BIOLOGICAL BULLETIN

OF THE

flOannc Biological laboratory

\VIH>I>S H' 'i.i-:. MASS.

2£Mtonal Statt

! ' . ( INKLIM

• ,i i iRGl 1 . M'" '

I . II. M i • K - . \ s
\\'. M. \\'ni i i i k
I R. \\'ii.""\

I K \NK l\. I II I II / •'•• Ut

X'oi.iMi: XIA I.

WOODS HOLE. MASS.

JAM Al<\ TO JIN

LANCASTER PRESS. INC.
LANCASTER. PA.

Contents of Volume XLVI

No. i. JAM AKY, 1924

K<>i'i ' . sn i \v Studies on the Influence of Inanition on

the Drcel'i/Jinent and the Duration of Life in Inserts ..... I
KM K. (". II. .1 Neu Field Method of Investigating '.he

I/ydrnlro/)isms "'. ^ebrates ........... ; =>

V >. 2. I'l I;KI \KY. i<)24

M\M, 5. O., A\I» I'i -i H, I.. (". M, Respoi

in A i' .............................. 55

M M AK I ill K. |MH\ \V. An /'.A fal .Vm/v and a

/'>: :! Interpretation "Inflation ami 1\

latt-il Mud if ' »: Embryos . '•<•

ALLEN, I.U..\K. / i Short Reproductive C)

in . 1 >:i"!"nt(i inihei His .............. . ........ v ^

H< i\n\, I |IK« 'Ki K- '. /' Rhal'tii i

\' i. ;. M \K( II. IM.' |

YOCOM, 1 1 \KK\ H. I :\ in i> :!a-

KiNV'ii'- Bki • i I ). /• Protoplasmic .1

/;/ Relati -dity and Environment in

. 1 rcclla f>o!yf)ora I "' >

\M-<>\. TiiiKiM\\ ('. 77n Attachment of Oyster Larva ij,

No. 4. Ark 1 1 . i'iJ4

Mi i/. ("n\-. \V., AMI NONIDEZ, J"-i I . H

the Xuclens and Chronio^<nr.<^ l^.trin^ .^f>, »!:•.

in the Robber Fly, Lasio{to\^on Inrittatus i - ;

l'\. K\KI>, CII\KI i -. The .^nscepti/>ility of Cells to Radium

Radiations i < ^

CLEVEI \\i>, I.. K. The Physiological and Symhio:;, Rela-
tionships Hctiiccn the Intestinal Protozoa <•' J'ermi/es ami
then Host, icith Speiial Referem e !<> Reticnliteri- ivi-

f>es Kollar . . 17*

iii

IV COMI-AI^ <>F VOLUME XI. VI

No. 5. MAY, i (^4

CLEVELAND, L. R. The Physiological and Symbiotic Rela-
tionships between the Intestinal Protozoa of Termites and
their Host, with Special Reference to Reticnlitermes
flavipes Kollar 203

KINDRED, JAMES E. The Cellular Elements in the Peri-
visceral Fluid of Echinoderms 228

GOODRICH, HUBERT E. Cell Behavior in Tissue Cultures.. 252

No. 6. JUNE, 1924

TURNER, C. L. Studies on the Secondary Sexual Characters
of Crayfishes, i. Male Secondary Sexual Characters
in females of Cambarus propinquus 263

HEILBRUNN, L. Y. The Surface Tension Theory of Mem-
brane Elevation 277

HOADLEY, LEIGH. The Independent Differentiation of Iso-
lated Chick Priniordia in Chorio-allantoic Grafts 281

MANX, MARGARET C. Cytological Changes in Unfertilized

Tubal Eggs of the Rat 316

Vol. XLVI January, 1924 Xo. I

BIOLOGICAL BULLETIN

STUDIES ON THE INFLUENCE OF INANITION ON

THE DEVELOPMENT AND THE DURATION

OF LIFE IN INSECTS.

STEFAN KOPEC.

! I'.-llli IK FOR A<;RICeLTCRAL RESEARCH. PlI.UVY. i

WITH 3 TABLES.

CONTENTS.

i I IM • mtliH ii. • ol ::. .1:. it ion on the duration of separate dcvclnpiin-nt.il -t. .

2. Inanition anil it- tx-arinK <>n the physiology of insect metarnoi;

<. Inaiiiiii.ii and certain problems of growth.. .

.}. A'l.i]>i.iti"ii of organisms to starvation. Some remarks on tin- |n..M.. in »\

death .......................... i ^

-iiiimiary .............................

Hit experiments were performed on the caterpillars oi tin-
ii'iiiniun moth Lymantria dispar I.. Each e.xprrinu-nt ri.n-i-ifi|
"I ~iAcr.il separate broods. The caterpillars <lcii\in^ It "in mic
\\i-re divided in two parts and put in t\\<> j.ir-. "iic p.irt
suliji-cted to starvation, the other, d.iiK ted. was u-cd .1- a
control. I lie inanition applied in the-r >nnlic- u.t- tnt.il but
intcnninciii. the caterpillars being totally de|>ri\ed < .|" t'.,. ,<1 dur-
ing (crt.iiu days and fed ad libitum during the rein. lining;. Th«-
distribution of these clays was different in different -eric- of ex-
periments, .1- may easily be -ecu from t.ible- i to 3. The obtained
pupa- beinu kept e.nli -ep.ir.iiely. it \\.i- p«'--iblc to -ludy the
iiillui'iuc M| -i.irx.itidii imt onl\ on tlu- duration of the stage "I
i .iterpillars, but also on that of e.u h rhr\-.ili- .md e.u I) moth.

hi the chief e\l)criiucllt- '-eric- .1 to /•' the whole material
consisted of u.Vi -tarxeil caterpillar-, and >>4^ -pccinicn- lor
control; 547 |>up.e \\crc obtaineil from the caterpillar- -ubjei ted
to -tarxation and 70^ from control lar\a-; the\ prodin'i-il 470

"-tar\ed" and <>N.^ comn.1 moth- a'f. Table I Phe increased

; Papei iii'in tin- 1 .ili..i.ttnt\- ni l-;\priinirin:il M.irphnli>i;y.. i i \l I n-

\ .:.' Polonai . Ruralt •: Pulm \'t>\. i. i

i
I

-Ill \\ Kol'l •( .

a

CS

-<

-

Si

"o
"5

O

Control.

So

> «

OOO O t

O 00 <-<

O IO IO O X X

~^ O -t T O» oo

Q P) PI

to r^ o

O H O (*•

c o t— 10

Ol X O

IO O ro

-~ 'o £"

— "O t 01 Tt 1-1

oo 0) c' C C~ -

O 10 M Ov O O

-T — to 01 t^ IO
M H O fO W IH

X O- X
"O 10 X

- r o

2 2- °°

ro w O

•

a> , _j CS .

•J *" > C '~

r^ r~- O O

M IO O» O

r — _

to t- 11 O "t o

-' : O IO o| ;
O 01 X O >O IO
Ol oo 01 to OO oo

M 01 IO

t^ C_ 10

O X 01

O M to

x 10 r

1 1 1

t— 01 O IO

Oi r- M x
O X Tt Tt

1 1 1

M x X
w oo C*

1 1 1

I- TO O O *t
IO M O1 01 IO ^O
NO O> X X I—

1 1 1 1 1 1

01 10 1^- 1^- oo X

o to -to c

O 1— X -OX

1 1 1

IO IO -t

- c -t

t^ 1^. x

i i

- ~! — '

— Ov Ov

Starved.

|l

01 0 11 10

r— w oo

O X O 01 O -i

X O oo 1^- _ ^*

X - 01

-CO)

O oo I— X

X O O ON
O O\ ON O

r*» r°"» oo

x ON •+

0 0 M

x 10 x x r*» to

10 to o >o o> O>
O I— O O> X X
O "t TJ- to O -t

cox

OOO M

Ol 01 00

X X 10

01 — r

Tt t OO

1

lo'lJIi

3 «>~~

IO X M N

oo 10 x 10

ON w i- X

III

ro ON 01 X
Tt O O oo

N t— "t
X to O
oo to to

~ " "

to O oo
IO 1— O
I>- ff IO

i O1 X •-• O ^O

O t— 01 Ov O r—

i T T i

O ^^ f^l C ^" *O
^^ ^H cs c*i ^* r^*

01 OX oo 1O IO

X C* "} O1 X O1
(^ to O C> O O

i i V i

C rO O 01 OO t^

oo o O- x r~ Tt

,r -f -) -/- -^ -r

OOOO M
X t^ 01
PI OO Tf

1 1 1

O — Ol

to oo 1^

: ; '

10^ —

1 1 1
O> to Ov
O> 10 X

01 OO 01

"b

•b

Control.

0 0

> u

<; a

t— — X to

^T O OO

M M W O W t»-

1-1 -t "t Tf IO —

N O X

Tt 01 r~

oo 1^- O 6

X O I- oo
ro to oo Tt

C> X C>
OO OO OO

O Ov ^t" ^ O r*-

a O -TOO -^f

T^- IO ^f "^ ^t" ^"

O^ 01 10 i^* ON C^

O> to t I— — rO

o - o

x - -t

OOO to

•* - o

Tt to 10

-

*• .i «! 3 K

jplfJJ

— ir _ -,

M o t- r-

IO IO t 10

1 1 1 1
01 c C *t

C - X X

r— to 1-1

Tt 10 10

i 0 to
£"*- oo r**

CS fN M

"-. ^ - !-!-!-

f** oc M O "1 ro
1010*0101010

1 1 '

OC f^* ^^ iO ^^ ro
^O ^^ ^^ ro *•* f**

*t O r** O O t""~

to O* ^2 oo O r*^
•t t IO IO IO IO

III 1

X >- — O> Ox oo

01 Tf Tt Tt M OO

PI r~ o

IO Ol Tl-
tO t IO

1 1 1

01 — IO

10 0| ,O
01 O IO
IO O to

IO IO O

O TJ- «o

Starved.

u .

O t)

> M

t— O O oo

ro O N

oc i"- (N ro O r-

01 10 10 O — O

O O 10

r^ to x

to O O t—

PI OO OO PI

to O\ O

\O ^O "t" ^ O *~*

O '"O '-' O O C\

01 n 01 IN O X
— Ov *t oo OO O
01 11 Ol 01 N M

Tt IO 01
tO OO M

oo 01 01
1^ X t^

- • — 3 •

— - r i -

O\ N OO M

till

O 01 IO IO

X O 1-1 01

— "^ -I •]

ro O O
^f >O IO
1 1 >
CS IO "-•

oj Q\ O\

10 10 lo r^ ^ O

fN O- *O I1^- O Ov

i 7s T 7^ T5 T5

O ro o* r» to o

-rf r^. «-o -rf 10 O

•"O fS N CS C< CS

Tt 01 M O X to

r~ 1-1 c i~- a> ov

01 01 OO 01 01 —

1 1 1 1 1 1

r<5 — to IO IO IO

O r— x x Tt to

r- x M

O O ro

? T

M X O

N - O

N O 1-
t^ — IO

? " ?

to x to
O 10 to

Number of
Moths.

Control.

•6 6

X to Tf M
M « M OO

N M O

CN t^» 04 O O W

00 N O Tt N .O

Tt O O

Tt Tt Tt

•,P,P

X IO 01 O

— , _l k-l <VV

M OJ ^f

N OO Tf Tf O O\

N «O to OO 01 O\

O O oo
— 11

Tt O *t

Starved.

. '6 6

— i — Tt M
M tH M IH

^ 00 O

CO O O W rt 00

x 01 r~ o O x

o o TT

t^> O X

•PJ>

t^ ON N O

oo c^ oo

M

^ ^f o ^ N r-

O ro O M M O>

M n — n

O ro O

O X t-

^c
o«

•2 -

Control.

•6 6

10 10 O X

-* -: ^~

0) « -

O\00 <s O O »0

M M IH M

O\ 10 O -t ro O

TT M O

O O Tt

-,p,p

l-< O 10 10

^t 10 r^-

rs oo O O O -•

oo 10 10 Tf 01 O

t** O ro

*- «-»

Tt X -t

— —

Starved.

'6 6

't Pi O oo

M M M M

Tf X O

O\ >~» O M oo Cs

H M M H

X oo ON Ol Ol O

|-| —MM

00 O O

X O O

•ee

O O\ to O

O N O

\o oo O w 10 O*

M OO O N 01 01

OTtO

C X X

—

tpi

•poojg

0

c
_g

-.

M OO Tt M

N O\ M

1-1

*"f >O O 1"* OC Ov

X O> O — 0) 00
M 01 OO 00 oo 00

O 01 10
t -t Tt

O t- X

Tt t -r

C | "^

M ,

c -r M
•- ^ c

111

fill

~% i, ' HI

£ x + 2
C " 1 "5

C 0

.2 ^

4~* 4-1

|l

Hi

O "O --j

fc " f" t"

c JS

•~ a «'
_t ca -a

° 1
ll

rt T3

> u
u • —

C/i •

u

Hi >> '

'£ cd i

t -a ^

Series D. Starvation lasting one
day everv second day ( -| h
- + -)

Series !•'. Starvation lasting two
davs t-vcrv second day ( + -
+ --)

§|

^ +
~ .

E ~

^ 'r
cS . ^~-

4^ ^. 1
* 1 I

<o K5 -J.

^ ^, i^

t-l 1

•h )M .MM |.»l\-.| JO l.i-ll< )

•sX3o uiojj

•.inn iiu|
l|ii(iui )>i:| .up .>.>tlis'

M \l -Inl'Ml \l AND DURATION" OF I.IFK IX INsl-XTS. 3

mortality of the -tarved caterpillars ought not to be ascribed
first of all to the emaciating influence of inanition, the chief cause
of tin- mortality i- tin- moult. Directly after moulting tin- animal
take- very much food, as during these processes it dor- not eat
at all and only digests its own substances. Then-fore when the
fa-tin^; day falls on the period succeeding moult tin- caterpillars
an- often unable to resist starvation. The mortality of chry-alid-.
on the contrary, was identical in starved and in control specimens.
The willow leaves on which the caterpillars were fed were al\va\ -
of the same variety and of the same freshness. The conditions
of space, light and moisture were identical in all jars during tin
whole t me of observations and the temperature was from 16°
to 19° ('. Food was put in great superfluity into control jars
e\ery morning and into those containing starved material during
iht feeding days and it preserved its freshness the whole day.
All experiments were performed during one season (1920).

i. Tin-: iNFi.t'KNCK OF IN AM i ION ON i HI: Di RATION OF SKPA-
i< \ 1 1 1 >i \ i i « i r\i i N i A i. Si A<;I-:S.

lor the experiments of series A caterpillars belonging to four
dillt-n-nt broods were used, from I to o hours after hatching,
These caterpillars wen- deprived of food every second day, in
the remaining days they were fed ad libitum. Starvation la-iiiu
one day and applied every second day may be designated by ilu-

-\mbol +• + -H • Table II. records the limits of individual

fluctuations in the duration of life of the caterpillars and chrysa-
lid- belonging to each brood, as well as the average duration of
lite in -i arved and in control specimens of the t\\o -e\e-. Table
III. finally shows the average differences of the duration of
larval and of pupal life of individual- deprived of fond for each
brood separately, calculated in penvir . t the average dura-
tion of the larval oj of the pupal stage- "I control specimens of
the same brood and the same sex. From the-e tables we see that
the life of caterpillars subjected io-n.!i -iai \ation was considera-
bly prolonged. In all lots \\ithoiii exception the longest-lived
control caterpillars underwent pupation earlier than theshortest-
lixed starved specimen- < f tin- -aim- brood and sex. (Cf. Table
I I . Series A). It we take the a\ era^e • >\ the averages of all broods
of >eries .1, a quantity which I -hall call "average of brood-."

STEFAX KOPI-:C.

a

Number of Days of Pupal Life.

o
c»

Control.

<t|

X X CN O

f « O

O o M o r^ CN

O t^ O 1^ ~5 O

O t^ ro

CN Tf O

Ox O O O
M CN CN CN

coo

CN CN CN

n - O O O O

CN CN CN <N CN CN

M O O M 6 H

(N (N CN rN CN CN

i Ox O

CN « CN

O O O

CN CN CN

*? .i'3 (J . IB

i CN — CN
CN CN CN CN
1 1 1
t— O t— Ox

CN -^ CN
CN CN CN

1 1

X O X

CN CN r-i « r-i —

CN CN CN CN CN CN

1

O 0 Ox O 0 Ox

CN — — ro CN -O
CN CN CN CN CN CN
1

O O O « Ox O

CN CN CN CN *• CN

CN — —
CN CN CN

1 1

O Ox Ox

CN n n

CN — —
CN CN CN
1 .

Ox O O

H- CN CN

Starved.

ifil

C C .CN -^

O ro O

X X O CN t— ro

00 ro r}- m O ro

ro O t--

11 in 10

M M M M

^™j in ^2

Ox Ox O1 Ox Ox Ox

Ox Ox X O O1 f^*

^ ro T

"-. -rxc

Him

CN CN ro CN

t~- i^- X

ro ro i O O O
CN CN CN CN CN CN
1 1 1 1 1 1

O O vC X OX

CN C O CN 11 C
CN CN CN CN CN CN
1 1 1
t^ X O Ox xO 1"

in in sj

^ ^ tx

I

O 0 C -

1 1

1 1 1
ro CN ro

~f ^ in

"b

"b

Control.

<' U M
O a

CN .T X CN

r^ O CJ

n CN CN ^- O X

O X X O Ox CN

X rtx

•"* ro x

S r? c? r?

CN CN CN

CN CN CN CN CN CN

CN CN ro ro CN ro
CN CN fN CN CN CN

CN CN CO
CN CN CN

CN ro r^

CN rN "1

•" ..!. "^ A i to

'-I— -n - i rt C
.E ° c3r" 3 o

CN CN CN CN

Ox CN CN n
*- <N CN CN

CN CN CN

1 1 1

M X O

CN CN CN CN CN CN

1 1 1 1 1 1

CN CN CN CN CN CN

1 1 1 i 1 1
CN CN ro 1 H CN
r-1 CN CN CN CN CN

CN CN CN'
1 1 1
CN CN CN
CN CN CN

CN CN CN

1 1 1

« ro ro

CN CN CN

Starved.

^£!

t— o O in

CN O X

CN in \c >n CN M

ro o mo o -c

O O ro

X O T

t C O vC

Ox Ox t^

CN CN n « ro CN
N N CN CN CN CN

M CN n CN CN n

CN CN CN CN CN CN

x x r^

O> Ox Ox

X i "p 1

•0 X 0 Ox

1 1 1 1

>i m CN ^

n n O
CN CN CN

1 1 1

^D t** ^^

CN CN CN CN CN CN
III II
CN 11 11 O CN M
0« W CN CN CN CN

CN rO CN ro -r r-g
CN CN CN CN CN CN

0 i ^Moi

-i IN -i ': -i ";

i-" Ox Ox
CN n n
1 1

\C I— O

MOM
CN CN CN

1 1 1

r- x r-

w

5

£ .

0 S

£ o

si

""* •*

•_ Cl

00

>• 5

Si

o

1*

5

0

Control.

' 1 0

O O CN CN

O mvO

>O lOxQ ^-O «

M n-CN in w CN

in rooo

oo r^ o

O OC O X

X CN M

^~ in in

Ox « O O O i-i
*- CN CN CN CN CN

Ox Ox X O O M
M « « CN CN CN

Ox Ox Ox

— CN —

Him

x in x Ox

V T"? V

Ox X -a-

Tf m TJ-

^f Tf Tf

O r"- CN 1 CN xO
CN CN CN CN CN CN

1
Ox 0 O 0 0 Ox

« CN O ro m ro
CN CN CN CN CN CN
1 1
1^- X >O X X O

ro CN n
CN CN CN
1 1

X X Ox

^f Tf O
N CN CN i

1 1

X O X .
n CN n

Starved.

>iS

< «J rt

c in in o

O ro 0

n O "St CN OO ro

x c m x CN M

O O 11

m c CN

X X \O O

ro X t—

xO xO ^D

X O Ox O ro O

CN ^O CN CN CN CN

o x Ox x o in

(N CN CN (N ro ro

t-- t- r—

ro ro ro

xO x ro

ro ro ro

yj , " i

5 ~>E~~

M l-l M 00

x x x r—
1 1 1 1

m in 11 in

t— r- r— t—

t^ CN 11

1 1 1

O ro t

*O ^G "O

I*** *o xc n x in
ro ro ro ro CN ro

1 1 1 1 1 1

ro O *1~ ro <N CN
CN CN CN CN CN CN

Ox O Ox Ox 00 ro
CN ro r>) ro ro rf

1 1 1 1

xO O ro ro ro 1^

CN CN CN CN CN CN

O m x

1 1
O X X
ro CN M

Ox Ox 1^
rp ro

ro l^ Ox

ro ro CN

•b

Control.

