Proposition XVI. Every group contains a
modulus with respect to its defining C-rule.

47. Definition. If a class which has a mod-
ulus wu for its defining C-rule contains symbols a, a
such that aoa =u=aoa then a, a are said to be
each the inverse of the other.

48. Proposition XVII. No member of a semigroup
can have more than one inverse in the semigroup. For
otherwise equals with unequals would give equals,

FRIZELL: FOUNDATIONS OF ARITHMETIC. 393

49. Proposition XVIII. Every group contains the
inverse of every one of its members.

50. Proposition XIX. A semigroup with modulus
is a group if it also contains the inverse of every one
of its members.

Proof. ao(aol b) =aodob=uohb= b, and
(boa)oa= bo(aoa) =bou=b, that is, § 45 is sat-
isfied by p=aob and g=boa.

51. Proposition XX. Given an abelian semigroup
G with reference to a rule denoted by o, if in the class
[(m,n)]=C of pairs of elements of G we declare
(m, 7) = (n, p) when and only when mop=nogq and
set up a rule defined by the relation (m, q)°(n, r) =
(mon,gor), then 1) C will bea K-class, 2)o a C-rule,
3) C an abelian group as regards the rule denoted by
the sign ©.

Proof. 1) Either moO p=noq or mop is not =noq
by hypothesis and §1. Hence either (m,q) =(%, p)
or(m,q) is not =(n,p). Obviously the assertions
(n,p)=(m,q) and (m,q)=(n,p) are identical in
meaning, since this is so for nogand mop. It re-
mains to verify the euclidean postulate.

Suppose that (m,q)=(l,r) and (l,r)=(%,p).

Then mor=loqandlop=nor.

Hence (mor) o (lop)=(no7r) 0 (log).

But since G is an abelian semigroup (mor)o(lop)=
morolop=(mop)o(lor) and likewise (nor)o
(log)=(noq)o(lor). Therefore (mop)o(lor)=
(noqg)o(lor). Whence mop=nogq by definition
of semigroup and finally (m,q)=(,p) by definition
of equality.

2) Let (m’, q’) = (m, q) and (n’,r’) = (",r). There-
fore m’og=moq' and n’or=nor’. Whence as
under 1) (m’on’) 0 (qor)=(mon) o (q’or’) and
hence (m/on’,g'or’=(mom,qor). That is (m’,q’)o

394 KANSAS UNIVERSITY SCIENCE BULLETIN.

(n’, vr’) =(m, 7) (nN, 17), or equals by equals give equals
and © denotes a C-rule. |

3) The fundamental group property is secured by
the definition. To prove the associative law we have
(m,n)°}(p,9)0(7,8)§ = (mopor,nogos)=(m,n)o
(p,q)o(7,8). Suppose that (m’,q’)o(n, 7) =(m, q)e
(n,r). Then (m’on,q'or’)=(mon,qor) by defi-
nition. Hence (m’on)o(qor)=(mon)o (qd or)
and (m’oq)o(nor)=(moq)o(nor). Therefore
m'og=moq by $14. That is (m’,q’)=(m,q) and
similarly for the other form of this property. Thus C
is a semigroup. But the element (m,m) is a modulus
and to every (”,q) corresponds an inverse (q,7).
Therefore by XIX, C is a group, which we will denote
by G.

Finally (n,r)o(m,q)=(nom,roq)

. = (mon, qor)=(m,q)o(n,7r).
Therefore by § 16, G is an abelian group.

52. The semigroup C and group G are connected by
Proposition XXI. If we declare (mogq,q) =m then
(mog,q)o(nor, r)=mon. For(mog, q)o(no7r,r)
=(mogonor,gor)=Monogor, gor)=Mon.

53. Scholium. The group G contains a semigroup
K holoedrically isomorphic with C in such manner that
the rule © becomes identical with the rule o for K.

54. Obviously Prop. XX may be applied to the set
of natural numbers in two different ways according as
we use the sentigroup on addition or that on multipli-
cation. An essential distinction between these two
procedures is furnished by ,

55. Proposition XXII. Given two rules of combi-
nation of which one is distributive over the other, no
set of symbols can form a group with respect to both
rules. For such a set would contain a modulus v for
the lower rule (XVJ). Then by the distributive law

a ——— a =< ee!

' FRIZELL: FOUNDATIONS OF ARITHMETIC. 395

a=bx=b(xov)=bxobv= aobv forevery a. That
is bv is alsoa modulus. But there can not be more than
one modulus (J/J). Therefore bv =v for every b so
that when b’ is not = ) we have }’v = v = bv contrary

to definition.

56. Corollary. It is not possible to define a set of
symbols constituting a semigroup with modulus as re-
gards both rules of § 55.

57. If Prop. XX is applied to the set of natural
numbers and the rule of addition, we lose the semi-
group on multiplication and can not recover it. This
semigroup, in fact, is destroyed if we only annex a
modulus for addition. Consequently we give the pref-
erence to the semigroup on multiplication and proceed
with the group G obtained by applying XX to it.

58. Proposition XXIII. Given a higher rule dis-
tributive over a lower, and a set of symbols forming an
abelian semigroup with respect to each rule, if we build
the group G of § 52 with reference to the higher rule,
then a necessary and sufficient condition of having a
lower rule over which the higher shall be distributive
and with regard to which G shall constitute an abelian
semigroup is the formula of definition (m, q)o(n, 7)
=(mrongd, qr).

Proof. The necessity of this relation is obvious. To
show that it is sufficient we first let (m’, q’) =(m, ¢)

mn’, 7')—(n, r)..(m'q=mgQ’) and n’r=nr’.
_ Therefore m/r’qr=m'qrr’=mrq'r' and n’q/qr=
nqq'r'. Hence (m/r'on’q)qr=(mrong)q’r’ by
distributive law and .°. by definition (m’r’on'q'q’r’)

=(mrongq, gr) or equals with equals give equals and
we haveaC-rule. Similarly equals with unequals give
unequals. For the associative law (k, 1)oj{(m, q)o
(n, r)} =(k, l)o(mrong, qr)=(kqrolmrolng, lar)
=(kgolm, lq)o(n, r)=(k, l)o(m, q)o(n, Tr).

896 KANSAS UNIVERSITY SCIENCE BULLETIN.

Obviously (n, 7)o(m, gq) =(m, q)o(n, r). Thus
we have an abelian semigroup and the distributive
principle follows by taking equals with equals.

59. Definition. The symbols of G shall be ordered
according to the convention that (m, q) shall precede
or follow (, q) according as m precedes or follows n
and by § 8. ,

60. Scholium. Ifa, » denote two natural numbers
that have no common factor other than unity, the set
of symbols (a, b) is simply ordered by the above defi-
nition. This set, which we will denote by R, may be
taken as representative of G since every element of G
is equal to some member of R.

61. Definition. The symbols constituting the set R
are called absolute, rational numbers. )

62. Proposition XXIV. Every two unequal ele-
ments of G are connected by a relation h = go where
h,g denote the given elements, g being that which
precedes.

For g = (m, q) =(mr, qr) and h = (n, r) = (nq, qr)
and mr precedes nq by hypothesis. Therefore h=gox
is satisfied by « = (d, gr) where ng=mrod.

63. Corollary. The set R is simply ordered. accord-
ing to the lower rule denoted by o.

64. Proposition XXV. If we exclude the modulus
u the remaining symbols of R form a set which is sim-
ply ordered according to the higher rule. For let x
denote any element of R which precedes and y any one
which follows wu.

Then gw precedes and gy followsg. That is, the
trios u,g,gv and wu,g,gy are ordered whether g
precedes or follows w.

65. Corollary. The modulus uw separates the re-
maining symbols of R into two simply ordered classes,
neither of which contains either a first or last element.

= al

S|.