J I 4)

.* u. t(J

< " rt

X O r^ ro

00 ^J" ^~

x x <j r-» m ao

O x o m 11 11

•* X CO

^- X «

ro !n *o in

Tj" rj- IT)

XX Ox Ox Ox Ox

x r» x x Ox O

XXX

X O X

M CN —

£ o c'H— s o

Ox t""" Ox Ox
1 1 1 1

CN CN "I" ro
Tl- rf •-!• Tt-

10 00 O

O O « i CN O
CN CN CN CN CN CN
1 1 1 1 1 1

XXX Ox X Ox

CN Ox Ox O CN CN
CN n n CN CN CN
1 1 1 1 1 I

O O t— l~- !•» X

Ox O M
n CN CN
1 1 1
X X t^

O '-J- Ox
CN CN M
1 1 1
X Ox X

t"= Starved.
Starvation Applied to: . :

> i &

<(, o cd

r— o oo in

M 10 ro

CN ^ in ro ^T ro

O O O >n ro O

ro CN ro

Ox O -

O t"** *i ^"
r-~ vo t** xO

M fN Q\

^O ^O ^O

CN CN CN ro i-i ro
CN CN CN M CN CN

t— o *t -t O Ox

CN CN CN CN CN CN

CN ro ro

X ro r~
CN ro N

ro . 73 i
.ti •- 2 o ' w

»J ^

X t— X O

i I i T

CN ro ro CN

\^"5 \O \rt \^i

O O\ r~

r^ o t~-

1 1 1

in in o

ro •* X X CN CN

CN CN CN CN CN rO
II 1 1
11 M O " O M

-, -1 -1 -1 "1 -1

O r^- in in ro ro

ro <N CN CN ro ro

I 1 1 1 1 1

-i ": -r -i -i -:

CN x m

rororo

r- m\o

CN CN CN

X T Ox
rororp

CN CN —
CN ro M

qo«3 01 5

a

M ro 't «

M

CN Ox CN

11

•*t mo t- x Ox

ro ro ro ro ro ro

X Ox O M CN ro
CN CN ro ro ro ro

o CN m

O I- X

--^ £

c 1 fe

— 1 ™

ill

ji|

rt ._ j c
CA e 'C

C 3

>. — .
^> w

^ cj O«

a ._
o -p
If

+j

£ '-c

_t 0 2

a
•a

c ^

rt I
~- 1

1 +

*-> N .
^ I^>

J-S

O *"
_fc o

a
•a

1?

•I '
~ i
|±

rt K*%
> «

B^

o
u

CJ w
X

!
_fe S

en
X

a

-o

'% 1

•3 1

§±
II

-i
itj
^ ^*

U!S

t?

p 1

II

A

j:

rt ~
W —

ft,'*1
C! ^*

•»-" o

•s33a tuojj
SuupiBtj oouiy

•ouo inq

1[I101U 1ST3[ Ot^ OJ

iJl-.YI-.l oI'Ml.x I AND DURATION OF I. IKK IN INSKi PS.

'.. ^1111 Silli i- III' 1 ' M.mlt Hilt < >llr.

-j.

- - 1

,. 1
^'_-- -

|?| +

m
& t

•spoojg

JO aSeiOAV

-i
<n

:
«^

»~

c
-^

xs

-7-

I>1

^

•o
5

o
B

Nninlier !<<•-

leu !l

•i Hinii'l.

•
—

t^.

r
•r

t^

-r

t^

Tj-

X

0

r-

M

S

IO

d
2

r-

—

00 u~
in 1^

-T

oo'

x

oo
p<

"7

»n

~5

o
•o
2

IO -^ C N

- -f 'f - ui

U-. 3C — ' ^

a

~i

t^«

u _

\ -_

'< 2-~

; n

tr

«ug

JO 3!lTM.>AY

r*) v^

"•; <S
>O '

t~.

H
N

1*1
<•>

0
C)

<^

a

N

;

]:

x~:3

r~.

PO

t-»

gQ

CO
t^.
~i

Y?

t^-

IN

M
10

PI

«

a
P)

-1

-T

^ t^
ui — '

C ' O1

C>

»

o"

~5

q\
p<
rn

•*

IO

PI

O
T

N 1^

6- 6,

m ao

T
ad

r
[^

q

PI
2

N

C
PI

.

JO

-•"•

fl

flf &

:-. £
SL

5«"

1

1 I

X

a

»AV

"J "". C «x

O O «i DC
> "n

«J

oo

*

*

•)
"5

n i/>

i ' £

00

-c

-
1^

CO

2

10

PI

1-

PI

~)

o

t—

2

P«

O

V)

Ol

If)

t

0

X

•<f
>n

IO

ad

\f.

~J

f

p«
2

T

e
2

1

•t

2

10

ad

'T

00

i

0
H

"7
po
2

q
<•«
<o

<q

Oi

o

0

T
Tt

10

1

0

PI

O ~5

<o 4
t —

m t^

p^ <:'

*?

PI

•r

0

T

oo
PI

q

o

m

"?

0
M

S

in

PO

q
ad

Starvation 1 >.iy
( + + - + + -)•

Numl>or Referring to Ivach
Brood.

25

a

»AV

S

00

**1

"1
^

•O '
f~

">

PI

j

rw*
fM

"3

>TI

•f

*^
ex

10

o

$

li

IO

«n

«?

P^

00
•f

r»

10

0

p«
od

0

ft

r% ^r

00 30

M i n

0

4

t^

0

4
i-

°.

90

s

!•* r-

f "i

_ ~

J—

p«

q
>o

r»

6

>o

6

ao

1

5.

U-,

6*
2

q

r^l

l1^
»0

IO

10

PO

-•

f
">

o

00*

"9

PO

t

PO

«r

to

q\

PO
r»

N
"-.

rimcnt Begun Since Il.tt'.liinj; ii

§**ji

-J

-_-

9 ^£^S

•r - ^J

••POOJH
jo OJIIU.JAV ri *•»

•^l

N
P|

Q

kq
^

—
•*

0-53

&M|

•fill

BE

^S

2

n

s

q

M

M

o

o

«?

s

—

O

—

c-

10

o

PI

O

e

~5

P*5
r^

fj

w

P4
PO

•<r

i

PI

^

PO

~5

<««

ci

PI

Cf
ad

IO

—
r^

1O
t-^.

ao

in

PI

O

J^ss

ca-rj

«"?^-
Sf~!

,- "is

l-^fc+s

t = t i-

"!-rii

> •" *•

JO . .• '• i . \ V

!•»

N

"

<n

•^

=

*^1

N

6

";
£

Number

rriiiK to
Earh Brood.

H

*9
«j

!-•

un

PI

t

q

po

o

O

s

V

t^.
m

«y
in

r»

PO

10

4

—

o

00

H

4
o

n

N

m

m

o

o

O

p«

i-

t

I—

r^

PO

O

r»

55

09

-

•*

O

"?

r

0

f-

"

2

T
j

10

>o

•c

ad
O

lP.P

6 6

.P^3

6 6

.p^

6 6

e , _ , ^ ,

A'. lion

nl llli' i>l tllf "~l.il \

iui|i;i-. iii tier cent, "i ili<
luration of \>\\\<
al life in O)iiti»l -|
11 n 'iis of the same brood
.UK! set.

^ S 5i

« -

V- " >-

~.E >-•«

-• ,t — Cl

•58 -
f I s S a

?o:^-

ic — '— tr ^

^!8l=s

^ u • •*« ^ •

j; -; - ^ _L _

<I •-> r- •— .

: =
: u ? ca a

3 8 * a - S ""

~. --- = '•--

it r — •- - — x

Bl8s2°§

- = -=_--" 5

£-° °-s B«g

~ y _ - "3

a--sliis<§i-
e<8eSfi9«i

. . -:-_=_--
> vi: — td <a •— > . —

x- •- - =

• — — - ^ -
— -^: -C-T= -

: . - — , — ^ =

6 STEFAN KOPEC.

we obtain for male specimens a prolongation of larval life of
52.7 per cent, for the female of 61.5 per cent. O". Table III.
As the processes of pupation were checked every 24 hours and the
caterpillars were taken for the experiments from i to 6 hours
after their hatching from eggs, the error concerning duration of
the larval stage could have in no case exceeded 30 hours.

The duration of the pupal stage is influenced by the starvation
of caterpillars in an essentially different manner, viz, the chry-
salids which have developed from starved caterpillars undergo
transformation into adult moths far earlier than control pupa?,
but the abbreviation of the pupal stages is smaller than the pro-
longation ot larval life. This abbreviation amounted in the
"averages of broods" in males to 31.0 per cent, in females to
44.5 per cent, of the average duration of pupal stage of control
specimens. Pupation having been controlled every 24 hours,
emergence of moths every 12 hours, the error in estimation of
the duration of the pupal stage could not exceed 36 hours.

The moths of Lymantria dispar L. do not take any food in
their imaginal stage. Control males lived as moths in separate
lots from I to 8, from 5 to 8, from 2 to 8 and from I to 8 days
and the " starved" specimens from 4 to 5, from 5 to ir, from 2 to
8 and from I to 6 days. Normal females lived from 3 to 13,
from i to io, from i to 12 and from 2 to 14 days, those derived
from starved caterpillars from 5 to 6, from 3 to io, from 7 to io
and from 6 to io days. Comparing the mean values of the dura-
tion of life of the normal and of the "starved" moths we obtain
in most cases a certain prolongation of life of the moths derived
from starved caterpillars ( + ) though rarely a certain abbrevia-
tion ( — ). This prolongation or abbreviation calculated in days
amounts in separate broods in the males to -f 0.3, + 2.5, + O.I
and -o.i days, in the females to -2.4, + 0.3, + 2.8 and -f 2.1
days. We see that starvation of caterpillars has no distinct effect
on the duration ot the imaginal stage. Emergence and death of
moths having been checked every 12 hours, the error in the esti-
mation may attain only 24 hours. The moths of the two se\e>
experimented upon both control and "starved" have never been
allowed to mate.

If we take into consideration the behavior of the "starved"
moths as wrell as the fact that the abbreviation of the pupal

!>1 \ l.l.nl'MI.N I AND I >I RATION OF LIFE IX INSE< fS. ~

period was smaller than the prolongation of the larval Mage, we
must draw the concln-ion that total deprivation of food of cat-
erpillars e\ ery -ccond day has^ considerable po-itive effect on
the total duration of their life from hatching until death. The
prolongation of i In- whole developmental period amount- here
in tin- average of broods in males to 16.5 and in female- to 20.4
days, or in penvntavie-; relative to the duration of life of control
-peeimeii- in males to 24.2 per cent, and in females to JO.O per
• «-nt.

The .ibove results find complete confirmation in my farther
lomp.tr.itive experiments in which inanition of various intensity
\\.i- applied (Aeries C, D and E). In these experiment- cater-
pillar- \\ere used from I to 12 hours after their last moult but one.
In -cries C food was administered two days, the third day the
ei- were starved (starvation lasting one day every third d \
+• • + +•-)• In series I) the animals were again de-
pri\ed of food every second day (starvation lasting one day
e\cry second day + •*+ • H — ) and the caterpillars of seru-
/. ueie depiived of I'CMM! i \\ o days and ate only every third day

irvation lasting twodays every third day +• + • + • -).

I in tin- absolute figures of Table II. as well as from I he
ptn tillages in Table 111 \\e ma\ infer that the duration of t he
lar\al period undergoes also, in these starved animal- coii-idei-
able prolongation (cf. analogous results obtained in silk-\\onn-
l>\ Kello-c .md Bell, '04 b), simultaneously with abbreviation <>t
the pupal stage. These changes become more and IIH-M
marUable in proportion as inanition in . the\ are larger in

-erie- /' than in series C' and the large>t in -eiit- /•.' in \\hich the
fa-ting da\'s were the most numeron-. Al-o in the-i . ..mpara-
ti\'e experiments the prolongation of the larxal life \\.i- greater
than the abbreviation of the pupal period, \\hile the duration of
the imaginal life n-mained unchanged. Hence it follows that
the total duration of the life of the insects from the beginning
of the experiment till death of the moth, i- more and more in-
creased in proportion as inanition increases within the limit- ot
the experiments. The duration of development from the la-t
moult but one till the emergence of the moth underwent in the
average of broods a prolongation ic.ilcul.ited in percentages of
the average duration of such control period-1 \\hich in the male-

STEFAN KOPEC.

amounted to 5.6 per cent, in series C, to 15.1 per cent, in series
D and to 16.8 per cent, in series E; in the females the prolonga-
tion was 13.3 per cent, in series C, 20.2 per cent, in series D and
28.6 per cent, in series E.

In series F the caterpillars were fed during two days and de-
prived of food during the next two. (Starvation lasting two days
every third day + +• ++ -). Notwithstanding a differ-
ent distribution of the feeding and fasting days, the quantita-
tive relation of these days was the same as in series D to which
starvation lasting one day every second day was applied (+•
+ • H — ); in both experiments the caterpillars experimented
upon were of the same age. It nevertheless turned out that the
prolongation of life was much greater in caterpillars of series F
than in specimens belonging to series D. (Cf. Table II. and III.)
In series F it amounted in the average of broods in the males to
54.2 per cent, of the average of the life of control caterpillars
from their last moult but one, and in the females to 78.0 per
cent while in series D the corresponding numbers were 40.5 per
cent, in males and 50.3 per cent, in females. The abbreviation
of the pupal period calculated similarly was in series F, in males
16.1 per cent of the duration of life of control chrysalids and in
females 24.6 per cent, while in series D the life of male pupje
underwent abbreviation of 5.6 per cent, and of the female 8.7
per cent. We see that changes of the duration of the larval and of
the pupal stages induced by inanition depend not only on the
mutual quantitative relation of the Listing and feeding period,
but also on the distribution of tlu-M- periods. The organism
responds by more energetic reaction to longer, though less fre-
quent, fasting intervals than to more frequent but shorter periods
of starvation.

3. INANITION AND ITS BEARING ON THE PHYSIOLOGY OF INSECT

METAMORPHOSIS.

When we try to discuss the foregoing results on the basis of
hitherto existing references in the literature, we meet certain
discrepancies which, however, may be cleared up by considera-
tion of the physiology of animal metamorphosis.

Kellner ('87) fed caterpillars of the silk-worm during their
whole life- on insufficient quantities of leaves, and obtained a

DEVELOPMEN1 V\I> DURATION OF LIFE IN INSKi DS. 9

short prolongation of the larval stage. Analogous experiments
\\ere afterward- made by Kellogg and Bell ('04 a and b). Pictet
('05) came to the convirtir.n that the larval stage is prolonged
and the pup.il -tage shortened by feeding caterpillars on plant-
containing few food -stuffs, but that the conjoint periods of
development do not undergo any changes. A prolongation of
larval lift- cau-ed by inadequate food was lately determined by
Northrop Ci~) Tangl. ('090), finally, maintains that the lar\a-
of flies reared on pure egg-white which they refused to take
underwent transformation approximately a week later than
normally. These results are, in general, concordant with my
observations on starved caterpillars. On the other hand, the
ob-ervation- of Krizcnecky ('14) and of Szwajsowna ('i<» on
the larva- of Tenebrio molitor which underwent metamorphi>-i-
earlier \vhen totally starved contradict the above determination-.
A -imilar discrepancy may also be remarked in analog .u- inve-n-
.uations on amphibians. While B.irfurth (87) and others ha\«
'•rtained that starvation of amphibians causes an eleration of
i heir metamorphosis, other investigators induced retardation
ot ihe-e processes by the administration of poor food to tadpole-.
It has been emphasized by Wolterstorff ('96), Morgan '07 .Kauf-
man ('18) and others that in amphibians the influence probably
ilillers in different developmental stages of the tadpole- u-ed f. ir
-tarvation. The former of the above mentioned in\ e-ti-ai..i ~
ill pri\'ed large tadj)oles of food, in part shortly before their
transformation, while the others used far younger -pecimen-.
SimilarK' the experiments of Kellner and the follouin- aiithm-
and also my own were made on youni; caterpillar^ \\hile the
ob-er\ations of Krizenecky anil of .s/\\aj-i '.\\iia refer to older
lar\a-. In another experimental -tudy on -tar\»-d tadpole-
l\"|)ec, 22d and b) I sin ceedcd in confirming the abo\ c \ ie\vs of
\\ I'lier-torll. Morgan and other-. My experiment- -h<>\\e<l that
metamorphosis of the tadpole- to \\hich -tarxation w.is applied
before the fiftieth day of their development undergoes retarda-
tion, where as (transformation i- accelerated when tadpole- are
-tarvedafter the-i\ty fifth day of their life. In t hi- -tudy on cater-
pillars similar behavior was al-o noted in one and the same ma-
terial of female larva-. It turned out that the caterpillars to
which total starvation ha- been applied approximately since

IO STEFAN KOPEC.

the seventh day after their last moult underwent pupation in three
different broods after from 7 to 13, from 7 to 16 and from 7 to
15 days (or on the average after 10.6, 10.8 and 10.0 days),
whereas the control specimens of these broods become chrysalids
as early as after from 6 to 12, from (> to 12 and from 5 to 12 days
(or on the average after 8.2, 9.3 and 8.6 days). On the contrary
the larval life of the females of the same broods in which starva-
tion began approximately from the tenth day after the last moult
was much shorter than that of control larvae, as pupation occurred
here from 4 to 8, from 5 to 8 and from 4 to 8 days (or on the aver-
age after 5.5, 6.6 and 6.8 days), in control specimens from 5 to
10, from 6 to 10 and from 7 to 9 days (or on the average after
7.4, 7.6 and 8.1 days). Hence follows: (i) The character of the
influence, whether accelerating or retarding, exerted by starva-
tion on the processes of metamorphosis depends on the period
of life or the developmental stage on which the factor begins. (2)
This moment may be accurately determined by means of ex-
periment, vi/, in the females of Lymantria dispar L. In nix-
cultures the critical moment during which the influence of in-
anition delaying metamorphosis is changed into an accelerating
one fell on the period between approximately the seventh and the
tenth day after the last moult.

An essential explanation of such different behavior of cater-
pillars may be found in my former studies on the importance of
the brain for insert metamorphosis (Kopec, '17 and *22r). It
was shown that the female caterpillars of this moth, deprived ot
brain the seventh day after their last moult, live far longer than
control animals, but they die without undergoing transforma-
tion. On the other hand, the removal of the brain from rater-
pillars the tenth day after their last moult has no influence either
on the processes of pupation or on the emergence of moths.
From various experiments supported by observation on control
material it follows that the brain plays here most probably the
role of an organ of internal secretion. From these results we may
infer that by depriving caterpillars of food since the tenth day
after their last moult we afford exceptionally favorable condi-
tions for metamorphosis, viz, the substance or substances al-
ready produced by the brain in sufficient quantities lind in such
starved organisms less material which ought to be transformed

DEVELOPMENT! A\l> DURATION O¥ LIFE IN [NSE< FS. II

in time. The prolongation of the larval period in younger cater-
pillars starved before tin- -cventh day after the last moult m.iy be
»•••.])!; lined, on the contrary, by impeded secretoric function of tin-
brain or by certain anomalies of its activity in starved organism-.
If the pro • sses occurring in the pupa were a continuation of
i hose Liking pi, ice in the caterpillar and leading to pup.it ion,
the abbreviation of the pupal stage in starved specimen- could
noi In- reconciled with the above explanation. But these proc-

- dilfi-r not only energetically as it has been shown by Tangl

The processes of pupation may morphologically be

reduced to the hi-tolysis of almost all larval tissues into a'homo-

eous mass \\hich fills up the pupal body, whereas the trans-
loniiation of the pupa into the fully-developed insect con-i-t-
in the de\ elopcinnt of the so-called imaginal discs, i.e., of small

1 1 in n la t ions of cells. In my present experiments < -cries B)
I -ucceeded moreover in establishing certain physiological char-
acters which point to another difference. In series M the cater-
pillars \\ere deprived of food every second day for -?o days after
their hatching from eggs, i.e., at a time when there .ire not e\ni
any traces of histolytic processes in the larval body. From the
twenty-first day till the end of the larval period the animals \\ en-
led every day and underwent the normal number of moiilt>.
From the appertaining items on Tables II. and III. we see that
tlu- caterpillars belonging to this series not only remarkably
dcla\cd the term of their pupation, but also that the duration
of the pupal stage was considerably shortened. Thi> .ibbn -\ia-
tion amounted in the average of broods to 17.4 per cent, of the
average duration of life of normal pupae in male- and t" JJ.S per
cent, in lenialc-. Consequently, the evolutive processes charac-
teristic o! t In- pupa are not a continuation of i ln.-e c li.m^e- \\ Inch
are ( -hara< tei i/cd by llistolysis ( ,f larxal ti-sur-. \\ V -t-i- that by
temporarily slarxin^ \oun- caterpillars ^cric- /•> u ( nia\' segre-
gate these pro- . —<- am I |>ro\e that processes of the development
ot the imaginal discs may bci;m far earlier, i.e., XKHI after the
hatching of the caterpillar from the e^- It "u-ht therefore to be
interred that the brain has t \\o ^eparate functions in normal Con-
ditions, (I) it cause- In-loly-i- of l.ir\al tisSUCS, J it delaN'- | he
evolution of imaginal di-c-. Muring inanition of i .lU'rpillars the
inlluence of the brain hindering the development of iniagin.il discs

12 STEFAN KOPEC.

is decreased, owing to its lowered function. The starved chry-
salids therefore begin their pupal stum* when the discs are better
developed and this causes acceleration of the development of the
imaginal body and hence abbreviation of the pupal life.

The assumption that the larval brain exerts two different in-
fluences is neither astonishing nor incomprehensible in respect
to the well-known data concerning the physiology of organs of
internal secretion. The final development of the imaginal discs
and their definitive differentiation sets in before pupation or even
in the pupa, i.e., at a time when almost all larval tissues are un-
dergoing histolysis or have undergone it. Therefore it ought to
be admitted that during this period the influence of the brain
hindering the development of embryonal discs normally becomes
annuled by certain processes which are unknown to us. It is very
probable that the development of imaginal discs and the histolysis
of the larval tissues become, at least towards the end of larval
life, physiologically correlated with each other. In the contrary
case it could be hardly understood why the caterpillars deprived
of their brain the seventh day after their last moult do not, it is
true, exhibit any histolytical changes in their tissues, but the
imaginal discs contained in their body do not undergo final
growth and differentiation. (Kopec, '17 and '22 c). Indeed,
if there were no such correlation, the imaginal discs of the cater-
pillars deprived in that period of their brain (and therefore of
the organ retarding their evolution) ought to develop the organs
of the imago in spite of the absence of histolytical procesM-s in
the larval body.