FRIZELL: FOUNDATIONS OF ARITHMETIC. 397

66. It is possible to divide R into two ordered
classes so that every element of R belongs either to the
one or to the other, but neither class has either a first
oralast member. For example, the class R, of all ele-
ments whose squares precede 2 and the class R, of all
whose squares follow 2 together exhaust the set R,
while neither R, nor R, has either a first or a last ele-
ment. If we take 4 instead of 2, the element 2 is not
included either in R or in R,; it separates them. We
usually make this separation in practice by selecting a
well ordered set, e. g..according to the decimal scale.
We take first the highest integer in R,, then the high-
est number of tenths, hundredths and soon. Similarly

we pick out first the lowest integer in R,, then the

lowest set of tenths, hundredths, ete.

67. Definition. A well ordered set which has no
last element is called a series.

68. Definition. The series of all symbols of the
well ordered set 1,2, . . . w,w-+l, .
eo |. Ow, 7) Hie" A ee aaa which pre-
cede the element a taken i in the above order, is said to
be a series of type a.

69. Definition. An ordered set which has no first
nor last element will be called an unbounded set.

70. Given an unbounded set S ordered according to
a rule with respect to which S constitutes an abelian
semigroup, suppose that a series K[f,] of type w has
been selected from among the elements of S and let
W, denote the set of all symbols of S which follow, V,
all which precede every f,, let V, denote the set of all
symbols s which precede W, and W, the set of all that
follow V,._

Then one of the sets W,, V, must be unbounded and
both may be. So also of V, and W,.

3—Univ. Sci. Bull., Vol, V, No, 21.

398 KANSAS UNIVERSITY SCIENCE BULLETIN.

71. In every one of the four cases when
W, has a first element or V, has a last
V, has a last element or W, a first
this element is called the limit of the set to which it
belongs.

72. Incase W, and V, or V, and W, are both un-
bounded. we will introduce a new symbol f,, which we
will call the limit of K[f,,].

73. Lemma. If W;and V; are both unbounded for

the series K[f,] the same is true for the series —

K{[aof,] where ais any symbot inS. For if either
the W or V set corresponding to K [aof, | had a limit
lL we could write 1/=aovw and x would be a limit for the
same W or V_ belonging to K[f,| contrary to hy-
pothesis. —

74. . Theorem. If the series K[f;|, K[g,;]| both
divide S into two unbounded sets the same is true of
the set K[f;og;]. |

Proof, Let hi" 1,0 2;, h, = tyoey7s 123

2 Pe = f,02,, hes = Ty OB; ea,
New41 = f,0¢,; Regis = f,02>, woe

Then by the Lemma each of the series K[h,], |
Kho) Kh oe eT i ao

into two unbounded sets. Hence the same is true of
the set fy hj) The ase FO Ta |

75. On the basis of the preceding theorem and defi-
nitions is built up, by laying down the usual definitions
of equality, order and two rules of combination, the
set X of limits of series K[f,,] selected out of R. The
set X forms an abelian group with reference to the
higher rule of combination and an abelian semigroup
on the lower rule. The higher rule is distributive over
the lower and possesses the modulus wu. X may be
simply ordered according to the lower rule and is di-
vided by the modulus u into two unbounded sets each

FRIZELL: FOUNDATIONS OF ARITHMETIC. 399

simply ordered according to the higher rule. X is itself
an unbounded set and contains a sub set holoedrically
isomorphic with R as regards each rule of combination. |

Thus the whole set of absolute numbers is deduced
from the set of absolute rational numbers by using only
principles of order.

76. The preceding development does not take ac-
count of the possibility that K[f, |] may not really di-
vide S. There may be no symbol which follows every
f;, Then W, is an empty class and V, is taken to coin-
cide withS. Thus if K/f,| and K|g,]| are both
series of this kind, they both have the same W and V,
whence if new symbols f,,, g,, were to be introduced
the principle of definition used in § 75 would lead us to
declare f,,=g,,. Therefore we assign to the totality of
the sets K | f, | for which W, is an empty class a single
symbol Z which is to follow every element of X. Then
following § 75, Zo g,, is to be defined as h, where h, =
f,og,andf,=Z. But for K[h,], W, is empty,
therefore Zox=Zog,=h,=2Z. That is, Z+e2=
L=2+2Z,Ze=Z=<2xZ, and similarly Z+Z=Z= -
ZZ=Z*=Z%. The associative, commutative and dis-
tributive laws still hold as well as the first group prop-
erty, but the semigroup is destroyed for both rules by
violating the law of equals with unequals.

77. It is also possible that no symbol in S precedes
every f;. Then V,is an empty class and W, coincides
with S. On the same principle as in § 76 we assign to
the totality of series of this kind a single symbol v to

_ precede every x and define vo x = h,, where h, = f, og,,

g, =, f,=v. And now the two rules must be dis-
tinguished. Since no «w precedes every f; the sets
K|f,+g, | and K[g,] separate S into the same V
and W. Thereforev+«=h,=g,=«+v. And for
the same reason no s can precede every f, g,

Ves(if) =—v=2Vv

400 KANSAS UNIVERSITY SCIENCE BULLETIN.

Also in like manner v-+v =V=vVv.

In the set K|[h, v] the associative, commutative
and distributive laws and the first group property are
all preserved. The law of equals with unequals is
violated for multiplication but still holds for addition,
for which v is a modulus.

78. Since the introduction of infinity would destroy
both semigroups while zero destroys only that on mul-
tiplication, it seems preferable to admit zero to our
arithmetic and exclude infinity. But this removes the
only objection to enlarging the addition semigroup into
-agroup. This is effected at a stroke by applying Prop.
XX to the set X and thus building an abelian’ group
with reference to addition. In this group we then de-
fine a higher rule by the formulas

@ob—ab=aobd and aob=ab. *

This completes the system of real numbers, forming
an abelian group on addition and, when we exclude its
modulus, an abelian group on multiplication, which is
distributive over addition. The system of real numbers
can be simply ordered according to the lower rule, but
the abelian group on the higher rule can not be simi-
larly treated.

79. We might have proceeded by first applying XX
to the natural numbers as a semigroup on addition.
This yields the whole set of integers, positive, negative
and zero. Omitting zero and defining by the law of
signs we should have an abelian semigroup on multi-
plication. Applying XX to it, there results an abelian
group composed of all rational numbers except zero.
Then the introduction of the limits would supply the
whole set of irrational numbers and close by reintro-
ducing zero.

80. As long as we postulate two rules of combina-

tion, one distributive over the other, and demand semi-

Si oe My yy:

a

FRIZELL: FOUNDATIONS OF ARITHMETIC. 401

groups with reference to both, we are led inevitably to
the transfinite types in case we impose no restriction
on the multiplication table and, if we require a mod-
ulus, to the natural numbers, whence the system of
real numbers results from the attempt to build groups.
The two rules of combination connected by the distrib-
utive property may be regarded as defining arithmetic
up to date; it has not yet been found profitable to pos-
tulate in any other way. Then transfinite arithmetic
is distinguished by postulating the semigroups and
finite arithmetic by the postulate of the modulus.

81. The real numbers form the most general finite
system with a single unit and must enter into every
system with more than one principal unit, one unit
must always be the modulus. Systems with two or
more principal units have been studied exhaustively by
Weierstrass. Here the abelian group on the lower rule
is postulated universally.

With two principal units, if we exclude the modulus
of the lower rule, the necessary and sufficient condition
of an abelian semigroup on the higher rule is the exist-_

ence in the system of asymbol i such that ii=w.

Thus the common complex numbers form the most
general system with two principal units satisfying the
postulates for real members.

82. Within the system of common algebra are dis-
tinguished different number bodies each built on a root
of a given algebraic equation as aunit. The algebraic
numbers of a given body form an abelian group on ad-
dition and, excluding zero, an abelian group on multi-
plication, and must contain a given symbol, the root of
the given algebraic equation. For example the system
of common complex numbers is the number body which
contains a root of the quadratic x*+-1=o0,

402 KANSAS UNIVERSITY SCIENCE BULLETIN.

83. An integral algebraic number satisfies an equa-

tion
aM ay NT) ei. Sa a ee ei

where every a;is an integer. A good illustration of
the serviceableness of the group theory process for
arithmetic is the theorem: The whole numbers of a
quadratic number body form an abelian group on addi-
tion and, if we exclude zero, an abelian semigroup with
regard to multiplication.

84. A more striking illustration is furnished by
Dedekinds’ ‘‘Ideals.” An ideal is defined as a combi-
nation of the whole numbers of a number body posses-

sing the first group property for both addition and —

multiplication. The ideals of a given body form an
abelian semigroup with respect to multiplication.
There is no addition of ideals.

85. Similar applications are found in the expression
of an ideal by aid of its basis and in the cognate formu-
lation of the transformations of a collineation in space
of N dimensions. The latter question resolves itself
into that of complex numbers with N principal units.
Here it is no longer possible to preserve even the re-
stricted abelian semigroup on the higher rule.

Not only is the abelian character lost, as in quater-
nions, but the semigroup property may be violated on
account of the possibility of a combination by the
higher rule being zero when none of the factors is zero.

86. It is now possible to describe more concisely the
relation of the transfinite arithmetic to common arith-
metic. The symbols defined in § 22 forma set of num-
bers with infinitely many principal units w, w, w*, . .
i. €., the principal units form a series of type w. In
finite arithmetic, it is true, the symbols w?, w3, . . . all
belong to the set generated from w as principal unit,
but if we allow this analogy in the transfinite system
there will be still more new symbols.

7G

3 4 ~ me

5 ee, a ee ee —-

ee
.

FRIZELL: FOUNDATIONS OF ARITHMETIC. 403

For just as the removal of restrictions on the higher
rule carried our series of symbols beyond the finite set
of type w so the introduction of a still higher rule, cor-
responding to involution, will extend it beyond the set
defined in § 22 if we impose no restrictions on the new
rule except that of order and the special form of the
group property.

87. Let us postulate a rule expressed by a” higher
than the two preceding, that is, the combination ab
shall precede a’. To distinguish we will now call ab
the lower and aob the lowest rule. |

The lower rule has been shown to be associative and
commutative for the set K[wo]|, which forms an
abelian semigroup with regard to it. A set of symbols
shall be generated from w by the higher rule in accord-
ance with the postulate of order and the restricted
form of the group property for w*, where a is a pre-
viously defined element. By Prop. I this set forms a
series w, w’ = w’, w’” = w", wr'w'”, . . . and now w’
must follow every wX (N =2,3,...)[w® is nota
combination by rule, the N is a mere index |. Then
by § 22 we obtain a series holoedrically isomorphic with
the set there considered if to the lowest and lower rules
respectively we make the lower and higher correspond.

88. Thearticulation of the two lower rules in §§ 28-35
leaves room for much freedom in definition. There we
defined the lowest rule first and made the lower de-
pend upon it. Here we have already defined the lower
rule to the extent implied in the application of § 22.
This, however, involves no further restriction than the
inductive associative formula of § 19. But the defini-
tion from the lowest rule carries this property with it.

It is therefore permissible to define the lowest rule
so that the lower shall be distributive over it. Then
by the process of § 22 new combinations are made by
the higher and lowest rule together, from each series

404 KANSAS UNIVERSITY SCIENCE BULLETIN.

of type w in the present set of symbols a new set of

the same type as that in § 22. And all new series
being fitted in by the nature of the process between
consecutive elements of the series previously defined
we have at every stage a totality of symbols forming a
series.

89. The group properties of the two lowest rules
are preserved throughout the preceding process. On
every new symbol not obtained from the preceding
by postulating them is built up a set forming abelian
semigroups on both the lower and lowest rules, the new
symbol playing the part of a unit precisely like w in
the series of § 22, and in this set the lower rule is dis-
tributive over the lowest. This process is to be con-
_ tinued as long as new symbols can be obtained by it.

90. There remains one more possibility of building
new symbols by introducing a combination defined one
way of the higher symbols with those of the series
K [wo] according to the lowest rule. Thus to every a
in the higher series will be assigned a w-series a, aow,
aou'., acu’, Pete) oeseame

This series fits in- between a and its next following
element in the former set so that we still have a to-
tality forming a series.

91. For completeness it is convenient also to define
combinations one way of certain symbols with those of
K [wo] according to the other two rules, viz.: those
expressions in which the symbols of K[wo] have
already been used as marks or indices, e. g.:

ww, ew ee eee et

u'w’, wl; ES AE A 1) haa ee". UF atae
Ticpreséiall WORSE Oboe. a? ; ae.

Ut!” WR ogee re not defined, neither are ‘there

any ‘combinations according to the higher rule except

the set w* where a belongs to this set, The higher

ee ee

Se Je nm

i i

FRIZELL: FOUNDATIONS OF ARITHMETIC. 405

rule has not been made a rule of combination for any
whole set. This would involve assigning properties to
it and completing the arithmetic of the symbols we
have defined, which is beyond the present purpose. It
is enough to have them in series; that forms the foun-
dation of their arithmetic. -

92. The mere existence of the series of symbols de-
veloped in the preceding § § is enough to solve a great
variety of problems arising in analysis, the nature of
which will be illustrated by comparing this series with
series consisting of absolute numbers. The natural
numbers form a series of type w if they are arranged
in the order of their genesis. It is, however, possible
to arrange them in series of higher types for e. g. the
odd numbers alone form a series of type w. If to this
we add on the even numbers successively we obtain
series of types

wt+l,w+2, ... w+N, '
and thus the whole set is ordered in type 2w.
If we order the set R of § 60 as follows:
(1,1), (2,1), (8,1). (4,1), .
(1,2), (3, 2), (5, 2), (7, 2),
(1,3), (2,3), (5, 3), (7, 3),
we have a series which can be put ordinally into
one to one correspondence with the series 1, 2,

Deas wa 2) ek as Bwy Qwt+ ly oes. ies

is of type w?.

The same set of numbers (the common fractions)
can, however, be arranged ina series of higher type.
For every finite, simple, continued fraction is equal to
some common fraction and conversely. Now let us
order the finite continued fractions according to the
values of their successive quotients

Gi, V2, 93, + + + Qn:
Thus the continued fractions containing each a single
quotient form a series of type w. Then from each

406 KANSAS UNIVERSITY SCIENCE BULLETIN.

member of this series by adding a second quotient
results again type w. Therefore the fractions of the

form me 7 may be arranged in series ordinally sim-

Har tol, 2 suns. What ws ee ee
1. @. of type wr, Annexing a ‘third quotient replaces

each element of this set by a series of type w so that

we include types w?+1, w?+2,... 2w*%...

ers
3w?, . .. t%4¢@ the class of fractions — oe

‘“ qit got qa
is ordered in type w®.
Therefore the whole set of simple continued frac-

tions, comprehending types w,...w,...w,...
w, . . . constitutes, as ordered, a series of type
w’=w’'. !

93. Now it is easy to arrange the natural numbers
also in a series of type w’ as follows. We know that
the class of prime numbers can be put into one to one
correspondence with the whole set of natural numbers
(the number of primes is infinite).

Therefore we can set up a one to one correspondence
ordinally between the prime numbers and the simple
continued fractions with a single quotient. By thesame
reasoning the continued fractions with two quotients
are shown to be ordinally similar to the class of prod-
uct of two primes, 1. @., If to every quotient g, we
assign that prime p, for which q, is the ordinal number
in sequence (so that e.g. to the quotient 7 we assign
17) and likewise for a second prime p, and quotient q,,
and so on.