In contrast to my experiments on the starvation of caterpillars,
Loeb and Northrop ('17) have lately convinced themselves that
each of the separate stages of life in Drosophila occurs more
slowly in lower temperature and more quickly in higher ones.
The changes of temperature were applied by the mentioned
authors during the whole development of the animals, while in
my experiments the factor of inanition had a direct influence only
on the processes taking place in caterpillars, as the chrysalis
do^ not take food in normal conditions either. Loeb and Northrop
obtained the same changes of the duration of the larval and of
the pupal stage first of all owing to changed celerity of metab-
olism in the larva as well as in the pupa. Experiments in which

DEVELOPMEN1 AND DURATION OF LIFE IN INSE< tS. 13

lower or higher temperatures should be applied only to larvae,
the chrysalids being kept in normal conditions, might solve the
problem as to whether and to what degree the changed tempera-
lure has an influence not only on the celerity of metabolism, but
also on the -upposed function of the brain.

The attempt- hitherto made to explain the metamorphosis <>t
in-ert- refer to the last stage of the processes. The appearance
of phagoi \ losis, of degeneration, of asphyxia and of other
'causes" of transformation is not yet elucidated. My experi-
ments on the function of the larval brain as well as the discovery
of tyrosinase in caterpillars and chrysalids made by Dewit/
''05 and '16) and confirmed by Steche and \Vaentig ('13) together
\\iih the present results on the starvation of caterpillars may
li ad to a better knowledge of the cause of insect metamorphosis.
Hut my "secretory" theory of metamorphosis will not be well
Lioimdcd until we succeed in finding in the structure of the brain
an adequate base for the theory, i.e.. until certain specific chai
in the brain not only during pupation, but also during moult-
\\ill ha\e been ascertained. A great support might be .iltunled
by positive experimental results on transplantation of brains, or
on injections of extracts of this organ into brainless specimens.
The latter experiments only would be able to dispel every doubt
in regard to the validity of my previous conclusion that the in-
dubitable influence of the brain on metamorphosis is due to an
internal secretion of this organ.

I Vrgrncr ('09) lastly considers the imaginal form to be ph\ 1« .
netii-ally older than the larval form, which developed -«v(.mlarily
In 'in the hilly developed insect owing to numerou- -n-ondary
life conditions "unter fortgehender Ketanlation der I ait\\ i( klung
imaginaler Organe" (p. Ii). I think that my experiments of
series B distinctly prove that in n< >rmal cirrmn-i .m< »•- the devel-
opment of the imaginal discs is retarded during larval stages.
I cannot deny that in the light of these re-ults the supposed
retardation of the development of the imaginal organs gains an
experimental base, at lea-t in ontogriu-tii- evolution.

3. INANITION AND Ci KIAIN I'KMM i \is OF GROWTH.

In this chapter I take into account only the final stages of
growth of caterpillars c\pre--rd by the \\eiglit of new-formed

14 STEFAN KOPEC.

pupa1. The chrysalids were weighed from i to 24 hours after
pupation, control weighings having proved that the weight of
pupa1 decreased in my broods during this period only from 0.18
per cent, to 0.33 per cent. The limits of fluctuations in the weight
of pupa.' are recorded in Table I., together with their average
values. On Table III. we see moreover the average weight of
"starved" chrysalids of each brood calculated in percentages of
the average weight of control chrysalids of the same sex and
brood .

On comparing the data of series C, D and E we see that the
weight of chrysalids decreases more and more in proportion as
the number of fasting days increases. Assuming in general that
the number of feeding days in separate series corresponds to the
quantity of food which has been taken, we may say that the
weight of pupa^ is in direct relation to the quantity of food
consumed. If we also take into consideration the data of series
F in comparison to analogous items of series D, we draw the
conclusion that the average decrease of chrysalids due to inani-
tion of caterpillars is larger when the feeding intervals are longer
though less frequent. As it was pointed out, the prolongation of
the stage of caterpillar and the abbreviation of that of pupa
increases in proportion as more and more intense starvation has
been applied. Consequently, the weight of the pupa? the cater-
pillars of which had been starved is in inverse relation to the
prolongation of the larval period as well as to the abbreviation of
the stage of chrysalids. Adopting the ratio of the pupal weight
to the duration of larval life as the approximative measure of the
rate of growth, we can infer from Tables I. to III. that this
rate decreases in proportion as more and more intense starvation
has been applied and in relation to the distribution of the feeding
and fasting days (Series F and D). The above-stated principles
refer to separate series of experiments but not to separate speci-
mens, either control or starved, in one brood. I have often ob-
served that although the processes of transformation lasted in
the control or in the starved material in every brood several
days, the heaviest and the lightest caterpillars underwent pupa-
tion the same day, sometimes the first and sometime:- not until
the last day, although the one was two- or threefold heavier than
the other. The same may be said as to the duration of pupal

DEVELOPMENT AND Dl RATION OF I.1FF IN INSF.i PS. 15

life in regard to the weight of separate chrysalids of one brood.
The problem ari-e- whether the capacity to grow i- checked
by age of the .uiimal. Such limits in rats have been adopted by
Aron ('12 .ind l.itely by Jackson and Stewart ('20), in contra-t
to < )-borne .ind Mendel 14) who are inclined to the opinion
ili.it the i.ip.tcity to -ro\\ is exhausted by mere growth, without
,rd to the factor of time. In series B of my experiment - the
• rpillar-. depi i\ cd of focxl intermittently for 2O days since their
hatching. \\eighed the twenty-first day in separate broods, on the
avei :ily 7.6, 3.6 and 5.0 mg., while control individuals of

tin -.line brood- had the mean weight of as much as y^>, 45.5
.ind 40.} in-. The starved specimens fed daily since the twenty-
tir-i day attained and partly exceeded in time the weight .md
-i/e of ei >MI ml caterpillars (cf. weights on Table I. ami III. \-
iii thi- series the processes of metamorphosis of original!) -Mixed
• aierpillars were retarded and as, on the other hand, caterpillar-
ha\ e grown until the end of their larval life, it ought to be interred
that in t hese animals the capacity to grow is not -npprc--eil at a
period at \\liich caterpillars normally cease to gn>u . M"ie..\er,
it may be concluded that pupation also may take place mm h
be\(»nd the age at which the control specimens undergo pupa-
II«MI. But the inference that the capacity may not at all be
limited by the age of animals ought not to be drawn from Mich
results. 1 1 is possible that the rats experimented upon \>\ Mendel
and < )sborne as well as my caterpillars would lose the capacity
to obtain the weight and size of control -pecinu-n- it they had
been kept at maintenance or starved c\cii a feu da\- longer.

4. ;\D\ri \1ION OF OR( IAN ISMS TO S I \ \< \ \ IK »\ . Si All l\l \I \KK-
ON TIIF. PK()l;l IM n| I )| \ Ml.

The \\eiijit of pupa- from M-rii - .1 in which >tar\ation la-ting
01 H- dav e\er\p second day, + • + • + •-, \\a- a|>plied in cater-
pillars during the whole lar\al life i- exideiitly different from
that of series I) in \\hich older -pecinu-n- after their la-t moult
but one were subjected to the same -tarvation (cf. Table I.).
We see that or^ani-m- may in time ^et accu-tomed to the detri-
mental effects of -tarvation. which prevent the animal from at-
taining its normal weight: the compari-on of the limits of indi-
\idual lluct nations -hown on Table I. point- to the coin ln-ion

16 « STEFAN KOPEC.

that the chrysalids of series D which have been deprived of food
only during the half of their larval life are lighter than those from
series A the caterpillars of which have been starved during their
whole larval life. From Table III. we see that the male pupa1
from series D weighed in the average of broods 48.5 per cent.,
the female pupa? 48.4 per cent, of the average weight of the con-
trol chrysalids, whilst the male chrysalids from series A weighed
as much as 70.2 per cent., the females 78.3 per cent .

As the prolongation of the larval period was much greater in
series A than in series D (cf. Table I.), it might be supposed that
the caterpillars of series A are heavier owing to the circumstance
that they lived longer, and therefore could take and digest food
during a longer period. But the following argument contradicts
this opinion. At the outset of the experiment in series D the
caterpillars weighed in the average of broods 94.5 nig., the male
specimens underwent pupation in the average of broods after
26.1, the female after 29.7 days. If we take as starting point for
series A the day on which the caterpillars of this series had an
analogous weight, which in the average of broods amounted to
93-5 m£-> the duration of larval life in this series from this day
till pupation was in the average of broods 27.6 in males and 36.2
in females. We see that when the starting point had been made
uniform the duration of farther larval life in series A, especially
in males, is almost identical with that in series D. In other words,
the specimens of series A attain considerably greater weight than
those of series D during approximately the same period. The
ratio of the produced number of milligrams of body-weight to
the total number of days of the period during which the larvae
had been deprived of food being considered as rate of growth, we
may calculate that the rate of growth amounts in the average
of broods in series A to 7.4 in starved males and to 19.8 in females,
whilst in series D this quantity attains only 4.5 in starved males
and 15.8 in females. It follows that the starved organism may
in time get accustomed to the metabolism of inanition, i.e., the
ratio of assimilation to disassimilation becomes during lon;^
starvation changed in favor of the organism.

By histological research I have convinced myself that cater-
pillars which died from starvation contain no adipose tissue. In
a certain contrast to my observations, Bialaszewicz ('19) draws

DEVELOPMENT AND DURATION OF LIFE IN IN^i I- \~

from his experiments on the starvation of leeches as well as from
the papers by several authors on other animals the conclusion
supported by numerous grounds that fat has no great import. in. v
in the hunger nirt.iboli-m of cold-blooded animals, but that it
undergoes only small reduction. The farther investigations of
Hialas/ewi<-/ will undoubtedly show whether my supposition
based on histological research is right, or they may explain the
cause <>f this discrepancy which possibly consists in the physio-
logical capacity of caterpillars to digest their store of adipose
tisane during their frequent moults.

In contemporary research the cause of natural death is regarded
• ui>ed by the accumulation of detrimental products of normal
metabolism which cannot be removed from the mtilticellular
organism. This view is not at all proved; it is based first of all
on the known investigations of Woodruff on the infusoria whose
conclusions are in discrepancy with the results obtained by
Viewegerowa and \ ieweger ('22) who by methodically exact
research proved that the products of metabolism have no great
importance on the development of Colpidium and tli.it tin- di\i-
sions are hindered in unchanged surroundings first of all by inani-
tion. Ii nevertheless seems to me to be unquestionable th.it the
duration of life depends on the character of metabolism, in oilier
words, that natural death is a function of metabolism. The
experiments performed by Kellogg and Hell ('04 b), by 1'ietei
<>j) and by Northrop ('17) on larva? as well as those 1>\ I .oeb
.UK! Northrop ('17) on fully-developed insects of 1 )n.-o|)liila
simu th. LI by means of inadequate food we may elicit considerable
changes of the duration of development or of life of tin -<• »igan-
i-ms. As investigations on the hunger metaboli-m e\ idence
distinct differences of the digested substances, .md <>t tin- character
of their disintegration in comparison t" normal nu-tabolism, it
ought to be inferred that, if death i- a I'UIH lion oi iiH-tab«.li-m.
insufficient feeding may also influence the in.mirnt of natural
death. Schultx ('05) emphasixes that certain animals under-
going periods of hibernal (or ,r-tival) sleep which is < . .ntu-i i-.-d
with \'er\' restricted metabolism live \ ery long in com]i.iri-on
with those having no such periods of rest. Stn— must be laid
on the fact that in my experiments moths deriving from starved
caterpillars in which the development was much prolonged or

2

1 8 STEFAN KOPEC.

delayed are much older organisms than the control moths. As
the imago of Lymantria d is par L. never takes any food, the quan-
tity of provisions stored in its body decides the duration of this
life-period. From histological research it follows that the amount
of this provision which may be noticed in the imago (the so-
called adipose body) is in the starved specimens, in relation to
their decreased body, not at all smaller than in the controls.
This is due to the fact that the starved caterpillars do not form
normally sized chrysalids, which contain smaller quantities of
the mentioned body, but they are transformed into smaller,
even dwarfed specimens. By this circumstance it might be
explained why both categories of moths live in general equally
long. But in the case of the "starved" moths derived from cat-
erpillars the life of which has undergone a very remarkable pro-
longation it is obvious that the hunger metabolism during devel-
opment caused retardation of natural death of the organism.
This I consider to point clearly to the conception of death as a
function of the character of general metabolism. According to
Ruzicka's researches ('17), the newts, when totally deprived of
food, undergo moults more rapidly than control specimens. In
connection with my experiments it would be very interesting
to ascertain whether the duration of life of such newts undergoes
changes.

More detailed considerations on the cause of especially favor-
able influence of intermittent starvation on duration of life of
animals are to be found in my former papers (Kopec, '22 a and b.}

%

5. SUMMARY.

1. Intermittent starvation of young caterpillars of Lymantria
dispar L. causes considerable prolongation of the larval life as
well as a certain abbreviation of the pupal period but has no
influence on the duration of life of the imago. These changes
increase in proportion as more intense starvation is applied.
Larger effects are elicited by longer and less frequent than by
more frequent, but shorter, feeding intervals.

2. The differences of results obtained by various authors in
regard to the influence of starvation on metamorphosis depend
on differences of age of the animals experimented upon. Tin-
caterpillars subjected to inanition approximately from the seventh

DI-.VI l.nl'MI-.Vl A\l» I -I RATION OF l.IFK IX 1NSKCTS. IQ

day after their last moult had retarded pupation, whereas this
proces> i~ .« i < -leraied \>y starvation of animals approximately
since the tenth day after the last moult.

.v The development of imaginal discs is not the consequence
(•I' hi-tolyti( al [processes which cause pupation of caterpillars,
hut they i.ike place simultaneously from the first days of larval
life. Tin- brain causes, probably by its secretion (or secretions),
hi-tolysis of larval tissues and it also seems to check in the cater-
pillar the development of the imaginal discs.

4. The prolongation of the larval peruxl in starved specimens
may be explained by certain disturbances in the hypothetical

•etory function of the larval brain, which are caused by inani-
tion; the abbreviation of the pupal stage may be ascribed to
analogous decrease of the influence of this organ, which retards
i !)»• development of imaginal discs.

5. The average limit of larval growth expressed in the average
\\cight of the new-formed chrysulids is in direct pro|M>rtion to the
<|ii.mtity of food given and inverse to the prolongation of larval
.Hid to the abbreviation of pupal life. (The decrease of weight
of caterpillars is larger in cases of longer though more rare fo<xl-
intervals than in cases of more frequent but shorter ones.) These
rules may be applied only to differently starved whole experi-
mental materials, having no application to separate specimens of
-i p. irate broods, both starved and control.

6. The capacity to grow as well as the capacity to undergo
metamorphosis exists in starved specimens far beyond the age
at which control caterpillars cease to grow and undergo trans-
formation.

7. During long lasting starvation organi-m- get ,n i u-ioined to
the abnormal conditions: the rate of gro\\th of tin- caterpillars
starved every second day during their whole life becomes in time-
considerably greater than that < •! specimen-, .m.ili igic.dly deprived
of food since their last moult Inn one.

M. Hunger death of caterpillar i> probably c.m-ed first of all
by exhaustion of reserve sul»tance-. Natural death of the imago
probably is a function of the character of metabolism, as death is
delayed by the changed metabolism of intermittent starvation.

2O STEFAN KOPEC.

BIBLIOGRAPHY.
Papers marked by an asterisk are known to me only from abstracts.

* Aron, H.

'12 \Veitere Unteisuchungen iiber die Beeinflussung des Wachstums durch

die Erniihrung. Verh. d. Gesell. t". Kinderheilkunde, Miinster.
Barfurth, D.

'87 a Versuche iiber die Verwandlung der Froschlarven. Arch. f. mikr. Anat.,

Vol. 29, p. i.

'87 b Dor Hunger als forderndes Prinzip in der Xatui. Ibidem, p. 28.
Bialaszewicz, K.

'19 Etudes comparees sur le metabolism chimique et energetique. I. L'inani-
tion et la nutrition chez les Hirudinces. Trav. de la Soc. des Sc. de Var-
sovie, III., Nr. 32.
Deegener, P.

'09 Die Metamorphose dei Insecten. Teubner, Leipsic and Berlin.
Dewitz, J.

* '05 Untersuchungen u'ber die Verwandlung der Insectenlarven. II. Arch. f.

Anat. und Physiol., Phydol. Abt., Suppl.
'06 III. Zusammenfassung friiherer Mitteilungen. Zool. Anz., Vol. 47.

* Jackson, C. M., and Stewart, C. A.

'20 The Effects of Inanition in the Young upon the Ultimate Size of the Body
and of the Various Organs in the Albino Rat. Journ. of Exper. Zool.,
Vol. 30.
Kaufman, L.

'18 Researches on the Artificial Metamorphosis of Axolotls. Bull. Ac. Sc.

Cracovie.
Kellner, O.

'87 Chemische Untersuchungen iiber die Erniihrung und Entwicklung des

Seidenspinners (Bombyx mori). Landw. Versuchst., Vol. 33.
Kellogg, V. L., and Bell, R. G.

'04 a Notes on Insect Bionomics. Journ. of Exper. Zool., Vol. i.
'04 b Variations Induced in Larval. Pupal and Imaginal Stages of Bombyx mori
by Controlled Varying Food Supply. Science, N. S., Vol. 18 (Leland
Stanford Junior University Publications, University Seiies, No. i).
Kopec, S.

'17 Experiments on Metamorphosis of Insects. Bull. Ac. Sc. Cracovie.
'22 a Further Research on the Influence of Inanition on the Development of
Animals. Experiments on Tadpoles. Mem. de 1'Institut Nation Polonais
d'Economic Rurale a Pulawy, Vol. 3.
'22 b Experimental Studies on the Influence of Inanition on the Development

and the Weight of Amphibians. Bull. Ac. Pol. Sc., Cracovie.
'22 c Studies on the Necessity of the Brain for the Inception of Insects' Meta"

morphosis. BIOL. BULL., Vol. 42.
Krizenecky, J.

'14 Uber die beschleunigende Einwirkung des Hungers auf die Metamorphose.

Biol. Centrbl., Vol. 34.
Loeb, J., and Northrop, J. H.

'17 On the Influence of Food and Temperature upon tin- Duration of Life.
Journ. of Biol. Chem., Vol. 32.

DKVKI.ni'MKN T AND DURATION OF LIFE IN IN^I < I-. 21

Morgan, Th. H.

'07 Experimental Zoology, Xew York.

* Northrop, J. H.

'17 The Effect of Prolongation of the Period of Growth on the Total Duration
of Life. Journ. of Biol. Chem., Vol. 32.

* Osborne, T. B., and Mendel, L. B.

'14 Tli<- Suppr«-Mon of Growth and the Capacity to Grow. Journ. of Biol.

( l..-m.. Y.il. 18.
Pictet, A.

'05 Intlu'-iit ••• •!«• 1 alimentation et de I'humidite sur la variation des papillions.

M«'-m. de la soc. dc physiques et d'hist. nat. de Geneve, Vol. 35.
Ruzicka, V.

'07 Beschleunigung clcr Hiiutung durch Hunger. Arch. f. Entw. Mech., Vol. 42.
Schultz, E.
'05 Ubcr VerjQngung. Biol. Centrbl.. \'ol. 25.

* Steche, O. u. Waentig, P.

'13 L'ntersuchungen uber die biologische Bcdeutung und Kinetik der Katalase.

Zoologica. \'ol. 26.
Szwajsowna. P.

"16 Le m«kal>olisine pliysiologiquc chez Ics larves du Tent-brio niolitor. Comptes

rend. Soc. Sc. dc Varsovie, \'ol. 9.
Tangl, F.

'09 a /.ur Konntnis de« StofT- und Energieumsatzes holometaboler Insckten

wiihrend der Mi-tamorphosc. Arch. f. d. ges. Physiol., Vol. 130. p. i.
'09 b Embryooale Entwicklung und Metainorphost- voin enerm-ti- ln-n xand-

ptinktc aus U-trachtet. lljidi-in. p. 55.
Viewegerowa, J., and Vieweger, T.

*JI Rt'-cherches sur les causes du developpcment des cultures du Colpidium
colpotla. Trav'aux du lalx>r. physiol. de I'lnsl. M. Nencki. \'arsovie, \'ol. i .
Wolterstorff, W.

'96 Uber die Neotenie der Batrafhicr. Zool. Garten, Vol.

ON THE HETEROGENEOUS INFLUENCE OF STARVA-
TION OF MALE AND OF FEMALE INSECTS
ON THEIR OFFSPRING.

STEFAN KOPEC.

GOVERNMENT INSTITUTE FOR AGRICULTURAL RESEARCH, PULANVY, POLAND.'

(WITH 4 TABLES.)

Our knowledge of the influence exerted by insufficient feeding
of the parents on their offspring is almost entirely confined to
certain observations made on fetuses of the Mammalia. It has
been noticed that these fetuses are able to take from tissues of
the mother the substances which are more or less lacking in her
food. The problem as to whether and to what degree starvation
of either of the parents separately has any influence on the total
development of their progeny from birth till sexual maturity has
so far as I know, not as yet been methodically investigated.

Kellogg and Hell ('04) ascertained that starved silkworms
mated with each other produce caterpillars which develop more
slowly and produce lighter moths, in spite of the normal feeding
conditions of this generation. But, as the authors have exclusively
mated the starved specimens with each other they have not
discriminated the influence of starvation of either of the parents
separately. For this discrimination it was necessary to mate
both females subjected to starvation with control males and
normal females with starved males. These two experiments
were undertaken in the present research.

The experiments were performed on the mot h Lynidutria dispar
L. The wrhole material came from series D of my investigations
on the influence of starvation of caterpillars on the development
of insects, (Kopec, '21 and '23). As Lymantria dispar does not
take any food in its imaginal stage, starvation of the parents
can take place only during their larval stage. In that series of
experiments the caterpillars were fed every second day since
their last moult but one. Adult moths which develop from these

1 Paper from the laboratory of Experimental Morphology. Cf. Mi'm. dc 1' I >i-
slitiil Nat. Polonais d' Economic Rurale £ Pulauy, Vol. 2, 1922.

22

INFLUENCE "I -IARVATIOX OF INSECTS OX OFFSPRING. 23

< aterpillars were mated to control moths which were at my dis-
posal. In the filial generation the conditions of rearing were
identical in each lot. Food was given daily and in superfluity.
A- .ill lot- originated from one pair of moths caught in 1917.
the material is to a certain degree genetically homogeneous.

Tin following physiological characters of the filial generation
uere taken into account: (l) The duration of the larval life,
\\hich in the parental generation had undergone considerable
prolongation owing to inanition, (2) the duration of the pupal
stage which in the parents had undergone a certain abbreviation,
(3) the limits of growth of the caterpillars expressed by the weight
of fresh pupa- which in the starved parental generation were
much smaller than control specimens. Apart from this, the sixe
of the eggs out of which the first filial generation developed,
ilicir number and their capacity to develop as well as the mortal-
ity of caterpillars and of chrysalids have been examined. The
changes of the above-mentioned characters were in the filial
generation in part much smaller than in tin- parent-. I therefore
did not use for my calculations in these experiments the averages
of a character for each lot separately, but in each experiment I
joined .ill specimens of either sex of all lots into class-frequencies
as usual and I calculated, the average values for all specimens
together with the probable error of these averages (A ± E,\). On
comparing .my character of the control and of the experimental
material only such differences are considered as essential, i.e.,
biometrically well grounded, which are larger than four times

... / DifT. \ Di IT.

their probable error I =— > 4 \ When the ratio is < 4

V Enut. i tL.\i\n.

the differences are considered as biometrically not essential and
the character as not changed.

Tin-: OFFSPRING OF FEMALES Sunji-:» IM» \^ si \m \TION M.\n:i»

\\ 1 1 ii ("<>\ i k< ii M M.I-IS.

From Table I. we may infer that the number of eggs laid by
these specimens was much smaller than that prodmed by control
females mated with control male-. < )n the other hand, the sixe
of the eggs was biometrically identir.d in both cases, as is evi-
denced in Table II. Almost .ill eggs from -t.ir\ed females were
i .1 |).ible of development, although a large number of the develop-

STEFAN KOPEC.

-C

. en

c|Ef^|

t**" C O t^* O O "}

(/_

'5

•o
o

u fcc'i'S a

£5 "w

oo 10 »o 10 m o o

^O i"** t** ^D t"* ^O t"*

oc

a

s

00

CO

. l~

+-* rt i C

M o^ ^^J ^ O ^o ^o

00

6-3

Ji-

£3" <S

o o o* w ^f w 10

H M H

o

ll

-^ u.
C C3

DO

u .

C M e J3S a;

*> M E 0 -2 Si

t^- t^* W r» rf O O\

-T

<3«

0 '

M

Ufcd o^ £^

O ^f ?*l O GO to O
O\ O\ 00 oc r*» w O\

00

Offsprin

u,

^ta S

E O M

\C ro r^ 10 O O\ *H

00
PO

. ^

•"!5 tr"0

'oMr-O^t^-OoOOfOOO^-

>0

•o

V

£d"l;^l

O 0s oo t** oo t** r*- oo oo oO O\ O^ I"*-

M

oo

a

s

en

• l-i •

00.000^0,0^00,00

ro

rt^

11

if1^!

R^-aSS5B8»*35a

-O

00

T3 0
0, U,

11

t- .

*i m C3 73
w bo - .C S 3j

^ bC c u -^ ^3

oo .o,w~oo,ao 0^-^,0

00

l<3

1*4

O
bfl

cu ^w

ro ^^ ^^ ,0 ^f PI O^ ^D oo ,0 oo t~^ PI

ft

Offsprin

u

Cft

1^1

O^ ^^ oc PI r^« o O^ O^ Pi O^ ^* Pi r>*

M P) PI M M PI M

tH

PI

si A^i

OvOooOOOOO-r

^

^
•o

4)

£5 ^^1

O t--O\oooo O\oo O\

oo
oo

od

s

<n .

si i g

00^.0-000-

rO

^3.2

E.2

£j

S *- ^^3 ^ 3

O O\PI t— t~- >O oo r~

PI PI IO PI PI PI PI

P)

]!

s§

>•« .

r~* PI ,o ,o o O O t-*

-t

c3'^

1*4

o

•

O ti] O '^ u" U

COI-IPIPIIOPI-J-O

0

Offsprin

u

"2<*- bo
COM

SSSS2JES

O
00

P)

o

rt y.

~ c3

IO

Average. .

INFI.UKM K OF STARVATION OF I\-K<TS ON oFFM'KIXi..

ing eggs died in time. (The number of eggs which did not develop
at all was not larger than I per cent, of all eggs in the control
material as well as in that experimented upon.) The percentage
of eggs from which caterpillars hatched was approximately iden-
tical in thi> experimental material and in the control specimen-.
The same may be said as to the percentage of caterpillars which

TABLE II.

A — average; EA — probable error of the average; a — standard deviation; « — num-
ber of specimen- \t-.\\. — difference between the control and the experimental
material; Eit\n. — probable error of this difference.

DATA CONCERNING MAXIMAL DIAMETERS OF EGGS, IN p.
DIFFERENCES OF THE ABOVE AVERAGES. IN /i.

Material*.

-4 ± EA.

a.

w.

of control female* mated with control males

1234.65^2.319

53-81

245

P ( .-.I, i i

1 237-09 ±1.783

5I.6I

381

• >{ control females mated with starved malt

238. oo ±2.064 ' 44.35 2io

Average* Compare*!.

Diff. ± /•:,„„

Diff.
'•'niff.

I >:tterencc between the average size of
control eggs and of that of the eggs
from starved females mated with con-
trol males.

±2.925

0.83

Not essenti.tl.

I Mttrrence between the average size of
control eggs and of that of the eggs
from control females mated with
starved males.

1. 08 ' nti.il.

underwent pupation and of the chrysalids which de\ eloped into
adult moths (cf. Table I.). Pupation of caterpillars and hatching
of moths having been checked once daily during one and the
same hour, the limits of this life-period were determined with
exactness up to 24 hours. The data of Tables III. and I Y referring
to the duration of the larval and of the pupal life in the progeny
of starved females mated with control male- as \\ell .is the same
data of the control material (control 9 9 X cm mil ofd") show
that the duration of the larval life i- ixn e»entially (hanged but
the duration of the pupal period of the offspring of the females
Hibjected to starvation ha> undergone e--cntial abbreviation.
The limits of larval growth lu\e been determined, as in the

26

STEFAN KOPEC.

experiments with the parental generation, by the weight of new-
formed chrysalids. As pupation wras checked once a day it ought
to be emphasised that during the first 24 hours the chrysalid did
not lose more than 0.15 to 0.28 per cent, of the original weight
of the chrysalid immediately after its emergence from the larval
skin. If we take into consideration the appertaining strict bio-
metrical data of Tables III. and IV., as well as the rati

10

'Dill'.

we may ascertain that the weight of the chrysalids underwent no

TABLE III.

DATA REFERRING TO THE DURATION OF LARVAL AND PUPAL LIFE AND TO THE

WEIGHT OF CHRYSALIDS.

A — average; EA — probable error of the average; a — standard deviation; n —
number of specimens.

Materials.

Sex of
the

Speci-
mens.

.-i ± EA.

a.

n.

O)

1*4

15

CS «

~* **•.
•— rt

O "C

l-s

4-t

5

3
P

Caterpillars from control females
mated with control males.

0*0*

52.8i± 0.247

4.42

146

9 9

63.21 ± 0.309

6-93

229

Caterpillars from starved females
mated with control males.

0*0*

5I.66± 0.217

4-23

173

9 9

03-73± 0.320

6.14

167

Caterpillars from control females
mated with starved males.

0*0*

52.22± 0.325

4-57

90
71

9 9

63-39± 0.762

9-52

:-=

"«
a

u.

'o-g
°*

4J

rt

u

Q

Chrysalids from control females
mated with control males.

0*0*

22.83^ 0.0568

0-93

122

9 9

ig.i5± 0.0525

1. 12

20?

Chrysalids from starved females
mated with control males.

0*0*

22.21 ± 0.0521

0-93

145

9 9

18.51 =t 0.0420

o.7<.

149

Chrysalids from control females
mated with starved males.

0*0*

21.01 ± 0.0771

0.84

54

9 9

i8.45± 0.1275

1-44

58

tO

TO

~&

00
>»

1-
— tt

.'" £

"o _c

~ ""

!s?
"v

z

Chrysalids from control females
mated with control males.

0*0*

283. 56 ± 4-63

82.92

146

9 9

758-5I ±13- "4

292.64

229

Chrysalids from starved females
mated with control males.

0*0*

303. 76 ± 4.17

81.38

173

9 9

794.61 ±18.02

345-25

if>7

Chrysalids from control females
mated with starved males.

0*0*

2i6.67± 4.14

58.21

90

9 9

609. 86 ±19. 34

241.65

71

1NF1.I I \< I OF STARVATION OF INSECTS ON OFFSI'KIV,.

TABLE IV.

DlM I KIM is (il Avi.KAOE DERATION OF LARVAL AND OF PVPAL LlFE AS \\M I

AS OK \\ I.K.HT OF CHRYSALIDS BETWEEN THE CONTROL AND THE
EXPERIMENTAL MATERIAL.

Diff. 'litt'-ii -nee between the contiol and the experimental matt-rial.
— probable error of this difference.

1 om pared.

Sex of

the
Speci-

Diff. ±

•

^ 'ire of
rence.

P

c
3
a

I >ni. r> in c Ix-tween the
average duration of lar-
val life of control speci-
mens and of that of
specimens deriving from
starved females mated
with control males.

1.15 ± 0.329 3.40

52 ± 0.445 1-17

N'ot essential.

Dm. rence Ijetwifn tin-
average duration nf lar-
val life of control speci-
mens .HP I of that of
s|K-cinu-nH deriving from
control females mated
with starved males.

0.59^0.408 1.44 Not essential.

o.i8± 0.822 0.22

Data concerning pupal life. *
in days.

Difference between tin-
average duration of pupal
life <ii control npecimeru
and of that of specimens
deriving from starved
females mated with con-
trol males.

1

o.62± 0.0771

8.04 Essential.

0.64 ± 0.0672
0.92 ± 0.0958

9-52

Difference In-tween tin-
average duration of pupal
life of control specimens
and of that of specimena
deriving from control fe-
males mated with starved
males.

9-6o Essential.

o.70± 0.1378

5.08

Data concerning weight of
chrysalids. in mg.

Difference between the
average weight of control
pupa.1 and of that of
pupa- deriving from
starved females mated
with control males.

dV

— 2O.2Odb 6.23

3.24 Not essential.

9 9

— 3&.IO±22.24

1.62

i

Difference between the
average weight of control
pupa- and of that of
pupa* deriving from con-
trol females mated with
starved males.

0*0?

- - 6.21

10.77 Essential.

9 9

^3-32

6-37

1

STEFAN KOPEC.

essential change in this experimental material. (The pupae are
here even heavier, but the differences are smaller than their
fourfold error.)

THE OFFSPRING OF CONTROL FEMALES MATED WITH MALES

SUBJECTED TO STARVATION.

As the control females mated with starved males were by
chance mostly larger than the control females mated with control
males, they laid somewhat larger quantities of eggs (Table I.).
The egg size, however, was identical in this material and in that
of the control, as may clearly be inferred from Table II. Also
in this experiment the percentage of eggs which did not develop
at all Was not larger than i per cent, of all eggs.

An exact inquiry into the data of Table I. shows that the capac-
ity of the spermatozoa from starved males to fertilize and to
timulate eggs to total development is, if not larger, at least nots
smaller than in spermatozoa of control specimens. On the other
hand, mortality of caterpillars as well as of chrysalids in this
experiment was undoubtedly larger than in control materials.
From data of Tables III. and IV. we see that the larval life in tfie
offspring of control females mated with males subjected to inani-
tion does not undergo any essential change. On the other hand,
the duration of the pupal stage is essentially shortened in this
experiment. On comparing the data of the same tables it is also
found that in contrast to the offspring of starved females mated
with control males, the offspring of males subjected to starvation
and of control females is lighter than the progeny of control
material.

GENERAL CONCLUSIONS AND SUMMARY.

The above experiments are the first to discriminate strictly the
influence of starvation exerted on the offspring by either of the
parents separately. The starved females the progeny of which
was used for this research were much smaller than the control
specimens, their limit, of growth being only 48.4 per cent, of the
normal limit of growth (Kopec, '21 and '23, Table III.). Never-
theless, the body-weight of the offspring of such starved females
was at least normal. This leads to the supposition that in females
hunger metabolism does not bring about any quantitatively

l\l I I 1 \CE OF STARVATION" OF INSECTS OX OFFSPRING. 2Q

unfavorable changes in the reserve substances out of which eggs
are developed in the chrysalids. On the other hand, we have to
do hen- with a special regulation of quantitative relations, con-
n-ting of two factors: the number of eggs laid by the females
undergoes in general considerable reduction, but the egg si/e
does not exhibit any changes, which points to unchanged quantity
of food substances for the developing embryo in each "starved"
egg-

The normal si/e of eggs laved by starved females together with
the decreased numl>er of the eggs speak in favor of the supposi-
tion that in case of insufficient feeding of an animal the food
substances may l>e exclusively used up by a certain number of
eggs which develop better than the remaining, owing to their
arrangement in the ovary or to other conditions. This is caused
by the competition between separate elements, which undoubt-
edly takes place in the starved ovary analogous to the struggle
between individuals of a brood, which develops in unfavorable
conditions. Owing to this rivalry, only part of the eggs un-
dergo development , directly or indirectly at the expense of neigh-
boring elements, as the nuinlxr of eggs contained in "starved"
chrysalids is considerably larger than the number of tully devel-
oped eggs in the mature females. Notwithstanding the capacity
of certain "starved " eggs to develop at the expen-e of other eggs,
thi- average si/e does not exceed essentially the normal average
limit, which points to the dependence of e^x si/e on certain
internal factors by which the limits of the growth of eggs are
determined. These factors are undoubtedly synonymous with
"genes" which determine in genetic, d sense the si/e of eggs
independently of the si/e of the female- by which the eggs are
produced. In spite of this evolutive independence during the
development of a fetus or of an e^u. there probably exists a certain
limit of degree and of duration of starvation which may be applied
to a female, beyond which limit the progeny does not attain its
normal si/e. The assumption that this limit may easily be
transgressed is supported by the fact that certain small unfavor-
able changes in feeding bird- cause a considerable diminution of
their eggs and even (heck the further processes of development
of the sexual element-. This limit was undoubtedly transgressed
in the experiment- made by YYoltereck ('08 -'n.) in which the

3O STEFAN KOPEC.

offspring parthenogenetically produced by starved Daphnia was
smaller than control specimens.

In contrast to the offspring of starved females the progeny of
similarly treated males were essentially lighter than those pro-
duced by control males. The unfavorable influence of starva-
tion of males was evidenced by increased mortality of caterpillars
as well as of chrysalicls, which could not be observed in the off-
spring of starved females. As egg size was also in this experiment
normal, the cause of these changes ought to be ascribed to the
specific influence exerted by starvation of males on their sperma-
tozoa.

A great influence on the offspring of the Mammalia exerted by
certain substances applied to their fathers was demonstrated,
e.g., by experiments on the influence of alcohol made by Stockard
and his collaborators ('12- 'i 8) and others, by the experiments on
lead performed by Cole and Bachhuber ('14) and by those of
Guyerand Smith ('18 and '20) on a special lens dissolving serum.
These researches have shown a considerable susceptibility of the
spermatozoa to certain extrinsic specific substances. My experi-
ments prove that normal development of the spermatozoa may
also be influenced by certain natural changes of metabolism
caused by inanition. Histological research has shown that the
males which develop from starved caterpillars contain the same
quantities of reserve substances relatively to their diminished
size as control specimens; as the testicles of the moth which
develop in the caterpillar at the expense of these substances are
too small organs to play an important role in the nutritive balance
of the chrysalis undergoing metamorphosis, it ought to be sup-
posed that the influence exerted on the spermatozoa by hunger
metabolism consists in qualitative and not in quantitative changes
of the chemical constitution of the spermatozoa. In my experi-
ments I did not succeed in ascertaining any marked histological
changes in the developing "starved" testicle, neither as to the
evolutive rate, nor as to the structure of the sexual elements.
Physiological dimorphism of metabolism in the two sexes is
certainly the cause why starvation of females does not elicit
analogous unfavorable quantitative changes in the egg.

From the above results we may infer that the decreased weight
of the offspring of two starved parents determined in silkworms

IM I.I l.\( E OF STARVATION OF INSECTS ON <>l I-1'RINC.. 3!

by Kellogg and Bell ('04) was exclusively due to the influence of
tin- -permato/oa from the starved males, unless the limit- beyond
which we "light to expect a noxious influence of the -tar\ .it ion of
It in. tie- (in their progeny had been transgressed, ow in g to different
decree dt" -tarvation. It ought to be mentioned that in certain
species the development of the sexual elements occurs in natural
condition- during periods of total physiological starvation.
Miescher ~ ascertained that in the salmon which dors n»i
take any food during the development of its sexual glands its
very large ovaries and testicles grow at the expense of niu-i
This fact, however, is in no discrepancy with my results as it may
be that certain special feeding experiments with salmon it they
are practically possible- would, as in moths, elicit in this li-h
changes as to the size of the offspring which in natural conditions
develops here from "starved" spermatozoa. The generalization of
my results obtained in insects to other animals, especially to
vertebrates, would be premature. I nevertheless believe that my
present contribution will have a certain general interest in regard
to the great biological importance of the discussed problem.

The duration of the larval period does not undergo any essen-
tial change in the offspring of starved females nor in that of tin-
starved males either. The pupal stage, on the contrary, is essen-
tially changed in both cases. My conception of tin- physiology of
insect metamorphosis is supported by such behavior of the prog-
eny. According to my opinion tin- larval brain elicits by its
secretion or secretions the histolysis of lar\al ti— ue-, which puts
an end to larval life (Kopec, '17 and '22 It .dso checks in the
larva the development of the embryonal discs from which the
organs of the imago are finally formed in tin- chrysalids ('21 and
'23). The prolongation of the larxal lift- and the simultaneous
abbreviation of the pupal stage c.m-ed by starvation of cater-
pillars has been interpreted a> due to delayed or decreased secre-
toric function of the brain. If the prolongation of larval life
elicited by starvation of the mother or of the father is due to
delayed or decreased production .if the substance by which histol-
\ -is of tissues is brought about, no wonder that in the filial
generation there is no change in respecl to the duration of larval
life: in these experiments the larval organism of the filial genera-
tion remains unchanged as to it- preparation for histolysis. The

32 STEFAN KOPEC.

duration of the pupal stage in the filial generation must be on the
contrary explained in a different manner. It has been shown that
the abbreviation of the pupal period and consequently the accel-
eration of development of the embryonal discs was brought
about even by temporary starvation of very young caterpillars,
viz., lasting 20 days after hatching from eggs, i.e., at a time when
there are in the larval organism not yet any traces of histolytical
processes characteristic of pupation (Kopec, '21 and '23). It
may therefore easily be assumed that not only the embryonal
discs of the animal subjected to inanition, but also the evolutive
factors contained in their sexual cells, which cause the evolution
of the embryonal discs in the filial generation, are stimulated to
development by starvation and by the decreased checking
function of the brain connected therewith. In the caterpillars
of this filial generation the brain finds therefore the embryonal
discs which are more physiologically advanced and this fact
eventually leads to a certain abbreviation of the pupal stage of
the progeny of starved females or of starved males.

On surveying the results of the foregoing inquiry the following
summary may be given :

1. The females of the moth Lymantria dispar L. derived from
starved caterpillars mated with control males laid smaller
quantities of eggs than control specimens, but the dimensions
of the eggs and the weight of the pupa? developing from them did
not undergo any essential changes. The capacity of the egg to
develop and the mortality of caterpillars as well as of chrysalids
are not changed by the starvation applied in these experiments
as compared with control material.

2. The morphological structure of the spermatozoa of males
deriving from similarly starved caterpillars is unchanged and
they have the normal capacity for stimulating eggs to develop-
ment, but the mortality of the resulting caterpillars and chrysalids
is distinctly larger than that of the control material. The weight
of the pupa? in the progeny of starved males undergoes essential
decrease.

3. Distinctly injurious effects were consequently brought about
in the offspring by inanition of the males, in contrast to inanition
of the females. The causes of the different behavior of either of
the starved parents ought to be referred to metabolism probably

INFLUENCE OF STARVATION OF INSECTS ON OFFSPRING. 33

different in the two ~.-\es. Changes elicited in the structure of the
spermato/o.i by starvation of males are most probably qualita-
tive.