Proceeding in this way the class of all products of N
primes is exhibited as ordinally similar to the class of

1 1
continued fractions —— -+—, But the class of
qi : q2 ” Qn

all products of primes is the whole set of natural num-
bers. Therefore by § 92 the natural numbers may be
arranged in series w’. Q. E. D.

94. In other words we have here a method whereby

ee

FRIZELL: FOUNDATIONS OF ARITHMETIC. 407

the elements of any w-series may be rearranged SO as
to produce a series of type w”. Applying this method
successively to the w-series in the preceding we obtain

in place of w, 2w, ...w’, 2w’ ... that is, instead
of w? we getatypeww’. Thenw?+w, w?+2w,...
are replaced by ww/+w’, ww’+2w’, ... making

in all type 2ww’ replacing the former 2w”.

Then come in succession types 8ww’, 4ww’, ...
so that the original w* expands into a type w?w’. It
is clear that in this way we retrace precisely the steps
of the process of § 22, building up on each new symbol
the abelian semigroups according to the lower and low-
est rules. |

Therefore by successive applications of the method
it will be possible to rearrange the natural numbers in
series of types as high as any of the set hitherto de-
fined. This result may also be stated. The transfinite
ordinal series so far defined may each be put into one
to one correspondence with the set of natural numbers.

95. Thesymbolsw, w+1, w+2,... 12 4., the
transfinite symbols, are said to form the second ordinal
class, the finite symbols constituting the first class.
The latter class was said to be of type w, where w is
the symbol following next after all the finite symbols
of K[wo]. Likewise we shall say that the first and
second ordinal classes together form a series of type
W, introducing this new symbol without assigning to
it any properties except that it shall follow next after
all symbols of the second ordinal class, 7. e., after all
the set K[w*]: w, w’, w”,...

96. Wesee that the ordinal symbols play a double
role. They were defined as symbols forming a well
ordered set. But to each symbol which has no imme-
diate predecessor corresponds a series of which it is the
type, viz.: the series of all its predecessors or any
ordinally similar series. The arrangements of the nat-

408 KANSAS UNIVERSITY SCIENCE BULLETIN.

ural numbers in § 94 are only a part of these series.
But they are also only a part of the permutations of
the whole set of natural numbers. Other permutations
are obtained by the following Lemma. From any w
series of symbols may be obtained a set of permuta-
tions of the symbols forming a series of type w’ = w”.

Proof. Leta,, a,...a€y,.. . denote the given
symbols in the given order. Without changing the
order of the higher a’s put a, successively in every sub-
sequent place. This set of permutations is obviously of
type w.

From every one of them by repeating the process on
a, we obtain again a w-series .”. in all a series of per-
mutations of type w?. Repeating the process with a,,
ay, .. . . Successively we have a series of permuta-
tions whose type is w’. Q. E. D.

97. Thus the arrangements of the natural numbers
furnish a series of type not lower than W. For by $96
we first deduce from the series 1, 2, 3, . . . . 4. e., the
normal order, a set of permutations forming a series of
type w’. That is, to every ordinal symbol preceding w’
is assigned a permutation, and vice versa. Then by §
94 arrange the natural numbers in series w’ which ob-
viously is a different permutation from any of the pre-
ceding *.: it can not be obtained by the process of § 96.
Now repeating the method of $96 on each w-series of
this permutation we use up all ordinal symbols between
w’ and that which results from it by a second applica-
tion of $94. This holds step by step as long as the
latter process can be carried on. Thus to every ordinal
symbol in succession preceding W is assigned a new
permutation. That is, we have a set of permutations
of all the natural numbers forming a series whose ordi-
nal type is W.

98. The process just described for making permu-
tations of all the natural numbers can not yield a series

FRIZELL: FOUNDATIONS OF ARITHMETIC. eee

of type higher than W since, as we have seen, it gen-
erates precisely the series W of ordinal symbols. That
there are types higher than W is obvious, for we can
proceed with W just as with w to generate new e-sets
and new abelian semigroups, and there is no limit to
the possibilities in the way of still higher symbols and
rules. Now it is quite conceivable that there may be
further permutations of natural numbers not obtainable
by the above process. As a first step toward investi-
gating this question let us consider the simpler one
whether the natural numbers themselves can be ar-
ranged in a series of ordinal type higher than W. For
this purpose we. will establish the following

Lemma. Permutations of the natural numbers can
be made by the process described and ordered so as to
form a series of type higher than any series of all the
natural numbers, however arranged. For by the proc-
ess in question we can always form a new permutation
which differs from the first permutation in at least its
first element, from the second in at least its second
element...... from the (w+1)st in. at least its.
(w-+1)st element, and so on, therefore is not included
in any set of permutations ordinally similar to any
possible arrangement of all the natural numbers.
From this Lemma we readily obtain the

Theorem. Every possible arrangement of the natural
numbers is a series of the second class. For by the
Lemma to every such arrangement in series can be as-
signed a set of permutations forming a series of higher
ordinal type. But by § 97 the process by which this is
effected yields a set of permutations forming a series of
type W. Therefore every possible arrangement of nat-
ural numbers in series is of type lower than W, there-
fore belongs to the second ordinal class ( being ipso facto
>). .Q. HE. D.

410 KANSAS UNIVERSITY SCIENCE BULLETIN.

99. The arrangement of the natural numbers in
series of type higher than w finds application in the study
of infinite continued fractions. By aid of the euclidean

algorithm for greatest common divisor every absolute .

irrational number less than unity can be expressed as
an infinite continued fraction

1 1 pe ky
01 Og ee ae

and conversely. The class of infinite simple continued
fractions may therefore be taken as the representative
of the class of irrational numbers between zero and 1.

An infinite continued fraction can not be obtained
from a finite one merely by annexing quotients; it can
only be described by assigning a law which determines gy
for every value of N. An infinite continued fraction
_ may be formed, e. g., by the law that every quotient
shall be 2. It is sufficient, however, to consider the class
in which the quotients are all different; this can be put
into one to one correspondence with the whole class.
Accordingly we are concerned with the class of all
possible permutations of all the natural numbers.
These permutations may be examined in the same way
as the permutations of a finite set by imagining a frame-
work of places to be filled, but the number of places is
infinite. ‘Moreover, we must provide for the possibility
of filling the places in a series of order higher than w.
Thus an infinite continued fraction can be formed by
. filling the even places successively with the odd num-
bers in their natural order and the odd numbered places
with even numbers in the same way, 2. @.,

ois duis aa ae aes ee Bei
or by filling the odd numbered places with primes and
the even places with composite numbers. Or we can
select first the places whose indices are primes, then
the indices which are products of two primes, three

‘
ae Sw ee ee

Re OR oC ean | ee lena ee eo te tee

FRIZELL: FOUNDATIONS OF ARITHMETIC. 411

primes, and so on, and fill these sets with the correspond-
ing sets of numbers permuted in any way we please.

100. We have seen that every possible arrangement

of all the natural numbers in a series is of. type lower

than W. Therefore the quotient places of our frame-
work in the order in which we propose to fill them (or
rather state how they shall be filled) constitute an or-
dinal type of the second class. That is, the nwmbered
places in a permutation of all the natural numbers, ar-
ranged, however, in the order in which they are to be
filled, form a series obtainable by the process of § 97.*
Hence there can be no permutation of all the natural
numbers not obtainable by this process, since such a
permutation would be eozipso different from all of
those hitherto obtained and well ordered in some defi-
nite numbered places. In other words, this would mean
that the series of the places as they are filled, or the
series of numbers with which each place is to be filled,
was not obtainable by the process of $97. Therefore
all possible permutations of the natural numbers form
a series closely related to the series of type W.

BIOGRAPHY.

I, ARTHUR BOWES FRIZELL, member of the Protestant Episcopal Church,
was born in Boston, July 14, 1865. My parents were Joseph Palmer Fes-
senden Frizell, civil and hydraulic engineer, and Julia Anna (Bowes) Frizell.
My early education was chiefly at home and in a private school in Dorches-
ter, Mass. I graduated*from the high school in St. Paul, Minn., and spent
three years at the Massachusetts Institute of Technology, where I after-
wards served as assistant instructor in mathematics, 1888-’91. I received
the degree of Bachelor of Arts from Harvard College in 1893 and that
of Master of Arts from Harvard University in 1900. I served as instructor
in mathematics at New York University 1895-’96, and at Harvard 1897-1906,
when I resigned to study abroad. After three semesters at Gottingen, I
returned to America, and was appointed, 1908, Professor of Mathematics
in Midland College, Atchison, which position I resigned, 1909, to accept an
instructorship in the University of Kansas.

*And the preceding statements apply verbatim to every series of values
admissible in a given numbered place.

5 wes

PO Ne Fete F ne : ; ape cus oe Pe RF a

SCIENCE BULLETIN, UNIVERSITY OF KANSAS PLATE 46
VOL. V, No. 12

-

ER CRG

SCIENCE BULLETIN, UNIVERSITY OF KANSAS PLATE 47
VOL. V, No. 12

SCIENCE BULLETIN, UNIVERSITY OF KANSAS
VOL. V, No. 12 :

PLATE 48

-

: PLATE XLIX.

oo, _- Fig. 1. External view of the ei mandible of Bryops
vi Drawing by s. Prentice. a=ar i

Fie. 3. Fragment of cranium, inchiding sisetiont of
e ; e ing very distinctly the suture bounding the anterior
ane ig frontal. A portion of the oe line
- Seat ol es , : detected. x *- O= sien :

ae

SCIENCE BULLETIN, UNIVERSITY OF KANSAS. PLATE XLIX.
Vol. V, No. 18.

eS :

eS a a

)
:

wr

al view of the sysenemn

C = coracoid po: mea
L= on O=h Tus.

SCIENCE BULLETIN, UNIVERSITY OF KANSAS. PLATE L.
Vol. V, No. 13. : *

: \w ‘
AWS WA fi
tI AK ‘e \\* SW >
yi ‘ WY ANS /
ah ANS Bate:
uss

J}

Ite

ee

Meta Buly

Mine

Wi

Tyee:

-
Pes

5:

Vol. V, No. 18.

i) ee y/ /. ff:
| ll Hp
| | /

1

i
f

Wie 5
yin
fi NM ty
sabe / Ay

| / Mf
iy]

| ey
Tt

an
ao Ee

(3

Sid

x

; ; 4 i ; i <> E ; ;
: ae
<
i: | eN eryese ; '
= es
3 3 : : |

cs
;
;
:
"
:
4

ee ee a eee ee ae

SCIENCE BULLETIN, UNIVERSITY OF KANSAS. PLATE LIt.
Vol. V, No. 13.

—— i ow
er — a

: okt =
<s
\ =f
Y

\
\Vy

Pac
Ebi hs
Apt |

‘|

Ba
See Se
2

Fig. 3. Ventral v view of the "rn
Eryops willistoni, x Noe

arterial Sonapien
Fia. 8. Cross acta of the dlaviele.

' SCIENCE BULLETIN, UNIVERSITY OF KANSAS. PLATE LIII.
Vol. V, No. 18. ;

See

a;
De

Fig.
Fie. 3

ee i
wea : 3 8s
| | 8 Bsy
WM z
E

Fie.
Fie. 4.

PLATE LIV.

Vol. V, No. 18.

SCIENCE BULLETIN, UNIVERSITY OF KANSAS.

= PLATE
The dorsal ilium ¢
plates, XH

SCIENCE BULLETIN, UNIVERSITY OF KANSAS
VOL. V, No. 14

é
j

PLATE

55

[p> SOM eile oe)
; pa) Sea

siete ap
: rs

Ms e =

er

PLATE LVa

Fic. ae large dermal spine which was possibly
regis ee sates ee x Me

one-half iatuesl size. pas
Fic, 9. A pga dermal vials: which may }

ee

SCIENCE BULLETIN, UNIVERSITY OF KANSAS
VOL. V, No. 14

PLATE 66

SCIENCE BULLETIN, UNIVERSITY OF KANSAS PLATE 57
VOL. V, No. 14

eas
we

PLATE LVI.

Fic. 1. ‘The external surface of the ight pubis, x he
Fic. 2.
Fic. 3.
vertebra. xX %4.
Fic. 4. The lower portion of the right tibia,
the astragalus. One-half natural size.
Fic. 5. Lateral view of one of the dorsal ve
size.
Fig. 6.

hice %

PLATE LVIII.

SCIENCE BULLETIN, UNIVERSITY OF KANSAS.

Vol. V, No. 14.

Fic. 8. The left ulna seen fre
Fic. 9. The left ulna seen from

PLATE LIX.

SCIENCE BULLETIN, UNIVERSITY OF KANSAS.

Vol. V, No. 14.

oe
2 Mg
re

ary

G

;

Bee ori = The longitudinal muscular |

SCIENCE BULLETIN, UNIVERSITY OF KANSAS. PLATE LX.
VoL. V, No. 15.

imen of T

The spec

SCIENCE BULLETIN, UNIVERSITY OF KANSAS PLATE 61
VOL. V, No. 15

Fic. 2 ae Photogra h re the alim ry |
bubalus if rom the venta

form of the intestines oe vt
intestinalis Moodie. as

‘<

coral fn

ug

SCIENCE BULLETIN, UNIVERSITY OF KANSAS
VOL. V, No. 15

PLATE 62

PLATE LXMIL. . =

ee Upper figure:
Re and sp.
a Second sie

Next to boa ea ee figure:
Scirte tica ritensis, new species. x2 ee

Next to bottom ‘row, ri
Scirtetica ritensis.

eee ritensis. x ies

u Bottom row, right figure: a ac y of
ritensis. X 2. He elieeg ( sais

SCIENCE BULLETIN, UNIVERSITY OF KANSAS : PLATE 63
VOL. V, No. 17

ON

3 —
eae E
it *

OF KANSAS

a >
AY >
x00

lew Dorsal ew Side vi
Spermatophores of Gryllus

¥

“

e Abdomen of Gry)l

ssections of th “a
‘ass ew:
>=

Three Soran Di
Int 1140

¢ inus and Sperm fro
secreting tubules spermatophore

stodqoismiega

SCIENCE BULLETIN, UNIVERSITY OF KANSAS PLATE 64
VOL. V, No. 19

Rel
ay 3

ae

Mel is
gf tae

- PLATE LXV.

Fics. 1 to 10, inclusive—Polar views of the equaturial plate from
spermatogonial cells, each showing the entire number of chromosomes.
In figs. 1, 2 and 4 the possible pairing of equal elements is shown. There
are ten pairs. Nos. 1-1 are the largest two spermatogonial chromo-
somes, which form the largest chromosome of the spermatocyte group.
No. 11 is unpaired, and is the accessory chromosome. In figs. 5 to 10, in-
clusive, the connecting bands of linin are not shown.

Fic. 11.—Lateral view of the spindle in metaphase from a young
spermatogonial cyst.

Fig. 12.—Same from an older spermatogonial cyst.

Fig. 13.—Beginning anaphase from an older spermatogonial cyst,
showing the usual confluent condition of the chromatin.

Fic. 14.—Late spermatogonial anaphase.

Fic. 15.—A section of an entire spermatogonial cyst, showing the cells
in different stages of division. The polarity shown is typical.

Fig. 16.—Early spermatogonial telophase.

Fic. 17.—Polar view of the same, in which all twenty-one members of

the complex can be distinguished. (a#) Accessory chromosome.

Fic. 18.—Late telophase, from a young spermatogonial cyst.