4. The prolongation of larval life which had been noticed in the
starved male as well as female caterpillars could not be a- ;ied

in their offspring. The pupal stage, on the contrary, whirh had
Lien much shortened in starved specimens underwent in the
priny of -t.irved females as well as in the offspri- starved

in les an analogous change which leads to acceleration of nu-t.i-
morphosis. Such behavior of the progeny of starved specimen-
.supports the .mihor's former opinion that insect metamorphosis
is checked l>\ the secretoric function of the larval brain.

BIBLIOGRAPHY.

.<•<! by an asterisk arc known to me only from abstracts.

* Cole, L. J., and Bachhuber, L. J.

'14 The Effect of Lead on the Germ Cc of the Male Rabbit and Fowl as In-
dicated by their Progeny. Proc. Soc. Exp. Biol. Med.. Vol. 12.

* Guyer. M. F., and Smith, E. A.

'18 vrin-s on Cytolysins. I. Some Prenatal Effects of Lens Antibodies.
Journ. of Exper. Zool.. Vol. 26.

'20 II. Transmission of Induced Eye Defects. Ibidem, Vol. 31.
Kellogg, V. L., and Bell, R. G.

'04 V. ni.ttions Induced in Larval, Pupal and Imagina! Stage of Bombyx >•
by Controlled Varying Food Supply. Science, ' • >!. 18.

Stanford Junior Publications, University Series, No. I.)
Kopec, S.

'17 Experiments on Metamorphosis of Insects. Bull. Acad. • vie.

'21 L'inlluence de {'inanition sur le d< a et la • e la vie

Insects. Mvm. de 1'Inst. Nat. Polonais d'Econ. i . Vol. i.

'22 Studies on the Necessity of the Brain ta-

morphosis. BIOL. BULL., Vol. 42.

"24 Studies on the Influence of In. ! the Dura-

tion of Life in Insects. BIOL. Hi i i ., \
*Miescher, F.

'97 Die Histochemischen und l'h\ hen Arbeiten. I .< ipsic.

Stockard, C. R.

'12 An Experimental Study o! Is Treated with

Alcohol. Arch. Intern. M

'13 The Effect on the Offspring of Intoj and the Trans-

mission of tin- I ' lions. Amer. Nat.. Vol. 47.

Stockard, C. R., and Craig, D. M.

'12 An Experimental Study ni tin- Ini: ,ind

the Developing Enihry- ol Mammals. A:> h- i. Entw. M--« li.. \'-l. 35.

34 STEFAN KOPEC.

* Stockard, C. R., and Papanicolaou, G.

'18 Further Studies on the Modification of the Germ Cells in Mammals. The
Effect of Alcohol on Treated Guinea Pigs and their Descendants. Journ.
of Exper. Zool., Vol. 26.

* Woltereck, R.

'08 Uber natiirliche und kunstliche Varietatenbildung bei Daphniden. Vern.
Deutsch. Zool. Gesellsch.

'09 Weitere experimentelle Untersuchungen iiber Artveranderung, speziell
iiber das Wesen quantitativer Unterschiede der Daphnien. Ibidem.

'n Uber Veriinderung der Sexualitat bei Daphniden. Intern. Revue d. ges.
Hydrobiol. u. Hydrographie, Vol. 4.

'n Beitrag zur Analyse der Vererbung erworbener Eigenschaften Transmuta-
tion und Praeinduktion bei Daphnia. Verh. Deutsch. Zool. Gesellsch.

A NTAY IIKI.D METHOD OF INVESTIGATING THE

HVDROTROPISMS OF FRESH-WATER

INVERTEBRATES.

%

C. H. TURNER.'
TVER TEACHERS' COLLEGE. ST. Loris. Mo.

INTRODUCTION.

During the past quarter of a century the tropism idea has
stimulated intensive investigation of the factors that serve as
directive forces in the behavior of animals. Gravitation, light,
heat, contact, wind and water currents, as factors of orientation,
have been much investigated; but, the directive influence of
bodies of water as such has received scant attention. In her
learned textbook,3 Professor Washburn does not discuss the
subject. In his book on animal intelligence1 Professor Holmes
devotes a whole chapter to tropisms. He discusses; chemotaxis,
geotaxis, thigmotaxis, rheotaxis, phototaxis, thermotaxis and
electrotaxis; but, nut a wurd about hydrotaxis. Professor
Botivier. in a recent work,1 says more on this topic than I have
seen in any other book on animal behavior. He writes:

"No less than heat, water is necessary to living beings, for it
constitutes the greater portion of their protoplasm and plays a
part in almost all their internal changes. Also, all organisms
are sensitive to variations of humidity in the space surrounding
them, and with a great number this sensitiveness takes the form of
a directive orientation which is called hydrotropism.

"We know the famous experiments made by Stahl. in 1884,
on the hydrotropism of the fungi of the genii- .Ktlmlinm (tanning
fungus). The plasmodial mass of thr-i- plants during the vege-
tation period, enters the tan, making fur the humidity necessary

I. Xote by the editor. This parx-r v. >r luiMicaium J.; s. 1923.

The author died February 14. 1923. Dr. Turn, t tudent ol P -or C. O.

Whitman, and received the degiee of DC ;li* University of

1 • .IKO in 1907. His thesis was on "Tin- Homini; <>f Ants: An Experimental
Stinly .-I Ant Behavior." He coiuimn-d his putimt. th i imlies of an

urior to the time of his death, as the present o>iHrii>uti<>n will t«--tify. Several
by Ur. Turner have api* .ufl in tin- p.i.yrs .>i the Hi' >i.< '<.!• AL BULLETIN.

36 C. H. TURNER.

to it, and remounts to the surface in a dry milieu when it is going
to form its spores which serve its multiplication. Its hydro-
tropism is positive. This is the case with the beetles of the genera
Ilaliplus and Hydroporus. Wheeler (1899) had taken from a
pool a tuft of aquatic plants where these insects swarm. He
says: 'As soon as the beetles could come out and disengage
themselves from the plant they turned, with a common accord,
toward the sea and to it directed their steps. As this was a
distance of about twenty feet, the little creatures could not see
the water, and I was led to believe that they had some means of
perceiving a source of moisture and acted accordingly.

"Aquatic bugs act the same way when they are taken from
the place in which they live, and we know that the land crabs go
a long distance to water when they are ready to place their
progeny. The proper degree of humidity differs, moreover, with
different species. Wheeler reports that Bembidium, Elaphrus,
Omophron, and other small Coleoptera which bury themselves
in the sandy beaches, leave their burrows and come out into the
open air when one throws a little water on their strand. This is
negative hydrotropism. It is well known to collecting ento-
mologists, who use it in making captures."

Weiss 4 reports two cases of what he calls positive hydrotropism.
When specimens of the wingless Gerris marginatus were removed
one to nine yards from the pond they immediately returned to
it. When removed ten yards from the water they had some trou-
ble in getting started in the right direction; but finally reached
the pond. Thirty yards from the water they seemed to be lost.
The case of Dinentes assimilis, a winged beetle, is even more
interesting. When removed nine or ten feet from the pond it
tried to walk to the pond, then arose and flew directly to it.
When removed to a distance of seventy-five feet, it walked
about in all directions, then arose, on its wings, to a height of
twenty feet and flew directly to the water. When removed
half a mile from the pond, it soared in a widening sub-spiral to a
height of seventy-five feet and then flew off in the direction of the
water. He does not know whether it reached the water or not.

It is not claimed that the above resum6 contains all that has
been written on this topic. It is quite likely that some articles
have been overlooked. However, the fact that a diligent search

HYUROTROPISMS OF FRESH-WATER INVERTEBRATES. 37

has -upplied only thi- suggests that very little attention has been
paid to the topic. This may be because no simple means of
iM\e-tiu,itin^ it has been published. The purpose of this papi-r
o supply that lack and to apply the method to the study of a
few forms selected from different strata of the water. As types
• .I" form- tint creep along the bottom of the water or along vege-
tation but <!<> not leave the water, I selected a snail [Planorbis
< I/i-lisoma') antrosnm Con.] and a dragon-fly nymph. As ex-
amples of form- that spend most of their time on the bottom, but
which, at times, leave the water, I have sleeted the crayfish
[Cambarns (Faxon i us) propinquns Girard] and a scaver
beetle (Tr<>/>i Remits sp. ?). As types of forms that live on the
surface of the water, I have selected the giant water st rider
(Gerris remigis Say), which, except during the hibernating season,
is practically a permanent inhabitant of the film, and the whirl;
beetle (Gyrinns sp.?) which occasionally dives and which mi-
cs from pond to pond.

TECHNIQUE.

To be able to study the movements of animals with a cer-
tainty that the movements were not influenced by either gravi-
tation, the sun's rays, chemical stimuli or contact stimuli, I
devised what may be appropriately called a checker-board
plate. This consists of a board twenty inches wide and twenty-
four inches long which is covered with white oilcloth that has
been subdivided, by printed lines, into one inch squares (Fig. l).
The four edges of the board were bisected by straight lines which
crossed, at right-angles at the center of the board. Diagonal
lines, bisecting these central right angles extended outward to
the edges of the board. The tips of the four lines mentioned
were named; o degrees, 45 degrees, 90 degrees, i;vs decrees, 180
degrees, 225 degrees, 270 degrees, 315 degrees. By means of a
level and leveling devices this board wa- made perfectly hori/on-
tal. Being horizontal, gravitation could not influence the move-
ments. Being an absolutely smooth surface, the movements
could not be influenced by contact -timuli. I'.ein- c« .\en-d with
oil-cloth, it was possible, by \va-hin-, to keep ii free from chemical
-tiinnli. It was protected from the sun by means of a screen of
-ort; this made it impossible for the -un'- ray- to influence

38 C. H. TURNER.

the movements. The checker-board pattern made it easy to
watch the movements of the animals.

Using a scale of one-eighth of an inch equals an inch sheets
of cross section paper were cut to resemble the board, and the
tips of radii labeled in the same manner as on the board. These
were carried to the field in a loose-leaf cover.

g

-

n

<

;:

!

/

\

/

/

7P

\

1

~i

jl

/

'

!

$

(

V>

FIG. i. Checkerboard Plate.

At the beginning of a series of experiments the board is placed
on the ground and adjusted until it is perfectly horizontal. If
necessary, it is screened from the sun. One of the sheets is
then numbered to correspond with the specimen to be investi-
gated. On the reverse side is recorded the position of the sun,
the position of the water and the direction in which the animal's
head is to be placed. In describing these positions the labeled
tips of the radii are used as reference points. The animal is
then placed in the center of the board with its head in the de-
sired direction and its movements traced on the cross-section
paper. When the animal reaches the edge of the board, it is
returned to its original position. This is repeated until it has
reached the edge of the board five times. It is then replaced in
the water a short time and then returned to the center of the
board with its head facing a new position. It is allowed to make
five trips in that case; these are recorded on a new sheet. This

BYDROTROPISMS OF FRESH-WATER INVERTEBRATES.

39

is repeated over and over again, using one sheet of paper for
each five experiments, until the creature has worked with its
head fari 11:^ all eight of the radii tips. This gives forty experi-
ment- and i- called a series. Work is continued until one has
secured record- of as many series as is thought necessary. In the
work recorded in this paper an attempt was made to secure
between twelve and thirteen series with each species investigated.
In practice it is not always possible to secure a complete series
•n an individual. Sometimes the individual would escape
before the series was completed, at others fatigue would put an
end to the experiments at an early stage. In some cases it was
necessary to return the specimen to the water at the close of
each experiment, instead of at the close of each five. Two invecti-
iMtor- (.in do more effective work than one; one giving his en-
'''••• time to looking after the animal and the other recording what
happen^. This is especially important when dealing with forms
that fly and with crawling forms that move rapidly. After a
little experience one recognizes the movements preliminary to
flight. I'sually a small glass cover placed over the specimen for a
moment will check the attempt to fly.

In the work performed in connection with this article, on
returning from the field the results of the days work were re-
corded on tables with the following captions;

TAHU-: I.

0

•u

Final Movement.

tn

E

£

rt

"5

,.

U

K

0

in

(9

0

u

0

ed

-a

i

u
^rt

t/}

1

E

3

Position ol

Posit i< in nl

ead Pointed

Number of

X

w

—
0

3

o
H

•JO)EM
par , |

—

' -.
- -

z

I"1'

•o
rt

SJ

o
Z

~

K

M

*

R

Z

*

Later these tables were condensed into those published in this
article.

In the experiments recorded in thi- article the side of the

4O C. H. TURNER.

board marked O, always faced the water. If there had been an
unlimited number of experiments, twenty-five per cent of
movements toward the water would indicate that the movements
were random and not influenced by the position of the water.
Since the number of experiments was limited to about 500,
unless the movements toward water were much more than twenty
five per cent, the movements were considered random. In
recording movements toward the position of the sun and the
initial position of the insect's head, eight possible directions
were considered. In each of these cases, if there had been an
infinite series of experiments, 12.5 per cent, of movements in a
certain direction would indicate random movements. Hence,
since there were only about 500 experiments, unless the per cent
of movements in a certain direction was much more than 12.5
the movements were considered random.

With the animals investigated in this connection, in addition
to the experiments with the checker-board plate, the creatures
were set free, at definite distances from the water, and their
movements watched. If the experiments demonstrated a tend-
ency to move toward the water, then the specimens were blinded
and the experiments repeated.

DRAGON-FLY NYMPHS (PLATE I).

In experimenting with the checker-board plate fourteen
individuals were used. In some cases the board was ten feet from
the water; in others it was fifteen. Five hundred and fifty-six ex-
periments were performed. The results are recorded in Table II.
In an infinite series of experiments, if the factors considered
did not influence the movements in any manner, 25 per cent, of
the movements should be toward water, 12.5 per cent, toward
the position of the sun and 12.5 per cent, to the way the head
originally pointed. The actual results were 25 per cent toward
water, 20.5 per cent, toward the position of the sun and 23.3 per
cent, toward the way the head originally pointed. Only tlmv
individuals made more than thirty per cent, of movements to
water. Evidently, dragon-fly nymphs, when moving about on
the checker-board plate are not influence by the nearness of
bodies of water.

IIVDROTROPISMS OF FRESH-WATER INVERTEBRATES. 4!

TABLE II.

DRAGON-FLY NYMPHS.

00

Per Cent, of Final Movement-.

1

"3

u

on

2

.

E

H

-

O

•3

a

.5

^

-;

'S

—

•a

.«

Q

O

lM

(2

•3 ^

u

—

RJ

K

'o

y

u

O

^^

£ £

~ .
« =

•: _•

.

|

y

•o

.2

2 ~.

— ^ > ^

£ ,=

=

E

'C

a

6

3

d

i

'c

• —

**

ro x

'o

f~

H

i

10

2

IO

IO

90

30

70

A.M.

2

3

I

O IOO IOO

IOO

A.M.

3

:

1

IOO 0

IOO

A.M.

4

18

4 7111

33

P.M.

5

43

30

79

30

70

A.M.

6

5

I

o

IOO

o

IOO

40

60

P.M.

7

So

8

46

54

37

63

21

79

P.M.

40

Pond

8

7

93

o

IOO

25

75

P.M.

0

80

8 17 | s

28

72

Both

10

40

22

IS

85

A.M.

1 1

167

Pond

57

4.1

1 1

89

Both

12

10

Pond

IO

90

20

20

A.M.

13

25

Pond

5

12

88

48

52

24

76

P.M.

14

30

Pond

4

10

90

20

80

50

A.M.

Av.-

4.8

25

75

20.5

70.5

76.7

NORMAL CRAYFISH [Cambarus (Faxonins) propinquus
GIRARD] (PLATE II).

In experimenting with the checker-board plate fourteen in-
dividuals were uscd.varying in length from one and a half
inches to three inches. In most cases the board was pl.n«-«l
within five or ten feet of the water, in some cases it was forty
feet from the water and, in a very few cases, it \\.i- fifty I« < i
away. In most cases the experiments \\i-n- romluru-d lu-.ir tin-
river or brook from which the crayfish \viv ohtaiiu-d. In others
the crayfish were carried to the bonK-r- of some lake or pond.
The results are recorded in TabK- III.

In an infinite series of experiment- t\v« -nty-live per cent, of
movements toward the WUUT i> what >hoiil«l !>»• expected if the
water has not directive intluemv i»n the indi\ idnals. The ex-
periments yielded 67.5 of mo\munt- toward the water. Three
individuals made IOO per cent, of movement- toward the water,
four between 90 and loo per cent. In only four cases did the
per cent, fall under 50 and the lowest of those was 36. Evidently

C. H. TURNER.

this crayfish, when moving on the checkerboard plate, is largely
directed by the nearest body of water. It acts as though it
were positively hydrotropic.

TABLE III.

THE CRAYFISH [Cambarus (Faxonius) propinquits Girard].

Per Cent, of Final Movement^.

1

• en

C

c

.

E

Ii

4^

_o

fT?

3

H

CJ

1

Q

•a

u

—

rt

o

Q

C

"3

N

i r\

*o

$

0H

"8 -

> u

"O
u

CS

o "0

S -p

"o

V

^

VJ

_

« i

6 S

rt c

o c

^""^ -^J

K*-I -*j

o

.E

o

"O

.5

£ «

w -*-j

* 5

H C

JS C

S

u

%

,0

E

8

•4-1

'c

h^

0^

o a:

H

o

1^

J§

H

IT.

S

12

2

o

I

13

/4 in.

Brook

fem.

6

38

62

IS

85

23

77

A.M.

2

7

in.

Brook

fem.

2

IOO

o

0

IOO

0

IOO

A.M.

3

3

in.

Brook

fem.

I

IOO

o

0

IOO

o

IOO

A.M.

4

in.

Pond

male

8

41

59

i

99

II

89

Both

5

40

in.

Pond

fern.

8

97

3

o

IOO

0

IOO

P.M.

6

40

in.

Lake

fem.

8

75

25

30

70

22

78

P.M.

7

40

i in.

Lake

fem.

8

IOO

o

2

98

IS

85

P.M.

8

40

1 1/\ in.

Lake

fem.

8

65

35

13

87

8

92

P.M.

9

40

3 in.

Pond

male

8

52

48

0

IOO

28

72

P.M.

10

80

i H in.

Pond

fem.

8

91

9

3

97

23

77

A.M.

ii

7

ij^ in.

Pond

fem.

2

57

43

o

IOO

14

86

A.M.

12

80

2^/2 in.

River

fem.

8

56

44

7

93

35

65

Both

13

45

i in.

River

male

8

36

64

ii

89

29

71

A.M.

14

40

i in.

River

male

8

38

62

IS

85

55

45

A.M.

Total

535

Averag

p .

6.5

67-5

32-5

6.9

93-1

18.8

81.2

In an infinite series of experiments, 12.5 per cent, of move-
ments toward the position of the sun would indicate that bright-
ness of the surrounding region (the direct rays of the sun did not
have access to the specimens) did not have a directive effect.
The experiments yielded only 6.7 per cent, of movements to-
ward the position of the sun. Four individuals made zero per
cent, of movements toward the position of the sun, three less
than four per cent., in only four cases did the per cent, become
more than twelve. Evidently, on the checkerboard plate, the
crayfish shows a tendency to move away from tin- brightest part
of the lidd.

In an infinite series of experiments 12.5 per cent, of movements
in the direction the head originally pointed would indicate that
the position of the head had not intluenced the movements.

HVDROTROPISMS OF FRESH-WATER INVERTEBRATES. 43

These experiments yielded 18.8 per cent, of movements in the
direction the head originally pointed. This is so near to what
theory demands that it is evident that the original position of
the head has not appreciably influenced the final movements of
the individual.

In tin- matter- con-idered no marked difference was observed
In -t\\een male- and females.

At the close of each series of experiments mentioned above,
t In- crayfish wa- placed on the ground, with its head away from
the water. The distances from the water ranged from ten to
fifty feet. In each case the crayfish finally turned and went to
the water, even going around obstacles.

On another occasion twelve crayfish were placed, one at a
time, on the ground, within ten feet of the body of water from
which they had been obtained. They were always faced away
!n>m the water. Kleven of these reached the water in a very
few minutes; the twelfth roamed about for fully half an hour
before- it found the water. Considered alone these experiments
uonld not permit us to draw any conclusions. The land sloped
io\\.ird the water. The sun was shining. Hence gravity and the
-mi rays may have influenced the movements. However, the
results are in harmony with the conclusions drawn from the ex-
periments on the checkerboard plate; and, when considered in
conjunction with some experiments to be described in the next
section, they are illuminating.

Bi INDED CRAYFISH (Cambaraus sp.? AND Cambarns (Fraxoni
propinquus Girard) (PLATE III).

The above experiments with the checkerboard plate prove
conclusively that the nearest body of \\ater. in -onie manner,
forces the crayfish to move toward?- it. < >ne naturally \\onders
if the eyes function in this behavior. To te-t the matter three
types of experiments were conducted \\ith blinded na\ii>h.

In the first series thirteen blinded < Ta\ \\-\\ \\ere te-ied on the
checkerboard plate. The e\e> of li\e \\ere Minded with an
opaque varnish, the eyes of the other eight wvre amputated.
The board was placed twenty feet from the water. The ronlt>
are recorded in Table IV.