Fics. 19 and 20.—Nuclei of spermatogonial cells, showing the chro-
matin at the point of greatest diffusion. The reticular structure is still
evident.

Fic. 21.—A typical spermatogonial nucleus, showing the beginnings of
the reorganization of the nuclear chromatin.

Figs. 22 to 28, inclusive.—Successive chests in the re-tormation of the
spermatogonial chromosomes.

Figs. 29 to 33, inclusive-—Nuclei in the process of synizesis. (#) Ac-
cessory chromosome. (t) Nuclear membrane. (c) Chromatin mass.

UNIVERSITY OF KANSAS SCIENCE BULLETIN PLATE 65
VOL. V, No. 20

. PLATE. LXVI.

Fig. 1.—An early stage of the spermatocyte nucleus, succeeding syni-
zesis. (#) Accessory chromosome. (p) Plasmasome.

Figs. 2 and 3.—Same as fig. 1.

Fig. 4.—Slightly later stage of the spermatocyte nucleus than is shown
in figs. 1, 2 and 8. The chromatin is more diffuse and the accessory
chromosome (#) is elongated, granular, and stains more deeply than the
other chromatin.

Fig. 5.—A spermatocyte nucleus at the period of greatest diffusion.
The chromatin is reticular. (p) Plasmasome. (#) Accessory chromo-
some more condensed than the form shown in fig. 4.

Fig. 6.—Same as fig. 5. The accessory is a homogeneous chromatin
rod and the plasmasome reaches its maximum size in this stage.

Fig. 7.—Spermatocyte nucleus in early prophase, showing the first
signs of chromosome reorganization. (p) Vacuolated plasmasome.
(x) Accessory chromosome. The difference in the consistency of these
two bodies is apparent in iron-hematoxylin preparations which were
stained for a shorter time, comparatively, or which were subjected to a
longer extraction process.

Fic. 8.—Spermatocyte nucleus, showing the beginnings of tetrad
formation and the accompanying phenomenon, the decrease in size of the
plasmasome. ‘

Fic. 9.—Later spermatocyte prophase. The plasmasome becomes
spherical as it disappears.

Fics. 10 and 11.—Two parts of the same nucleus drawn from adjacent
sections. There are thirteen bodies, all of which are shown. The plas-
masome has diminished greatly in size. (m) The univalent halves of the
small chromosome.

Fics. 12 and 18.—Spermatocyte cells in early prophase. The nuclear

wall in fig. 12 is not shown. The plasmasome is very lightly stained, in
marked contrast to the accessory, which holds the stain much longer.

Fic. 14.—Accessory chromosomes drawn from spermatogonial cells in
metaphase. For the method of identification see text.

Fic. 15.—Accessory chromosomes from cells of the same stage and.

preparation as that of fig. 12, where there is no trouble in identifying
this element.

Fic. 16.—Tetrads from the same stage and preparations as figs. 12.

and 18. (a) Largest form. (b) and (c) Typical forms of the cross.
(d) The small chromosome, the halves of which are not usually so close
together at this time.

Fic. 17.—Portion of a spermatocyte nucleus in late prophase, showing
the diminishing plasmasome (p) and the halves of the small chromo-
some (m).

Fics. 18 and 19.—Portions of the same spermatocyte nucleus appear-
ing in adjacent sections, showing thirteen elements—nine tetrads, two
m-chromosome diads, the accessory chromosome, and the plasmasome.
The plasmasome has almost entirely disappeared.

Fics. 20 and 21.—Same as figs. 18 and 19.

Fics. 22 and 23.—Same as the two previous figures.

Fics. 24 and 25.—Same as above. The nuclear wall in fig. 25 is not

shown.

UNIVERSITY OF KANSAS SCIENCE BULLETIN PLATE 66
VOL. V, No. 20

;
{
;
x
|

PLATE LXVII. |

Fig. 1.—Nucleus in the process of synizesis. (#) er
some. la.—Drawings of accessory apincge vers: from nuclei

comparative drawings were made.

Fic. 2.—Spermatocyte nucleus, immediately after synizesis,
both accessory chromosomes and the plasmasome. 2a.—Accessory |
mosomés from the same stage as is shown in fig. 2. 2b.—Plasm
from the same nuclei as the accessory chromosomes in 2a. The o
elements in columns 2a and 2b are from the same nucleus. The sam
holds true for columns 3a and 3b, 4a and 4b, 5a and 5b.

ao 3.—Spermatocyte nucleus, showing a later stage than

masomes from the same stage as fig. 3. See explanutien: of so
plate.

Fie. 4.—Spermatocyte nucleus at the period of maximum
and largest plasmasome growth, “> stained. 4a.—Ac

as fig. 4. See explanation of fig. 2, same plate.

Fic. 5.—Spermatocyte nucleus, showing the beginning oka ch
reorganization. 5a.—Accessory chromosomes from the same
fig. 5. 5b.—Plasmasomes from the same stage as fig. 5. See
tion of fig. 2, same plate.

PLATE 67

UNIVERSITY OF KANSAS SCIENCE BULLETIN

VOL. V, No. 20

S
~-9@@e008

S~@eerte

-@@6@S8n0
sea@=meogd

2& @g 98
Sg@@ged

$008 @O@
8 @®esjoe

as Ge

Bx he eo 0 F a eae
ari, * printrss be rN r. Bree m r
H : i ; ie, Se fae
e : Rees 5 Hele ae : SA er a a
: ale. pine yuan is Bh UE ath eae Pie Peete

.
. Z
i
‘
© # rs x
_ rs >"
;
: 2
> . ay r : ’
‘ é
. i
x * yi
~ r? ay Fe i
‘ . , . i x
~ , ©
‘ x c
xe i ee ¢ *
: x .
¥ +) As , e 2 ~
* =) A _" t
; Ana . ra he é Ks e
4 3, r
a5 ; : a) :
- 7 rs .
Coa * * . : ee
f 3 I * = 5 sey Cs
: ies ¥ ee i
‘ e . + i . 3
5 me
a - a J mn J >
- = “
- * . rE t | * »
“y $ ‘ f zp * iS
ES 3 V f at 5
ie eae - - < \ i 2 x ns ‘e
‘ a) ,
AG ae Ee,
v bes .
i = » :
’ , 3 i.
RA b, 2%
t : 2: ” E <
4 "
©
r : # :
* i
i C 3 es
- we % 3
4 , 4 y .
a ‘
\ : x :
s ¢ read
rine
2h F a y eh
2 &
fj - rie Pa sy
‘
e 7 =
~ o
+ . ‘ %
fe X - rf <

PLATE LXVIII.

Figs. 1 to 10, inclusive——The chromosome complex of the first sperma-
tocyte in metaphase. (x) Accessory chromosome.

Fig. 5.—An unusual condition of the first spermatocyte metaphase,
showing twelve elements. a-a.—Probably the disjoined halves of one of
the larger chromosomes which normally occurs in the ring.

Fic. 11.—Drawings of the accessory chromosome from cells of the
cyst from which fig. 6 was made, showing the constancy in size in this
element within a single cyst.

Fic. 12.—Drawings of the accessory chromosome from cells of the
cyst from which fig. 7 was made.

Fic. 13.—Drawings of the accessory chromosome from cells of the
cyst from which fig. 8 was made.

Fig. 14.—Drawings of the accessory chromosome from cells of the
cyst from which fig. 9 was made.

Fic. 15.—Drawings of the accessory chromosome from cells of the
cyst from which fig. 10 was made.

Fic. 16.—Lateral view of first spermatocyte in beginning anaphase.
(«) Accessory chromosome. The cell membrane is shown.

Fic. 17.—First spermatocyte spindle, in longitudinal section through
the accessory chromosome (#). (a) and (b) Chromosomes of the ring
at equal distances from (m), the central smal! chromosome. The acces-
sory chromosome lies outside of the ring, causing the asymmetry of the
spindle, as drawn.

Fic. 18.—Same view of a slightly later stage, showing the typical
irregularity in the movements of the accessory chromosome (a) and the
central small chromosome (m) with respect to those of the ring (a)
and (0b).

_ Fig. 19.—Later anaphase, showing the icieiniti in organization be-
tween the accessory chromosome (a) and the ordinary chromosomes.

Fics. 20, 21 and 22.—Later successive stages of the first spermatocyte
anaphase. (x) Accessory chromosome.

Figs. 23, 25 and 26.—First spermatocyte telophases, in which the acces-
sory chromosome (a) may still plainly be identified.

Fig. 24.—Telophase of first spermatocyte. (x) Unusual element,
which is probably an undivided accessory chromosome or a persistent
plasmasome.

PLATE 68

UNIVERSITY OF KANSAS SCIENCE BULLETIN

VOL. V, No. 20

: sQre = s
© se ie 8eg0° & g Sok

)

@ee iw” 000°

& 0 -

—.2 c =

@0 p00. ote, 000.
Wa ot"
©
on @ee86_

| . ,. *
STW oie
it
yee

rs

ras ew *
IO EET a
4

a rr, ee

5 pW

se reo
Sr

CC) BY Sean Ne camer so
paws SPLEEN oes

SES 5 6 f

vip hyurely
ay
+.

4% +
ri ee

chy ee

PLATE LXIX,

Figs. 1, 2 and 3.—Polar views of the single daughter groups of chro-
mosomes resulting from the first spermatocyte division. This stage is of
very brief duration following immediately that in figure 25, plate LXVIII.

Fig. 4.—A group of drawings of the accessory chromosome from stages
such as are shown in figures 1, 2 and 3.

Figs. 5, 6, 7 and 8.—Polar views of the second spermatocyte complex.
Note difference in arrangement of the chromosomes compared with those
of first spermatocytes.

Fig. 9.—Oblique view of a late first spermatocyte telophase. One whole
daughter group is shown and part of the other.

Fics. 10, 11 and 12.—Lateral view of second spermatocyte spindle in
metaphase. (a) Accessory chromosome. (m) Small central chromosome.

Fig. 13.—Lateral view of early anaphase, second spermatocyte.

Fic. 14.—Later stage of the same. (x) Accessory chromosome,

Fic, 15.—Same as figures 13 and 14, showing all of the chromosomes.
(x) Accessory chromosome.

Fig. 16.—Same as figure 14.

Fic. 17.—Slightly later stage, showing the beginnings of the formation
of a dividing cell membrane. Note that the daughter groups of chro-
mosomes are entirely coalesced. This is typical of these later stages.

Fig. 18.—Same as figure 14.

Fic. 19.—Group of accessory chromosomes drawn from cells such as
are shown in figure 18, showing the constancy in size of the lagging ele-
ment («).

Fics. 20, 21, 22, 23, 24 and 25.—Later stages in the second spermatocyte
division, showing the behavior of the accessory chromosome (a). Note
the dividing cell wall and persistent spindle fibers which identify, beyond

a doubt the two cells shown in each figure as daughter cells of the same

first spermatocyte.

Fics. 26 and 28.—Later siniie of the same after the formation of the
nuclear membranes. (x) Accessory chromosome.

Fic. 27.—Later stage of the same in which the spindle. fibers are
fading.

Fig. 29.—Still later stage in which the two chromatin masses are
breaking up and the ordinary chromatin appears granular. 9 Acces-
sory chromosome.

Fic. 30.—A group of drawings, showing the typical appearance of the
daughter nuclei of the second spermatocyte which have received the ac-
cessory chromosome in the second maturation division.

Fig. 31—A group of drawings showing the early development of the
spermatid which received the accessory chromosome in the second mat-
uration division.

i cures, als

PLATE LXXxX.

These photomicrographs were made with a Zeiss 2mm., 1.40 N. A.
apochromatic objective and No. 4 projection ocular, the dances bellows
being extended so as to place the sensitive plate at a distance of thirty-
one inches above the object. Illumination was furnished by Welsbach
mantle, the light passing through a Watson Parachromatic oil immersion
condenser with the diaphragm set at N. A. 1.0. The original magnification
is 1500 diameters which, in reproduction, has been reduced to 1000
diameters. Considerable detail has been lost in reproduction, especially
in figures 3, 7 and 8.

Fic. 1.—Polar view of spermatogonial metaphase, showing 21 chro-
mosomes. '

Fic. 2.—Same as figure 1.

Fic. 3.—Same as figure 1.

Fic. 4.—First spermatocyte prophase, in which the accessory chromo-
some is bent in the middle. This is the nearest approach to the spireme
condition found in Anasa.

Fic. 5.—Later prophase with the accessory extended and showing the .
longitudinal split.

Fic. 6.—Same stage as in figure 5. .

Fic. 7.—A portion of three cysts appears in this photograph. In the
lower one the last generations of spermatogonial chromosomes are seen
in several cells. At the upper right-hand corner are four cells in
synizesis. te

Fic. 8.—Parts of three cysts are included i in this, picture. The cells in
each are in some stage of synizesis. The presence of the accessory chro-
mosome on the periphery of the nucleus and of the plasmasome in the
synizetic knot is demonstrated in several cells. The stages shown in
figures 7 and 8 precede the ones represented in figures 4, 5 and 6.

’ Fig. 9.—Post-synizetic stage, in which the peripheral accessory and the
more central plasmasome are shown in several cells. This may be com-
pared with a similar condition in cells prepared according to the method
of Foot and Strobell, as shown in their figures 11 and 12, in order to
judge of the correctness of their belief that the smear method preserves
delicate detail better than sections. It may be pointed out also that there
is little liability to confuse either of the nucleolar bodies with karyosomes
resulting from concentrations of the chromatin elsewhere. Wilson’s slide
949b. Photos 1 to 9 from sections.

Fic. 10.—Chromosomes of the first spermatocyte metaphase, smear
preparation. The lowermost chromosome is the accessory. It is seen to
have one plane of division, while the others have more or less distinct
second planes indicated. In this cell the five lower chromosomes have
dried thinner and spread more than the upper six and show wrinkles and
other distortions. Wilson’s Woods Hole “x” slide.

Figs. 11-15.—Photographs from sections on Wilson’s slide 9496, show-
ing the accessory chromosome and plasmasome in the first spermatocyte
prophase at about the same stage as that of figure 9. It is difficult to

get several nuclei close together with the two elements in focus at the
same time, but most of these show at least two cells. Figure 11 is in
focus at the level of the upper part of the nucleus and shows the ac-
cessory extended as in figures 5 and 6.

Figs. 16-20.—Photographs from smears on Lefevre’s slide 30, showing
the late prohases of the first spermatocyte. The accessory is distinguished,
as Foot and Strobell point out, by having two halves lying side by side
without lateral extensions. There is no indication of a second plane of
division, as may easily be seen in figures 16 and 20. In figure 20 there
appear thirteen elements.

Fics. 21-26. These are from smears; and show the chromosomes of the
first spermatocyte in metaphase. The great difference in the size of the
chromosomes, due to the variation in the extent of spreading, upon drying
may be judged by comparing figure 22 with figures 23 and 10. In figure 23
there are twelve elements, one much lighter than the others, which may be
interpreted as an unsually resistant plasmasome which has persisted much
longer than common. In all of these the accessory may be identified both
by its position and by the single plane of cleavage indicated. It has been
injured in the cell represented in figure 23, where it is seen next to the
plasmasome with one corner cut sharply off.

Fics. 27 and 28.—Photographs of the first spermatocyte metaphase,
polar view, showing the typical arrangement of the chromosomes. The
‘accessory lies without the ring and the m-chromosome in the center.
These prints are from the same negative and show how it is possible to
vary the apparent size of structures even by printing. Wilson’s section.