In an infinite >erie- of experiments, if the factors considered did

44

C. H. Tl FINER.

not influence the movements, 25 per cent, of the movements
should be toward water, 12.5 per cent, toward the sun and 12.5
per cent, toward the way the head was originally pointed. The
experiments yielded the following results; 25.7 per cent, toward
water, 15.3 per cent, toward the sun and 17.1 per cent, toward
the way the head originally pointed. Evidently the movements
of blind crayfish, on the checkerboard plate, are not influenced
by the nearby bodies of water.

TABLE IV.

THE CRAYFISH (Cambarus, sp.?) BLIND (PLATE III).

Per Cent, of Final. Movements

<U

•^2

Cfi

.3

.2

IH

•

E

h

(3

_5

4J

(f

•a

•o

03

a
Q

c
o

"o
o

o

N

c/5

"o

I

6

CM

•H ±

a v

i*

n! .

> -

0 £

—
rt C

L-

rt

S-g

o

ffi-
>,£

(4-1

0

O

8

ja

*O

[rt

* rt

> =

' 5

r^

« c

s

"C

E

C

'^j

0 >

^

0 CA;

^ C/^'

£?'3

•S'o

*

w

'5

H^

0 "^

H

4J

0

1^

c^

^

C/j

z

o

I

40

4 in.

Pond

fem.

8

12

88

0

IOO

IO

90

P.M.

2

20

4 in.

Pond

male

4

40

60

25

75

5

95

P.M.

3

40

4 in.

Pond

male

8

17

83

7

93

22

-

P.M.

4

40

4 in.

Pond

fem.

8

42

58

20

80

35

65

P.M.

5

38

4 in.

Pond

male

8

45

55

ii

89

21

79

P.M.

6

40

4 in.

Pond

fem.

8

25

75

15

85

IS

85

P.M.

7

40

4 in.

Pond

male

8

32

68

22

78

17

83

P.M.

8

40

4 in.

Pond

male

8

20

80

18

82

18

82

P.M.

9

80

4 in.

Pond

male

8

23

77

29

71

40

60

A.M.

IO

!

4 in.

Pond

fem.

8

42

58

13

87

15

85

A.M.

ii

65

4 in.

Pond

fem.

8

26

74

9

91

15

85

A.M.

12

10

4 in.

Pond

male

2

10

90

30

70

IO

90

A.M.

13

5

4 in.

Pond

fem.

I

O

IOO

o

IOO

0

IOO

A.M.

Total

498

Averag

?

6.9

25.7

74-3

15.3

84.7

17.1

82.9

These thirteen blinded crayfish were placed on the ground,
within fifteen feet of a pond. Their heads were faced away from
the water. The land sloped gently towards the water and the
sun was shining brightly. At the end of an hour none had reached
the water. Several had wandered off into the shrubbery and
become lost, three had moved to within two feet of the water and
then turned and walked off in another direction, some were
wandering about at a greater distance from the water than their
original position and three were in practically their original
positions.

In another series of experiments the crayfish [Cambarus
(Fraxonius) propinquus] were used in pairs. The eyes of one

HVDROTROPISMS OF FRESH-WATER INVERTEBRATES.

45

of each pair were painted with a thick coat of lampblack mixed in
liquid glue. The eyes of the other were normal. Twelve pairs
were used. They were placed fifteen feet from the water. In
ten cases the one with good eyes reached the water in a short
time and the one with blinded eyes failed to do so. In one case
neither reached the water. In another the one with blinded
eyes reached the water and the one with good eyes wandered
iv from the water. The experiment with this last couple was
repeated and yielded similar results.1 It is probable that the
eyes of this blinded crayfish were not entirely blind; because, on
tin- next day, after the eyes had been repainted, this individual
behaved like the other blinded ones. Its companion of the first
day died over night; hence, it was not possible to re-test it.

The-e experiments with blinded crayfish, when considered in
connection with the experiments with normal crayfish, demon-
strate that the stimulus by means of which a body of water
controls the movements of the crayfish reach the animal through
the eyes.

GIANT WATER BUG [Belostoma (Zaitha) fluminea] (PLATE IV).

The exceptionally dry summer caused many ponds to go dry
and it was impossible to secure enough specimens to conduct an
extensive number of experiments. Five individuals were used
and 109 experiments performed. The checkerboard plate was
placed ten feet from the water. The results of these experiments

are recorded in Table Y.

TABLE Y.

GIANT WATER BUGS [Relo^ioma ('/.

hi

•

1

tn

u

i.

•

E

3

o

m

3

"J^

td

_

u

•u

Q

H

flj

"£.

en

1

—

c

b

-g

1
I

*—
S

"i

a

"c

"w

£

dt

3

z

—
?

—

0

i 1

o
ii

S

H

(fl

2

x

I

8

Adult

Pond

s

3°.

o

23

77

P.M.

2

10

Juv.

Pond

go

0

IOO

IO

go

P.M.

3

40

Adult

Pond

S

95

30

70

I'M.

4

ii

Juv.

Pond

o

36

64

P.M.

S

40

Juv.

Rivr

8

32

68

2

98

0

IOO

A.M.

Total

IOO

A.VCJ

/17

=;i

1 !

10.8

1 In tl, i- tw.-i

one pair while 1 \vatrhcd thr "tl

w. i • a tinn.-; my wife watching

46

C. H. TURNER.

It is probably unwise to draw conclusions from only a hundred
experiments; but, these experiments suggest that the water does
have a directive influence upon this giant water bug.

A SCAVENGER BEETLE (Tropisternus) (PLATE V).

In experimenting with the checkerboard plate nineteen in-
dividuals of this species were induced to perform 571 experiments.
In most cases the board was ten to twenty feet from the water.
In one case it was fifty feet (number 18); in another case it was
100 feet (number 19). The results of the experiments are recorded
in Table VI.

TABLE VI.

A SCAVENGER BEETLE (Tropisternus) — ADULTS.

;J

Per Cent, of Final Movements.

O

V*

*r

.0

is

c.

C

^'

B

3

H

rt

+j

•a

•o

al

nJ

a

r^

"o

*

0

•a ,•

14

a

•a

Is

•a

V

K-6

V

IH

0
i

O

"5

is

l-H *•

nj £

H 1

i-

i §

o c

K 1

HI

_E

'o

E

.£5

*j

® <

*j fe-

0 0)

' C/3

c""o

^ 'o

0.

3

W

c

H ^"

o >•*

h

o

K» ^

*_> CH

w

*

O

I

7

Lake

5

87

13

o

IOO

13

87

A.M.

2

2

Pond

i

50

50

0

IOO

O

IOO

A.M.

3

9

Pond

i

55

45

II

89

II

89

A.M.

4

18

Pond

3

7

93

o

IOO

50

50

P.M.

5

40

Pond

8

3

97

3

97

13

87

1' M.

6

40

Pond

8

53

47

53

47

20

80

P.M.

7

8

Pond

I

25

75

25

75

88

12

P.M.

8

40

Pond

8

o

IOO

o

IOO

23

77

P.M.

9

40

Pond

8

o

IOO

65

35

25

75

P.M.

10

40

Pond

8

35

65

o

IOO

8

92

P.M.

ii

40

Pond

8

55

45

15

85

28

72

P.M.

12

9

1 Vn.l

2

1 1

89

o

IOO

44

56

A.M.

13

40

Pond

8

45

55

5

95

40

60

P.M.

14

10

Pond

2

10

90

0

IOO

40

60

A.M.

15

40

Pond

8

37

63

20

80

18

82

A.M.

16

80

Pond

8

6

94

8

92

38

62

\ M

17

40

Pond

8

47

53

10

90

15

85

P.M.

18

40

Pond

8

45

55

0

IOO

20

80

I'M.

19

28

Pond

4

53

47

M

86

20

7i

P.M.

Total

571

Av.- .

5-1

32.3

67.7

12

88

27

73

In an infinite series of experiments, if the factors considered
did not influence the movements, 25 per cent, should have been
toward the water, 12.5 per cent, toward the sun and 12.5 per
cent, toward the direction the head originally pointed. The

HYDROTROPISMS OF FRESH-WATER INVERTEBRATES. 47

results obtained from the experiments were; 32.3 per cent,
inward the water, 12 per cent, toward the sun, and 27 per cent.
inward the direction the head originally pointed. The per cent
of movements toward the water is too low to warrant the assump-
tii.n that the movements of the group, as a whole, have been
influenced by the water; however, certain individuals exhibit a
marked tendency to move toward the water. One individual
made 87 per cent, of movements toward the water; five, between
50 and 60 per cent.; three, between 40 and 50 per cent. Thus
nine individuals, practically half of the individuals used in the
experiments exhibited a tendency to move toward the water.

Bright ne— did not influence the movements. There was a
-li-Jn tendency to move in the direction the head was originally
pointed.

When these beetles were given their freedom, some flew and
others did not. Of those that flew, more than 90 per cent, flew
to the water.

SMM.I. |)i\i\<; BKKTLE (Laccophilns sp.?).
To test their fitness for work on the checkerboard plate, a few
experiments were conducted with a small diving beetle. On
account of its proneness to fly it is a difficult subject with which
to work; but, where one has sufficent patience it may be used.
It might be a good idea to clip its second pair of wings. On
escaping they always flew toward the water.

THE I.AR<;K WATER STRIDER (Gem's renrigis Sa I'I.ATE \ I

In working with the checkerboard plate, 27 indi\ iduals were
induced to perform 534 experiments. Seventeen individuals
were used near the stream where they were captured: tin- re-
mainder were taken to the shore of a lake. In nn>-t cases tin-
board was placed from five to ten feet fn-in tin- \\ater. In a few
cases it was necessary to place it nu-nty feet away. The results
are recorded in Table \ 1 1.

In an infinite series of experiment-, if the factor- considered,
have no directive influence on tin- animal-, then J5 per cent, of
the movements should be toward the water. u.,s per cent, to-
ward the sun and 12.5 per cent, toward the origin. il po-iiion of
the head. The experiment- yielded 34.7 per cent, of movements

48

C. H. TURNER.

toward the water, 16 per cent, toward the sun and 26.6
per cent, toward the initial position of the head. The low per
cent of movements toward the water is too low to warrant the
conclusion that the group, as a whole, is directed, in its move-
ments, by the nearest body of water. However, certain indivi-
duals are undoubtedly so influenced. Two individuals made 100

TABLE VII.

WATF.R-STRIDER (Gerris remigis Say).

Ui

Per Cent, of Movements?.

13

1-1

V

CO

C

H

&

•2

T3

£

<y

^

CO

•a

•a

•a

a

ai

C

u

60

o

&

"H K

c3 ^

•e

te

ri _:

<u ~

ffi-d

4>

a

(11
&

T3

is

g.£

fe ri

£ 0

o •<->

rt S
& ~

0 g

K -

rt S

'o

6

• £H

+J

b>

K*

0 C/3

C/}

>> *o

J> 'o

a

CO

3
S

w

'3

t— 1

H15

'o "*

H

"o

gft,

-J ^
O

!5

I

IO

Adt.

Br.

3

80

20

o

IOO

30

70

2

5

Adt.

Br.

4

40

60

o

IOO

40

60

3

i

Juv.

Br.

i

o

IOO

o

IOO

o

IOO

4

12

Juv.

Br.

8

33

67

I

99

50

50

5

2

Adt.

Br.

2

o

IOO

o

IOO

0

IOO

6

II

Tuv.

Br.

5

9

91

o

IOO

27

73

7

II

Adt.

Br.

6

9

91

0

IOO

27

73

8

4

Adt.

Br.

I

25

75

50

50

o

IOO

9

2

Adt.

Br.

i

IOO

0

o

IOO

o

IOO

10

12

Adt.

Br.

3

0

IOO

50

50

o

IOO

ii

2

Adt.

Br.

i

50

50

50

50

o

.IOO

12

2O

Juv.

Br.

4

30

70

25

75

50

50

13

I

Juv.

Br.

i

IOO

o

o

IOO

0

IOO

14

18

Adt.

Br.

3

II

89

89

II

17

83

15

9

Adt.

Br.

2

44

56

22

78

22

78

16

20

Adt.

Br.

4

20

80

50

50

50

50

17

6

Adt.

1 .1 1 ,

2

50

50

17

83

O

IOO

18

28

Adt.

1 .(!..•

6

36

64

14

86

61

39

19

40

Adt.

Lake

8

43

57

0

IOO

33

67

20

40

Adt.

Lake

8

25

75

7

93

52

48

21

40

Adt. '

Lake

8

52

48

3

97

35

65

22

40

Adt.

Lake

8

18

82

2

98

40

60

23

40

Adt.

Lake

8

57

43

25

75

32

68

24

40

Adt.

i , : -

8

32

68

5

95

40

60

25

40

Adt.

Lake

8

i?

83

8

92

33

67

26

40

Adt.

1 : .

8

30

70

7

93

42

58

27

40

Adt.

Lake

8

25

75

8

92

35

65

Total

534

Averag

e ....

5

34-7

65-3

16

84

26.6

73-4

per cent, of movements toward the water; one, 80 per cent; four,
between 50 and 60 per cent. On the other hand it must be noted
that three individuals made 100 per cent, of their movements

HYDROTROPISMS OF FRESH-WATER INVERTEBRATES. 49

away from the water; three, between 90 and 100 per cent.; four,
between 80 and 90 per cent.; five, between 70 and 80 per cent.;
four, between 60 and 70 per cent.; four, between 50 and 60 per
cent. Kvidmtly, although some individuals are undoubtedly in-
fluenced by water as a directive force, this is certainly not true
of the group as a whole.

Near a small pond in Carondelet Park, St. Louis, Mo., there
is a narrow valley. The floor, which is nearly horizontal is about
60 feet long and 20 feet wide. About eight feet from the pond it
slopes, at an angle of 60 degrees to the water. On the two sides
the walls slope upward at an angle of from 30 to 50 degrees. This
valley floor was selected as the place in which to perform the
following experiments with water-striders which had been cap-
tured in a brook near Creve Coeur Lake, Mo. One at a time the
water-striders were placed on the ground, at a certain definite
di-iance from the water. They were always faced away from
tin- water. They were watched continuously until they had
rit her reached the water or had gone from the starting point,
in -oim- direction other than toward the water, a distance equal
to that from the starting place to the water. Such individuals
were recorded as not moving toward the water. Sometimes an
individual would start in a certain direction and continue in
that direction to the close of the experiment. More often it
wandered about in various directions before settling down to
continue in one course. Sometimes an individual would move
toward the water a distance of ten to twenty feet and then
turn about and move away from it. In a few cases an indi\ idiial
would move away from the water a distance of ten to til'u-en iVrt
and then turn, meander about, and finally reach the \\.iitr.
Several of those that did not reach the water rlimU-d tin- -idr-
of the valley for a distance of fifteen to twenty feet. If. in its
wanderings, an individual happened to arrive on the steep slope
that extended from the horizontal floor of the valley to the pond,
it invariably continued, sometime.-, \\iih accelerated -peed, to
the water. The results are recorded in Table VIII.

These results harmonize with the conclusions l>a-rd upon ex-
periments with the checkerboard plate.

C. H. TURNER.

TABLE VIII.

Number of Individuals
Used.

Number of Feet
Placed From Water.

No. that Moved.

Per Cent that Moved.

I 1

Not Toward
the Water.

- £

•o .

£ "

20
IO
10

10

20
30

6

2
2

14
8
8

30
20

20

70

80

80

THE WHIRLIGIG BEETLE (Gyrinus sp.?) (PLATE VII).

In working with the checkerboard plate twenty-one individ-
uals were used. With the board from ten to fifteen feet from
the water, 551 experiments were performed. The results are
recorded in Table IX.

In an infinite series of experiments, if the factors considered
have no effect of the movements of the creatures, then 25 per
cent, of the movements should have been toward water; 12.5
per cent., toward the sun; 12.5 per cent, toward the original
position of the head. These experiments yielded 52.6 per cent,
of movements toward the water, 6.7 per cent, toward the posi-
tion of the sun and 22 per cent, toward the original position of
the head. In one individual the per cent, of movements toward
the water was 95, in four cases it was between 80 and 90, in
three cases it was between 60 and 70, in four cases it was between
50 and 60, in five it was between 40 and 50. In the case of only
three individuals was the per cent, of movements toward water
less than 30. Evidently, the whirligig beetle, when moving on
the checkerboard plate, is largely directed by the nearest body of
water. They act as though they were positively hydrotropic.

In addition to the work upon the checkerboard plate, tin-
following experiments were conducted in the open. Twelve
specimens were placed in a dry tray ten feet from the water. In
less than half an hour all had flown to the water. Ten spin inu-ns
were placed in a dry tray twenty feet from the water. In less
than half an hour all had flown to the water. Ten specimens

HYDROTROPISMS OF FRESH-WATER INVERTEBRATES.

TABLE IX.

THE WHIRLIGIG BEETLE (Gyrinus sp. ?.) — ADULTS.

Per Cent, of Final Movements.

1

u

Qj

s

.

c

h_

—

-

£

.c

•^

—
ca

a
Q

Z

IM

r*

2?

2

£ .

•o

o

0

C

1

~=

£

PL*

"ljs

~E r' -t '- "H .

- - 2s

1 ^ H £ ! ?.

•— ••* "? ••* i I™

pd

|§

• O)

o

glj
— -
M —

>

- £

u
H

C/3

Z

z

z

? "~

'o

Z

I

5

Pond 1 i o

IOO

0

IOO

A.M.

2

35

Pond

7

60 40 6

94

94

A.M.

3

17

Pond

4

88 12 0

IOO

29

71

A.M.

4

Pond

I

O IOO

IOO

IOO

o

A.M.

5

9

Pond

2

1 1 89

IOO

IOO

A.M.

6

5

Pond

I

IOO

IOO

20

A.M.

7

40

River 8 7

27

73

A.M.

5

River 1 40 600 100

IOO

A.M.

9

40

River 55 45 15 85

13

87

A.M.

10

17

River 3 37 63 5 95

12

P.M.

1 1

40

River 8 87 13 o 100

13

87

A.M.

12

41

River 56 44 100

20

80

A.M.

13

13

River 3 77 23 o 100

8

92

A.M.

M

40

River 8 30 70 15 85

IS

85

A.M.

15

40

River 8 88 i .' 3 97

IOO

P.M.

16

41

River 7"

17

83

1- M.

17

41

River

8

58 42 2

98

36

64

P.M.

18

40

River

8

43 57 15

85

15

85

P.M.

19

40

River

87

7

93

P.M.

20

20

River 4 95 5

IOO

20

80

P.M.

21

2O

River 4 45 55 30

70

5

95

P.M.

Total

551

Av...| 5.2 | 52.6 47.4 6.7 | 93.3 | 22 78

were placed in a dry tray forty feet from the water. In less than
half an hour all had flown to the water. In most cases they did
not ascend high in the air; but the further back they were the
higher they ascended. In one case a sidewise flight carried t he-
beetle behind a tall tree. In that case it arose to above the top
of the tree before turning toward the water.

At 3:15 P.M. the eyes of ten individual-; were blinded by paint-
ing them with lampblack mixed in liquid :Jue. The-e whirlL
were placed in a dry tray forty tVet from the water. Ten normal
individuals were placed in a similar tray, at the -ame di-tance
from the water. The two trays were only t\\«> feet apart. One by
• >ne, the individuals with normal >ight an>-e fn-m the tray and
tlew to the water. By 3:30 P.M. all but <»ne had reaehed the
\\.iter. At 4:00 P.M. the last one tlew to the water. Up to

52 C. H. TURNER.

4:15 P.M., when I stopped for the day, none of the blinded
whirligigs had reached the water. Two had become lost in the
grass. Eight were still in the tray.

These experiments show conclusively that the nearest body of
water exerts a directive influence upon the movements of the
whirligig beetle and that the control is exerted by means of
stimulus received through the eyes.

CONCLUSIONS

i . The checkerboard plate consists of a smooth board covered
with white oilcloth which is subdivided, by printed lines, into
one-inch squares. There is a leveling device by means of which
it can be made perfectly horizontal. Considering the center of
the board the center of a circle, the board is divided into eight
sections, each forty-five degrees wide. Starting at the middle
of one side the tips of these dividing lines are numbered from o to
315. This nomenclature makes it possible to arrange the speci-
men at the center of the board, with its head pointing in a certain
definite direction. When in use, by means of an adjustable
screen, the board is protected from the direct rays of the sun. It
is washed from time to time to free it from odors that may in-
fluence movements. Thus manipulated all possibility of geotac-
tic, chemotactic, and thigmotactic responses are eliminated;
hence this device forms an excellent means of investigating the
influence of bodies of water upon the behavior of small animals.

1. It may be used in studying any walking land insect, milliped,
centiped, crustacean, or annelid.

2. It may be used in studying flying insects that also walk;
but, in some cases it may be necessary to clip the wings.

3. It may be used in investigating the behavior of many fresh-
water invertebrates (small snails, amphipods, isopods, crayfish,
some backswimmers, giant water bugs, water-striders, larvae of
straight-winged flies, dragonfly larvae, scavenger beetles, diving
beetles, whirligig beetles) ; in some cases it cannot be used be-
cause the animals cannot remain out of water long enough
(damsel-fly larvae); in yet others it cannot be used because the
creatures are too awkward when out of the water (large snails,
water-tigers).

BYDROTROPISMS OF FRESH-WATER IXVERTEBRA I 53

4. Although the matter has not been tested, it should be possi-
ble to use it in investigating many marine invertebrates.

5. This device may be used, not only with creatures with sound
vision; but, al>-o, with forms that have been rendered blind by
manipulation.