Fics. 29-33.—Polar views of first spermatocyte metaphases from see-
tions. In order to get several cells to show together the focus was shifted
during exposure in making photograph 31 so that the outlines are not
sharp. In figure 29 the accessory does not show, being just out of focus
in the lower cell. From these illustrations it may be seen that while
there is a general agreement in the arrangement of the chromosomes in
the equatorial plate, it may be modified in details. Wilson’s sections.

SCIENCE BULLETIN, UNIVERSITY OF KANSAS PLATE 70
VOL. V, NO. 20

ae ae! aS)
pte Is hi =

a ie il

eK

ag

MCCLUNG, PHOTO

COCKAYNE-BOSTON

ae

+.
ae a

Me

+S ee

si

PLATE LXXI.

Fic. 1.—Lateral view of two second spermatocytes in early anaphase.
It is apparent in both of these that the accessory chromosome is already
in movement to one of the poles, in the upper cell to the right and in the
lower to the left. The accessory in the upper cell already has its longer
axis parallel to that of the spindle. Smear by Wilson.

Fic. 2.—Lateral view of first spermatocyte in metaphase or early ana-
phase. The accessory is the lowermost chromosome. All the elements are
clearly seen to be divided. This is a slightly earlier stage than the ones
of the second spermatocytes in figure 1 and should be compared with
them in order to see the difference in the behavior of the accessory in
the two divisions. Smear by Lefevre.

Fig. 3.—Mid-anaphase of the first spermatocyte, showing the full com-
plex of chromosomes in each daughter cell. The accessory is the lower-
most one in each group and is slightly removed from the others. Smear
by Wilson.

Fic. 4.—Late anaphase of the first spermatocyte mitosis. The acces-
sory lies to the right of each daughter group at the same level. Smear
by Lefevre.

Fic. 5.—Mid-anaphase of the first spermatocyte, oblique view. The
accessory to the right of each daughter group.

Fic. 6.—Mid-anaphase of the first spermatocyte anaphase, lateral view.
The accessory at the lower end of each daughter group.

Fic. 7.—Very late anaphase of the first. spermatocyte division, lateral
view, with the accessory chromosome of each group proximal to the
equatorial plate. Smear by Wilson.

Fic. 8.—Late anaphase of the first spermatocyte mitosis, showing the
divided accessory accompanying each daughter group of chromosomes.
Smear by Lefevre.

Fic. 9.—Like figure 8. Smear by Lefevre.

Fic. 10. Similar to figures 8 and 9.. Smear by Lefevre. This series
of anaphases of the first spermatocyte indicate, I think, that the accessory
is characterized by an independence of movement and position in the
group of chromosomes rather than by a “lagging” tendency. It does
not in any case, apparently, anticipate the poleward movement of the
other chromosomes, but takes part in the general division. On account
of its peripheral position and slighter relation to the plate of chromosomes
it is crowded back and therefore lies below the daughter mass of chro-
matin. Its position in the metaphase of the second spermatocyte depends
upon the degree to which it is crowded aside in the telophase of the first.
If it maintains its peripheral position at the level of the other chromo-
somes it is then an eccentric element in the metaphase of the second .

‘supermatocyte. If, however, it lies beneath the other chromosomes in

the telophase it may become included within the ring of chromosomes in
the next metaphase.

Fic. 11.—Polar view of second spermatocyte metaphase. In the lower
cell three chromosomes are shown in lateral view. The difference in the
arrangement of the chromosomes, as compared with the first spermatocyte,
is striking-and apparent. Section by Wilson.

Fic. 12.—Similar to figure 11. Compared with the cell shown in the
preceding figure it is seen that the arrangement of the chromosomes is
different and that both are unlike that of the first spermatocyte where
the m-chromosome practically always lies within a ring of larger chromo-
somes. The upper cell in figure 11 shows how close the resemblance to
the first spermatocyte may be, on the contrary. Section by Wilson.

Fics. 13-27.—Photomicrographs of smears, mostly from slides by Wil-
son, showing the behavior of the accessory in the second spermatocyte. It
will appear from these, I think, that while the accessory may lie near
the equatorial plate in the early and mid anaphases, when actual division
of the cell has occurred the position of the accessory in only one of the
two daughter cells may be determined. Of particular importance in this
connection is figure 15 where numerous cells in telophase are in focus at
one time. Each one of these demonstrates beyond question the unilateral
movement of the accessory. This area is but a small part of a much
larger one on the slide, where more than a hundred similar cells show,
without exception, the same undivided condition of the unpaired element.
The lagging chromosome is recognized by all to be the accessory. It is
seen here in all the divided cells to pass without division into but one of
the two daughter derivatives. In comparing these figures with those of
Foot and Strobell it should be borne in mind that these cells are without

- question completely divided, that numbers of them are shown together,

and that the whole cell body is represented, so that there can be no ques-
tion of the relation between the members of the pairs of chromosome
groups. Their figures, on the contrary, are of single cells, mostly in early
or mid anaphases, in which the cell body does not appear. The con-
sistent behavior of the accessory chromosome in the different cell gen-
erations is worthy of attention also. Wherever it is found it tends to
isolate itself somewhat from the rest of the chromosomes and to behave
with some independence. The argument of Foot and Strobell is based
on conformity to one type on the part of the accessory as well as on that
of the ordinary chromosomes. The figures on plate LX XI will demonstrate,
I believe, that the accessory does not conform to’ the processes of the
other chromosomes. Its isolated position in one of the two daughter
spermatids shown in figures 26 and 27 is just as apparent as it is in
each of the two daughter spermatocytes shown in figures 3-10. This
characteristic may also be seen in figures 20, 21 and 25. Certain cells
in the stages represented in figures 14, 17, 22, 23 may suggest that the
accessory chromosome might later divide, and it is upon such as these
that Foot and Strobell base much of their argument, but later and more
decisive stages demonstrate that it does not do so and this positive evi-
dence has much more value than the presumptive. It is my belief from
a careful study of the material and the photographs of Foot and Strobell
that not one instance which they figure is an indubitable case of a divided
accessory in the second spermatocyte. I do not believe that the in-
dentations in outline, or the light places in the middle of the accessory,
have any value as an indication of probable division, for such effects must
inevitably occur to delicate structures like the chromosomes in smearing.
I would submit that if such evidence is to be used, the chromosomes of the
upper group in the anaphase of the second spermatocyte represented in
figure 37, plate III of Foot and Strobell, are much more certainly divided
than are any of the accessories for which they claim divisions in figures
26-46 of the same plate,

4
Lip,
Ve)

SCIENCE BULLETIN, UNIVERSITY OF KANSAS
VOL. V, NO. 20

PLATE

MCCLUNG, PHOTO

Prod

COCKAYNE-BOSTON

Univer

sity of Kansas science

bulletin ENB

<

Tp pay alee

De

_

a

Tashshaess
ESrostaceses

get

im a deo bene ts

<4)
eechoasas Forgtasae: ‘
iets rhe) paceet reseee) : . ‘
« A 29
rs

Rae oo gS

ri

mg Le > . ¥.

3 ease eee :
ashes Se ae
eee

ter °

. “ ie
- =

M5
“tis
4

iectuseeetitePe-eoe

TOT Shy She Sete Tte Taran se

i

ni
Hate

paet ed

£
tt
ibe fe
felitats

eet
it

oy . ee y Y!
ie : patiatt
bE eects cos

CTE POT. TT

text
eed

tet Set,
ie

a
riesete.
mii serrbnas,

i}
43,
ae

Ty
a

ee oe

f co

€

ae
tosis
3
- ey iss
SER Tate

S ee se
Sh a alt 3.

hte
eutty
et 4

Hi

ih»

ui

ae :
= patted =
: ees estes

x.