II. In testing this device several ecological types of inverte-
brates were used: those that creep along the bottom of the
body of water or along water plants, but do not leave the pond
or stream (Ascllns, half grown dragonfly larvae, pond snails
[these occasionally venture beyond the water edge]); those
that creep along the bottom or up vegetation and occasionally
leave the water (crayfish, scavenger beetles); those that live on
the film and never depart far from it (giant water striders);
those that dwell on the film, but migrate from one body of water
to another (whirligig beetles).

1. The movements of neither pond snails, asellids nor dragon-
fly larva.' are influenced by nearby bodies of water.

2. In the majority of cases the movements of crayfish are con-
n "lied l>y the nearest body of water. When out of the water
they behave as though they arc positively hydrotropic. Blind

h do not so react; this demonstrates that the movements
of this crustacean, when out of the water, are controlled by visual
-(iinuli furnished by the nearest body of water.

3. About half of the scavenger beetles investigated act as
though t! ively hydrotropic; the other half are not
influenced, in .1 directive manner, by nearby bodies of water.

4. When on land, the majority of the water striders investi-
gated are not influenced, in their movements, by the nearest
body of water; a few act as though they are p • ly hydro-
tropic.

5. When out of the water, whirligig beetles are undoubtedly
influenced, in their movements, by the nearest body of water.
On the checkerboard plate the majority always move toward the
water; when set at liberty within forty feet of a 1>"<1\ of water,
they always fly toward it. If blinded they do not react in this
manner. This demonstrates that they are controlled by visual
-liinuli furnished by the water.

6. This paper does not contend that the-e movements which
.-nine of the forms make towanl watt r are po-itive hyclrotropisms

54 C. H. TURNER.

in the Loebian sense. These investigations prove that the move-
ments mentioned are caused by the directive power of some body
of water and not by gravity, nor the sun's rays, nor contact
stimuli, nor odor trails. As to whether they are responses to
sensations or are merely tropisms it has nothing to say. Either
interpretation seems equally plausible.

7. Although the board was protected from the direct rays of
the sun, almost all of the species used exhibited a marked tend-
ency to move away from that portion of the board which was
nearest the position of the sun. Since the sun's rays did not
impinge on the individuals, this avoidance cannot be a negative
phototropism. It must be a form of differential sensitivity.
Perhaps it should be called negative photosensitivity.

REFERENCES.

1. Bouvier, E. L.

'22 Psychic Life of Insects, pp. 23-24.

2. Holmes, S. J.

'n Evolution of Animal Intelligence, pp. 11-60.

3. Washburn, Margaret Floy.

'17 The Animal Mind, Second edition.

4. Weiss, H. B.

'14 Notes on the Positive Hydrotropism of Gerris Marginatus Say and Din-
eutes assimilis Aube. Canadian Entomologist, Vol. 46, 1914.

BIOLOGICAL BULLETIN, VOL. XLVI.

PLATE I.

'

•

A70

.2.70

3.10

0

H

C. H. TURNER.

PLATE I. DRAGON FLY NYMI-HV SERIES OF 40 TRIALS.

BIOLOGICAL BULLETIN, VOL. XLVI

a

• 0

-• TURNER.

PLATE II. NORMAL CUAYRSH.

BIOLOGICAL BULLETIN. VOL. XLVI.

PLATE III.

C. H. TURNER.

PLATE III. Bi i

BIOLOGICAL BULLETIN. VOL. XLVI.

PLATE IV

C. H. TURNER.

PLATE IV. GIANT WATER I

BIOLOGICAL BULLETIN, VOL. XLVI.

PLATE V.

.270 10

3. 10

470

5.70

C. H. TURNER.

PLATE V. SCAVENGER BEETLE.

BIOLOGICAL BULLETIN, VOL. XLVI.

PLATE VI.

lit)

3.10

MO

3.10 10

-'70

C. H. TURNER.

PLATE VI. LARGE \YAII-.K Snu

BIOLOGICAL BULLETIN, VOL. XLVI.

PLATE VII.

C. H. TURNER.

PLATE VII. Tin: WHIRLIGIG BEETLE.

Vol. XLVI. February, 1924. Xo. 2.

BIOLOGICAL BULLETIN

MODIFICATION OF RESPONSE IN AMU.I'. V

S. O. MAST AND L. C. PUSCH.
FROM THE ZOOLOGICAL LABORATORY OF THE JOHNS HOPKINS VMVI .K-IIV.

Gibbs and Dellinger ('08) and Schaeffer ('17) maintain that
AiiKJL-ba has the power of selecting food. The former hold that
the kind of food preferred changes under various condition-, .unl
they suggest that changes in the kind of food selected is the re-
sult of a process of learning based upon the method of "trial and
error." One of us (Mast '10) in preceding observations found
that when the tip of a pseudopod of an amoeba comes in contact
with a highly illuminated region in the field it stops; then others
are sent out in succession in the same general direction, each one
stopping when it comes in contact with the light, until suddenly
one is sent out in a markedly different direction, avoiding the
illuminated region altogether. It is maintained that this abrupt
change in tin- direction of the formation of the pseudopods.
constituting a profound modification in the response to a given
stimulus, is dependent upon the preceding experience of the
individual, i.e. repeated contact with the strong light.

Tin- specimens used in the following experiments were obtained
from the cultures in the laboratory, most of which had been
-torked \\ith material from a stream near the campus. They
were of the proteus type. Each amoeba was isol.ited on a glass
slide in a drop of its culture-fluid under a cover-.^l.i — . tin ed^es of
which were sealed with vaseline. Each anm-li.i was adapted to
darkness before using.

The observations were made in tin- d.irk room. An area of
intense light was obtained by forii-iiii; on the >lidc under the
microscope, by means of the reflector .md ,m Abbe condenser,
an image of a portion of the luniim >us til.unent < >l .111 incandescent
Ma/da lamp placed in a light-tight 1>«\. I lie IM\ - «if the filament
-ed through a very narrow slit in the box. The image h.nl

55

56 S. O. MAST AND L. C. PUSCH.

clearly defined edges and furnished a band of intense light across
the field. The slide was adjusted so that the amoeba was within
a distance equal to approximately half its length from the light
with the pseudopod or pseudopods progressing towards it. If
the pseudopod did not stop when it came in contact with the
light the amoeba was discarded. If it stopped the behavior of
the amoeba was observed until it definitely moved away from the
light, and the number of pseudopods which made contact with
it was recorded. After this the light was shut off, and the amoeba
was left in darkness until the next trial was made. In this way
each individual was given twenty-seven trials.

The interval between trials was approximately three minutes
except in the following as recorded in Table I: individual B,
trials 16-17 and 23-24, 24 hours; individual D, trials 15-16,
22 hours; individual E, trials 8-9 18 hours, 16-17, 48 hours, and
21-22, 24 hours; individual F, trials 8-9, 50 hours, and 20-21,
20 hours; individual G, trials 16-17, 16 hours, and 24-25, 3 hours.
During these intervals the amoebae were kept on their slides in
darkness undisturbed.

The behavior of a typical individual is illustrated in Fig. I by
sketches representing all of the trials in the series. By referring
to this figure it will be seen that while during the first part of the
series there wras something of a persistence on the part of the
amoeba to attempt to proceed in its original direction — notably
7 attempts in trial 4 — there was but little in the latter part of the
series — a total of only 5 attempts during the last 14 trials.

The results obtained in all of the tests are presented in Table I.
This table shows that the number of pseudopods which came in
contact with the light before the direction was changed decreased
in general as the number of trials increased in each series, that
there was a similar decrease in the total contacts of all amoeba!
for corresponding trials, and also in the totals of consecutive
groups of three trials. Amoeba B made 10 contacts with the light
in the first three trials; with trial 4 a general decrease begins. A
total of 20 pseudopods made contact in the first 14 trials of the
series, and a total of 9 in the last 13. In the series of amoeba D
there were 5 contacts in the first 6 trials but none at all in the
remaining 21 trials. Amoeba E made 21 contacts in the first
14 trials — II of these were in the first 4 — and 5 in the last 13

MODIFICATION' OF RESPONSE IX AMCEBA.

57

2:03 2.Q3L 2:04 2:05"

1

tf "9-40*9

/a

10.2

IO:58 U.oo n. 01

ICOS" M:06 M:07

H:i3 i •• IMS' M.io l|:20 H-'

Vi:3 7 " It 3.? I

1 1 f C

IK44 l'4-«r

U..C7

FIG. i. A series of diagrammatic illustrating the U ha\ i. i . i Amoeba E for twenty-seven

tri.ils. Thr straight lines of equal Icii.uth i< nd of li^lit. I K.- ..utlines of the amoeba de-

si-rilit- it- j...-ition in relation to the light at the time given below in each individual sketch. The
arro\v< shcu tin- din-ction of protoplasmic -tn-aming.

S. O. MAST AXD L. C. PUSCH.

trials. Amoeba F made no contacts until trial 8; between trials
8 and 15 there were 8 contacts, while in the remaining 12 trials
there was but one contact. There was one contact in the first
trial of Amoeba G, then none until the 7th, when a reversal began
and persisted very generally through the iyth, by which time
23 contacts had been made; in the remaining ten trials there
were 8 contacts.

TABLE I.

RESPONSE OF AMCEBA TO CONTACT WITH A HIGHLY ILLUMINATED REGION IN

THE FIELD.

Number of Pseuddpoda that

Designation
of

Came in Contact with Light
in Each Trial.

Total Contacts
for All Individ-

Total Contacts for
All Individuals in

\ji
Trials.

uals in Each
Trial.

Groups of Three
Consecutive Trials.

B.

D.

£.

F.

C.

i

4

o

0

0

I

5

2

3

0

3

o

o

6

3

3

3

I

o

o

7

18

4

o

o

7

o

o

7

5

3

I

0

0

o

4

6

o

I

o

o

o

i

13

7

i

o

2

o

I

4

8

2

o

O

2

4

8

9

O

o

2

I

2

5

17

10

O

0

4

0

I

5

ii

3

o

2

2

2

9

12

o

o

O

O

o

o

M

13

0

0

O

I

2

3

14

I •

0

0

0

4

5

15

o

o

I

2

I

4

12

16

I

o

I

O

I

3

17

0

0

0

0

5

5

18

o

o

o

O

o

o

8

19

o

o

I

0

o

i

20

o

o

o

O

I

i

21

2

o

2

o

0

4

6

22

O

o

O

I

o

i

23

I

o

O

o

3

4

24

O

o

O

o

3

3

8

25

4

0

O

o

I

5

26

I

0

0

0

o

i

27

0

0

O

o

o

o

6

The totals by groups of three consecutive trials fluctuate
between 18 for the first group, 12 for the middle and 6 for the
last. In other words there is a modification of response in the
sense that the number of attempts to continue in the original
direction after meeting the light decreases in general as the
number of trials of the series increases.

MODIFICATION OF RESPONSE IN AMCEBA. 59

It will be remembered that each series was interrupted ir-
regularly l>v longer intervals of time, usually about a day or
longer. Despite this there is a general constancy of the tendemy
throughout each series. This seems to indicate ability to retian
for some time the modification of response.

These experiments were repeated 18 months after the preced-
portion of the paper was completed. Two individuals were
used and a series of 18 readings approximately 3 minutes apart
uc ic obtained with each. The total number of contacts made by
ilu-se two individuals in groups of three consecutive trials be-
ginning with the first follows: 20, 9, 10, 6, 7, 4. The results ob-
tained in these observations are consequently in harmony with
ili">e obtained in the preceding observations.

SUMMARY.

\Vfim AiiKL'ba repeatedly comes in contact with a band of
intense light the number of attempts to continue in the original
iliicc tion decreases as the number of trials increases.

This indicates that there is some change in Amoeba that is
analogous to what is called "learning" in the higher animals.

LITERATURE CITED.
Gibbs, David and Dcllinger, O. P.

'08 1 IK- Daily Life of Amoeba prolfus. Am. Jour. Pay., Vol. 19. pp. !
Mast, S. O.

'10 Reactions in Amoeba to light. Jour. Exp. Zool., Vol. 7. pp. 265-
Schaeffer, A. A.

'17 ' 'od in Amoeba. Jour. Animal Behavior. V"' .'^0-258.

Schaeffer, A. A.

'17 Reactions of Amoeba to Light and the Effect of I.i.nht .:UK. BIOL.

BULL.. Vol. 32. pp. 45-72.

AN EXPERIMENTAL STUDY AND A PHYSIOLOGICAL

INTERPRETATION OF EXOGASTRULATION

AND RELATED MODIFICATIONS IN

ECHINODERM EMBRYOS.

*

JOHX \Y. MACARTHUR.
DEPARTMENT OF BIOLOGY, UNIVERSITY OF TORONTO.

INTRODUCTORY.

In a study of the means of altering and controlling the course
of deYelopment in echinids by electrolytes the writer had oc-
casion to note the action of lithium chloride in effecting remarka-
ble form changes in the early ontogeny of sand-dollars and sea-
urchins. An attempt was then made to analyze further these
effects and the similar ones so extensively studied by Herbst
and others, and to provide for the wealth of facts obtained some
more consistent and satisfactory explanation than has been
offered.

Application to these cases of the susceptibility and differential
inhibition methods of Child has supplied what seems a secure
and reasonable basis for interpreting such puzzling teratological
forms as exogastrulse, "holentoblastula?," etc. For I believe that
the first impression of an exquisitely specific action of the Li' ion
is not sustained by a closer examination of the facts, and that
"lithium larvae" fall logically into the series of abnormal forms
producible quite at will by other means.

The work on Echinarachnius, Arbacia, and Asterias was carried
out at Woods Hole in the summer of 1916 and that on Strong-
ylocentrotus and Orthasterias at Friday Harbor in 1917. For
materials and facilities at these stations the author is indebted to
Dr. F. K. I.illie and Dr. YV. C. Frye, and for many suggestions,
discussions, and even for generous contribution of experiment, il
data special thanks are extended to Dr. C. M. Child.

EXPERIMENTAL.

Physiological Axial Gradients. — In scores of species <»t plant *
and animals of all degrees of complexity of organization a gradient

I \<x, \-I RULATION AND RELATED MODIFICATION'S. 6l

of metabolic activity and of other associated properties and
conditions (COo production, O2 consumption, electrical pott-nti.il,
vital staining capacity, etc.) has been demonstrated by the help
of many different method- (Child, '156 '20). Of these nu-thods
ih.M of determining the relative susceptibility of various levels
along i IK- -irintur.il axis is most readily applicable in embryos,
and ha- been found a generally reliable, if somewhat gross, in-
dicator of the presence and nature of the physiological gradients,
the -ignith -.nice of which for development has been emphasized
by thi- writer.

Mt-thods. — As applied to ontogenetic studies there .ire really
interrelated susceptibility methods: (a) direct determination
of -ii-< eptibility of different parts of the egg or embi\o; .m.l
differential inhibition, differential acclimation and diffi-reiui.il
i«, overy in development.

ents in concentrations determined to be lethal within a
few hours commonly cause death and disorganization tir-t of the
part- of the egg or embryo with most rapid metabolic acthiiv
and then later attack in succession in the same way tho-e portion*
\\ith progressively lower rates of metabolism — the susceptibility
paralleling roughly the metabolic gradient (direct susceptibilit\ }.
In lower concentrations which kill after a day or more the on Id-
ol -n-« rpiibility was found to be reversed, since regions with
IIH-IV active metabolism acclimate more rapidly and complc1
(indirect susceptibility). For these purposes a \a-t nnmbd of
lethal a^ent- have been employed with generally concordant
rc-ult-, but the more toxic agents, such as cyanide*, ane-theti.-,
>ah- of hea\\ met.ils, basic dyes, etc., usually give be-t re-ult*;
and \\ith >oine of the-e tests were made on the direct -n-cepli-
bility of the lithium-modified as compared with normal embi

(p. 73)-

When developmental stage- are treated \\ith the -aim-

agents in parti. illy lethal concentration- or -ublethal coiu-i-ntra-
tions pemiitting acclimation or tolerance, or are returned to the
normal medium after temporary e\po-nrc. a dirfeieiiti.il effect
on development results. That i-. the le\d- of hi-he-t metabolic
rate in any gradient are more retarded or inhibited, or in case of
acclimation or reco\ er\ . acclimate or reo»\er more rapidly or
more completely than level-, of lower rate. In differential in-

62 JOHN W. MACARTHUR.

hibition then the levels of lower rate may be relatively accelerated
and in differential acclimation or recovery the levels of higher
rate may be relatively accelerated after the primary inhibition.
Direct differential acceleration is also possible under proper con-
ditions. These differential effects of external agents in relation
to the physiological gradients provide means of controlling to a
greater or less extent the course of development and the sizes,
proportions and relations of parts under given conditions (Child,
'16, '17; Bellamy, '19).

In short, the established gradient of the egg or embryo, being
physiological, is not necessarily permanently fixed and unalter-
able, but may be radically modified, obliterated entirely (as in
Asterias embryos by N/iooo HC1 in sea water), or even entirely-
reversed in direction, as we shall show for the lithium larvae

(Fig. 13).

The ordinary procedure in the differential inhibition method
was as follows: Eggs of freshly collected animals were removed
to a bowl, fertilized, and then placed during early cleavages in
small quantities in liter Erlenmeyer flasks of sea water to which
the agent used had been added. The flasks were loosely covered
but not aerated. Each day the solution was renewed and em-
bryos examined and notes and sketches taken recording the
course of their development as compared with the controls kept
in pure sea water. In most experiments some of the embryos
were also removed after various times of exposure to the salt
solutions to determine the possibility and nature of the steps of
their recovery.

The Lithium Modifications. — Best results with lithium are
contingent upon recognition of the prominent roles of several
factors; especially concentration of the agent, duration of ex-
posure, and stage of development at time of exposure. In higher
concentrations development is strongly inhibited and incomplete.
If the concentration is too low, or if exposure is too brief, begins
too late or is terminated at too early a stage, tin- effect is weak
and development approaches by all transitions toward:- tin-
normal. Herbst observed ('950, ch. 3) that typical eflect> are
not produced by temporary treatment merely up to fertilization,
cleavage or young blastula stages; but well developed blastulae
treated with rather concentrated solutions and then returned to

EXOGASTRULATION AND RELATED MODIFICATION-.

sea water do -how -ome of .the lithium modifications much ob-
-t tired by recovery. In general it maybe said that maximal
effect- may be- -ecured by sea water solutions of about X no to
X joo I.i< • inline to the species of the embryo used) acting

I'm- .1 sensitive determining period from cleavage continuously
through late blastula stages.

Echinarachnius parma. — This sand-dollar provides perhaps
ilu nio-t re-ponsive material yet found for the study of lithium
t-l h-fts, which are here, if anything more diverse and extreme
than in sea-urchins and far more so than in the star hshes used.

The action of lithium solutions producing the characteristic
lithium effects is such as to retard or inhibit development, the
ree of inhibition (and loss of power of recovery) varying
directly with concentration. I.iCI, 2 H-:< ). X 60 or more, in
\\..ier stops cleavage early; and even at X 80 only a few individ-
uals reach the blastula stage. These survivors are either partial
blastula- or very small, immobile, thick-walled stereo-blastul.i .
\\liose remarkably small segmentation cavity is packed so full
of precociously formed opaque mesenchyme that much of the
latter is forcibly extruded from the basal pole; these types,
though not "sickly," die shortly without gastrulation. In

X i leath occurs most commonly in late cleavage, but one

timK inactive at the bottom of the container, a few stereo-
blastuLe. which die as such, or, rarely, after assuming the con-
striction form described below.

In X i JS I.iCl a considerable minority of the egt- In. dine
opaque mesenchyme-filled blastuhe, the larger of which, e-pe-
cially in more <lilule (X 150) Lid, often show a \\ide /one ol
thickening of the basal pole and after some delay bev;in t !u- elonga-
tion preluding gastrulation; but the entoderm then eNa^inate-
after the fashion of that in lithium < 'nila-. There i- much

variation in the diameter of the rin- u~ual!\ ba-al ' where the
ectodermal (gastrula wall) and entodi-i in.il iai-i henteron \ |)or-
tions of the embr\o join, but tin- riir^ i- aKvay- enlarged "tlen
so much so that it mo\-e- toward, to. or e\ en pa-t the equator of
the blastula (Fig. I. ;/-</); in the extreme cases the e\o-astrula-
tion process appears to be niereU a con-trictioii bei\\een the
animal and \egetal |»ole>, forming two vesicles, one

64

JOHN W. MACARTHUR.

of the gastrula wall and the other of the gut entoderm. These
two components may be of very different relative size (Fig. 2)

J

FIG. i. Blastulae and early gastrulae of the Atlantic sand-dollar, Echinarachniiis
parma. a-c, normal embryos; others, typical embryos as modified by lithium. In
the lithium series note: the overdevelopment and frequent basal extrusion of the
mesenchyme; the various degrees of enlargement of the sharply defined, thick-
walled endodermal portion; and the occasional occurrence (s, w, x) of partial cn-
dogastrulation. a and /8 arc double cxogastruhe, two joined terminally by their basal
ends.

as might be expected from the mode of their formation. Some-
times the ectodermal portion exceeds the entodermal, often tln-v
are more or less equal, but in a great many the entoderm is
clearly much the larger. In the last mentioned cases the ccto-
•dermal component becomes less and less, and approaches the
vanishing point — a little knot of cells or nothing at all! — while
the archenteron undergoes a compensatory and reciprocal hy-

KXOGASTRULATIOX AND RELATED MODIFICATIONS. 65

pertrophy, until in a few individuals by its progressive enlarge-
ment the whole gastrula is entodermal (i.e. a "holentoblastula,"
better called for its physiological equivalent a holentogastrula).
There i> never any real difficulty in distinguishing ectoderm
and entoderm in these forms. In every species there are charac-
teristic differences in pigmentation, in thickness of walls, in
structure of the cells, in cilia, in size of cells, in frequency of

a f> c d c f g '

FIG. 2. Series of sand-dollar embryos (constriction forms) developed 27 hours
in M/i6o LiCI. Ectoderm (above) separated at dashed line from cnd"'k-im.
Mcsenchyme packing the blastocoels omitted. The ectodermal (thin-wall, .h
component, larger in a-c, is reduced in size in d-f, and nearly obliterated in >;,
\\liu li is practically a holentogastrula.

nuclei, etc. From a study of the lineage of cells from the pie-
men ted and nonpigmented zones of the early egg it may be stated
quite positively that entoderm grows at the expense of ectoderm,
some of whirh becomes converted into entoderm (and possibly
aUo into mesenchyme).

I ill in the solution many of these constriction forms show
early death and gradual disintegration, sometimes of the apir.il
eetoderm, or of both extremes, but usually of the entodermal
component from its free end towards the an. u lied p. in. 1<« -
turned to sea water the more resistant ectoderm. il portion may
recover and linger on awhile as a partial form, a holectogastrula
as it were. This type may arise al>o in .mother way: if the con-
stricted gastrulae are removed early from the lithium and allowed
to recover in sea water ectoderm and entoderm tend to separate
from each other, especially when the two eoinpoiieiii^ are of
nearly equal size (Fig. 3, v). It would ,-eeni that equality and
duality are inconsistent with development of mie individuality!

N/l6o is close to the optimal concentration for producing the
most striking lithium effects in thi> >perie-. Retardation of

66

JOHN \V. MACARTHUR.

development is less general and more localized and differential.
The eggs produce practically 100 per cent, modified blastuhr,
which are smaller and thicker walled than their controls; the
ectoderm does not thin out for some time even though the blas-
toccel is almost invariably solidly filled with mesenchyme. Such
overproduction of mesenchyme seems to lead to a basal extrusion
and loss of excess proliferated primary and secondary mesenchy-
mal cells from the basal pole of the blastula, which is commonly

s

FIG. 3. Commonest types of sand-dollar gastruki* and larvae from lithium rul-
Uircs. Show the wide range of lithium alterations. (See also other cxumpl'"- in tin-
originals of Fig. 12). a-i, from M/i6o and j-x from M/aoo LiCl, return* .1 at about
2 days to sea water. Mesenchyme omitted in most.

EXOGASTRULATION AND RELATED MODIFICATION-. 07

marked by the "cell ro-ette" of Herbst, composed sometimes of
i-ol.ited cells and -onetimes of loose masses, as if the bla>tnla
wall had ,n inally bnr>t at its weakest point through di-teinion
of it- i ,;\ ity. Shortly the entoderm begins to evert, all stages of
^astrulation an- present (Fig. 3), and soon the culture con-
-i-t- almost entirely of exogastrula? with only a few forms ap-
pro.K liin- normal endogastrulation. And here again one finds
the whole gamut of types with regard to the relative si/e of the
•dermal and entodcrmal portions. The ectoderm shows all
d< ^rees of reduction down to complete obliteration in the large,
relatively clear, holentogastrulu', whose walls and content -
exhibit all the distinguishing features of the archentenm. Noth-
ing is more evident than that the entoderm grows here at the
•ense of the ectoderm.

At 33 36 hours or more those types in which both component -
are present may shoxv further differentiation: the archenteron
^hows constrictions separating it into two, or its usual three
di\isions. This tri-partile gut is safely homologized with tore-,
mid-, and hind-gut, with the fore-gut of course, situated at the
most posterior free end of the embryo. The relative size of these
three divisions shows a most interesting series of variations
I ig. 4). The total gut is greatly increased in bulk over the

ft C <

Fit;. 4. Sck-rU'd series of types from saii<l-<l<>ll.tr i iiltun •- <l, Vi I.,IM-.I m Lit I.

M/I5O to M/2OO. Showing, a-e, the wi> :--iIiii-ti"ii-i in th- m.il

portion u'c/.i; ami tin- parallel ih pi< ipi>rt inn of tin- riiilncl.-i in n it-;
throe divisions: hind-gut (In; inid-giit (m); ami i

normal and could not coin ei\ abl\- be em -lo-ed in the ravity of
the ectoderm; and in thi> inrn-a-e the (We-yin t.ike- the greatest
>hare, the mid -gut next, and the hind-^nt .ittai he<l to the ecto-
derm) the least. There is an ol>\ ion- and -i^nilicant parallel
in thi> proce» to the relatixe increase of the etiloderm as a whole
at the expense of reduction of tin- ectoderm.

• - JOHN W. MACARTHUR.

Herbst demonstrated quite conclusively ('95 a, e.g. Figs. 17,
18, 22) that the thick entodermal layer of the lithium larvae and
of the intermediate types, is rich in nuclei and in the number of
its smaller cells in proportion to the size of the part. The ecto-
derm, on the other hand, gives every evidence of retarded ac-
tivity: at first very thick, it afterward becomes thin and stretched
but is never composed of more than a single cell layer (except
when and where the ciliary bands appear), and this layer is
peculiarly sparse in nuclei and is composed of a few extraordinarily
large cells, both in comparison with the normal ectoderm and
with the entoderm of lithium larvae.

The tip of the fore-gut often shows some disintegration, and
the resulting cellular debris and associated mesenchymal cells
are evidently somewhat sticky, for one finds frequently that
two exogastrulae fuse, telosynaptically as it were, by the ends
of their guts (not laterally as in Herbst's cases). These adhere
closely and more or less permanently to form " fused twins"
(Fig. 3, w, x) whose parts, of varying relative size, have an in-
teresting history of inhibition and reciprocal overgrowth, separa-
tion, etc., that might be followed further with profit.

Meanwhile differentiation has progressed somewhat in the
ectodermal component also; especially if this is fairly large, has
not been too strongly inhibited, and is then allowed to recover
partially in sea water. Notable deviations from the normal
development are seen in the marked reduction and generally
radial disposition of apical structures. The mouth usually
fails to form, but when it does it is practically terminal. The
ciliary bands, when formed, appear as an apical ciliated tuft,
plate, or small transverse ring. Apical outgrowths are common
(Fig. 3, d). By the third day skeletal spicules have often formed
in the mesenchyme, which is collected, not in two lateral masses
near the blastopore, but, rather, crowded far forward into the
apical end of the ccelome; these spicules stimulate the produc-
tion of the pluteus arms, \\hich are uncommonly close together
and narrow angled from the marked reduction of the oral lobe
and also from the tendency of the "anal" arms to be formed
very near the oral region. Herbst also noted ('95, a, Figs. 32-36)
that lithium larva? successfully recovering in sea water continue
their overproduction of mesenchyme and skeletal structures and

EXOGASTRULATION AND RELATED MODIFICATIONS.

69

lay down an excess number of skeletal spicules (3, 4, or 5 — the
total final number in ordinary development — or even 7, in
Sphcerechinus) and the pluteus arms are increased in number and
displaced accordingly and had besides an increased number of
radii '4 or ,S in-trad of 3).

\ joo I. id >lows down development noticeably, but the
bla-iul.i- are of nearly normal size and less crowded with mesen-
ch\mc, wliich is extruded from the posterior end in only about
half of the embryos. When gastrulation occurs there are found
about 60-70 per cent, entogastrulae and only 30-40 per rent.
exogastnilae and intermediate forms. The bla>tula iv« «\ n \\\^
in >ea water quickly resume active movements and I>L-«-..I
trul. i- differing from the lithium larvae in the slightly thinner
•drrni and entoderm; in the incomplete or merely tem|»orary

£• 3 g- j-ftl' t>) evagination of the gut; in the relatively 1

a

e

J

FIG. 5. Sand-dollar embryos as moditic-l l.\ < 50 i1 •. . MK! dilute

•. .ii IT. a. an »-at 1 trula after i day in CuSO«, M 1,250 . alter 2 days

it-nil II . < ,M • .ooo.ooo; c-«. I'ri'in diliiti A.UIT; j-m. from
K\i \l 750,000.

JOHN W. MACARTHUR.

gastrular wall and more fully divided gut; and in their larger
oral region and occasional belated evagination of the stomadeum.

N/25O LiCl yields mostly ordinary forms, but little retarded,
and only 5-10 per cent, exogastrula?. In still loxver concentra-
tions development quickly approaches that of the controls.

Other agents, such as KXC, CuSO,, HgCl2) HC1, CaCl2, XaCl,
and conditions, such as diluted or staling water, used in appro-
priate concentrations and intensities, were found to cause the
same sort of general inhibition of development as LiCl, stopping
entogenetic changes at cleavage, blast ula, or gastrula stages; all
may be employed to show the relatively greater inhibitory effect
on apical and ectodermal parts; many brought the blastula up to
a stage (Fig. 5) precisely simulating late Lithium blastulae; and a
few produced a small percentage of imperfect exogastrulae, as
in Fig. 5.

Arbacia pimctulata and Strongylocentrotus franciscana. — In
these sea-urchins a range of types of abnormality was produced

d

FIG. 6. Exogastruhr, Constriction Forms, and Double Emluyn »t .\rhacia,
produced by LiCl and XaCl. a-g from LiCl, M/8o to M/ioo; It j, I'min N'aCl,
M/4 to M/8.

i XOGASTRl I \lloN AND RELATED MODIFICATION-. JI

by lithium treatment, deviating from the normal in the -aim-
nd ( --entially to the same degrees as do the sand-dollar
lithium larva-. The concentrations required are rather higher,
the mo-t numerous and typical lithium effects being indiuvd in
the Atlantic species by about X 80 to X 100 LiCl, and in the

fie I'orm by X 100 or less.

Several hundred sketches of teratological types in these
un hins (the very common types seen in Figs. 6, 7) conform so

Fit.. 7. Modified > '• M joo; /-/.

from LiCl. M 150; m-o, from NaCl. M i M..MI KN MO; r-z, from

CaCI?. M/ifto to M/2OO.

closely to the sand dollar fonn> just dc-i -rilu-il .Mid to the modifi-
cations illustrated and described -in >m-h -real detail by Herbst
Ti)-1. 'i)5<;. '956) for Echinus nii<rotul>cr(it!<itns, SpharechinuS

72 JOHN W. MACARTHUR.

granularis and Strongylocentrotus lividus, that further descrip-
tion would hardly aid in our present purpose of analyzing and
interpreting the lithium effect.

It is important to mention the production of practically typical
lithium embryos without the aid of lithium. Herbst found a good
substutute for the lithium salts in Na butyrate etc., and Driesch
('95) obtained a high percentage of exogastrulae from develop-
ment at 30° C. for nearly a day preceding gastrulation. So it is
not entirely surprising to find that in CaCl?, N/8o to N/i6o,
often the majority of Strongylocentrotus eggs become character-
istic exogastrulae, etc. (and many of these fused twins) practically
indistinguishable in type from those in Figs. 1-3. Quite similar
exogastrulae are produced in Arbacia by NaCl, N/4 to N/8, and
in Strongylocentrotus by N/8 to N/2O NaCl.

Asterias forbesii and Orthasterias sp. — Starfishes are by no
means as responsive as the urchins and sand-dollar to the lithium
action. Though all characteristic stages of general inhibition are

FIG. 8. Experimental Alterations of Aslcrias, a-e induced by LiCl, M/ioo to
M/i6o; m-p, induced by 75 per cent. sca-\vatcr; and q, by KNC, M/25,ooo

EXOGASTRULATION AND RELATED MODIFICATION-.

73

readily obtained by the lithium salts and other agents, and apical
inhibition is evident in the characteristic reduction or entire
suppression of the preoral ciliary field, etc., yet the endogastrula-
tion process is not so easily interfered with.

But partial and occasionally fairly typical exogastruhe occur
with great regularity in Asterias cultures in N/ioo to N/2OO LiCl
solutions in KNC, in diluted sea water, etc. (Fig. 8).

In Orthasterias N 160 LiCl sometimes induces exogastrulation
or marked change of proportions of the germ layers, and not un-
commonly in later development the stomadeal pouch is turned
outward and forward from the mouth by a process apparently
similar to exogastrulation (Fig. 9.).

/> C

FIG. 9. Alterations of Orthastrrias embryos in M/i6o LiCl. d-t, showing sterna-
l-version; t.witli mixed endo-and exo-gastrulation; «j-6.with reduced ectodcim
and greatly enlarged endodi-rm. particularly in the fore-gut. Note the position ot
the c-iliary bands in c-c and in Fit;. 8. c.

Comparison of Susceptibility gradients of Normal and Litliittm

Embryos.

i. The Xormal Embryos. — The susceptibility relations in the
i ,nl\ ontogeny of Asterias and Arbacia have been described by
Child ('i5&; '160); these observations have- been confirmed and
extended to include the sand-dollar (data by I >r. Child and tin-
writer) and the Pacific coast forms.

According to the direct susceptibility method p. <\\] tin- fertil-
ized or dividing eggs, blastulcT?, gastrula- and larvae were placed
in loth. il solutions of KNC (M iooi,,M rooo), HgCl M -«>ooo
or less), CuSOi, concentrated lu-utral red or methylene blue, and
to a few other salts, etc. and the course and relative times of
death of parts noted by the criteria of pi»-ibility of recovery in
return to sea water and of disoruani/ation of cell structure.
Disorganization is often made more \i-ible either by a brief
urn to sea water, or by previous or simultaneous staininy in

74

JOHN W. MACARTHUR.

dilute neutral red since the stained tissue decolorizes sharply at
the death point. Since general susceptibility increases steadily
from fertilization to gastrulation the concentration of the agents
was correspondingly decreased in the later stages.

The order of susceptibility in the succeeding stages of develop-
ment is shown in Fig. 10. Disorganization, manifested by droplet

a

o

f

FIG. 10. Disintegration of normal sand-dollar eggs and embryos, showing re-
lative susceptibility of parts. Mesenchymc omitted, a-b, eggs; c, 4-celled; d-f,
blastulie; g-j, gastrulte and early larvae. When the polar axis becomes recognizable
disintegration proceeds from animal to vegetal pole (i.e. is apico-basal) .

formation, cell separation, roughening of contour, visible dis-
integration, and decolorization, according to the particular agent
used, begins at one point of the surface of the egg, or at one pole
of the blastomeres in tin- 2- or 4-cell stage, or at one pole of the
cleavage mass or early blast ula, and then extends gradually over
adjacent parts to reach last the opposite surface or pole.

As soon as the blast ul.i differentiates to show the b.is.il thick-
ening and begins its elongation, one may orient the embryo in

EXOGASTRULATION AND RELATED MODIFICATION-. 75

its chief apico-bas.il axis at a glance. Here the manifestations
of olc.it li are tir-t visible at the animal pole or apex and pn>^:

dually and -ic.idily towards the basal region which is reached
l,i~i, often hour- .tfter the opposite end had been affected. The
apparent n-ver-e in some slow-acting solutions was obviously
due eiiiin-ly to the basal extrusion of the mesenchyme.

Tin- same gradient persists into the gastrula stage, where the
apex -hows the greatest susceptibility, while the inva^inated
• •MI., derm is so much more resistant (except probably for a /one
around the blastopore) that it usually survives in M [OO KNC an
hour or more after the ectoderm of the entire gastrular \\all has
disintegrated. By returning non-disintegrating gastrula1 to
sea water after an appropriate exposure the ensuing dish ion

may be checked midway in its course and recovery I.ir\.r
ol it, lined showing absence or partial loss of apical (or even nioir
basal) ectoderm and more or less complete survival of the ento-
dermal parts which are accordingly large in proportion to the
whole. As in the case of Arbacia "in the gastrula the apex of the
conical both' represents the apical end of the major axis, and the
Mastopore the basal end, while the entoderm represents a still
more basal region of the egg and blastula, and the mex in livme
the most basal region" (Child, 'ib/>, pp. 68-69).

Vital staining with neutral red and other basic dyes shows an
apico-ba-al progression in blastul.e and early gastruhe quite
paralleling in all essentials the wave of advancing disintegration
as de-cribed above.

The experimental embryologists have described the i-oinni"n

a d c

FlG. ii. Diagram illu-tratiiiK tin p..|.ir ;nr.m.o nn lit ..I part- in tl
the normal i-inlu iiicl in litliiinu i-mlii-. •.iiinili-iin

mill-, and hind-gut. The lithium rml>ryr> i~ ii-.i!l\ an <-!i iny.iti <\ .ni'l

with the germ-'
tlu- .

JOHN W. MACARTHUR.

plan of organization of the Echinid egg as such that the prospec-
tive significance of different levels of the egg along its main struc-
tural axis is as follows (see Przibram, '08): The animal pole
half (diagram, Fig. n. Ect.) furnishes the ectoderm and its
differentiations; the subequatorial zone (End.) the gut and its
derivatives; and the vegetal pole (Mes.) the primary mesenchyme
and the larval skeleton. In Fig. n (6) and (c) these parts are
homologized in the normal and the lithium-modified gastrula?

a

y

FIG. 12. Disintegration of typical lithium-modified sand-dollar embryos and
larv;f. Mesenchyme omitted, a-c in neutral ivd; </ /, in HgCli M/sooooo; j-q,
in HgCl: plus Neutral Red; r-y, in KNC, M/IOOO; /, q, y, fused doubles. In the
less modifier! forms (a, b, d, j, r) susceptibility is about equal apically and basally.
In the more modified forms with large development of endoderm a distinct basi-
apical disintegration gradient is always found. In general the larger the relative
size of the endoderm and of the fore-gut the more distinct is this reversed gradient.

EXOGASTRULATION AND RELATED MODIFICATION- 77

respectively. Obviously in the lithium-modified embryo tl it-
egg ha- merely elongated and differentiated without tin- usual
infoldii

2. Lithium Embryos. — Tested with the same agents in prc-
cisely tin -aim- way as were the normal embryos the exogastrular
.UK! con>triction forms (p. 63) showed new and different and
charactc ri-tir susceptibility relations. General susceptibility is
much less than in the normal, and the original apico-basal
gradients of susceptibility, vital staining and decolori/atiim are
less marked, absent or reversed (Fig. 12).

Quite commonly the basal end of the entoderm ha> already
in disintegration in the LiCl solution itself, and much of tin-
second a r\' mesenchyme may have been removed and tlu- primary
• •nchyme may have escaped from the blastocoel (Fig 12, /, /).

Newly formed lithium exogastruhe and constricted forms
removed intact to M,. 1000 KNV, M 500000 HgCl-_>, or concen-
trated neutral red. disintegrate in the next few hours. In this
process disintegration and decolori/ation after staining generally
I « Jus in the secondary mesenchyme-forming region of the an h-
enteron.

Sometimes the two extreme ends disorganize at the same time,
the apical ectoderm of the gastrula wall being as suscrpiibile as
tlu- basal entoderm of the archenteron. And in a few cases the
ectoderm goes a little in advance of the opposite end. In these
hour glass forms the susceptibility varies with the size of tin-
components; the larger component being quite uniformly tin-
more susceptible; and indeed it is apparently susceptible in din •< t
1 'i' 'I ioi lion to its relative size, as the sketches str^e-t. I
greater the reduction of the ectoderm the higher i- it- resistai

Exogastruiac removed from the lithium in a more differentiated
state with bi- or tri-partite gut, etc., exhibit regularly the re-
versed order of susceptibility of parts I

DISCUSSION AND IMI-RI-KI i \ i h >\.

i. Physiological Gradients in I-'.ihiniil /-'.»: hryos.

It will be taken as highly probable that in the e.^g and blastula
of the echinoderms so far studied the >u-ceptibility declines from
the apical animal pole region to the ba-al \e-etal pole levels. In

78 JOHN \V. MAC-ARTHUR.

the late blastula, gastrula and larva the susceptibility demonstra-
bly declines sharply from ectodermal to entodermal and mesen-
chymal regions (p. 74). This is chiefly significant as indicating a
corresponding gradient of metabolic activity — as a prerequisite
to normal development — which is further evidenced and mani-
fested by structural differences along the apico-basal axis
(differences in cell size, cilia, pigment, vacuolization of proto-
plasm, etc.) and by functional differences in rate of develop-
mental processes (growth, cell division, morphogenesis, and
differentiation).

It is a familiar fact that teratogenic agents and conditions
applied to whole eggs and embryos affect different regions of the
developing organism differently and to a different degree. On
the quantitative metabolic gradient conception regions of great-
est susceptibility are subject to the greatest retardation in the
developmental processes, while those of lowest susceptibility are
correspondingly least inhibited, as is experimentally proven for
Arbacia (Child, 'i6a), Annelids (Child, '17), the frog (Bellamy,
'19), etc. Thus most teratological forms may be reproduced
according to expectation and explained consistently.

Lithium effects also seem to be produced by a differential in-
hibition along a quantitative metabolic gradient. The facts
relevant to this interpretation have been collected in the next
section.

2. Analysis of the Lithium Effects.

Exogastrulation aside, lithium larva? are characterized mor-
phologically by the abnormal proportions of their parts, which
involve all the germ layers in a manner we may summarize as
follows:

(a} Increase in amount and change in position of the mesenchymc
and its derivatives:

Excess mesenchyme of stereoblastulae (p. 63).

Excess pigment cells (p. 63).

Excess skeletal structures, e.g., spicules, arms, radii of arms, etc.

(Herbst).
Conspicuous production of serous fluids and swelling of serous

cavities in main' annelid, fish, and amphibian embryos.

KX<>i.\-I Kll ATIOX AND RELATED MODIFICATION-. 79

(b) Ilvperlrophy of entodermal parts, especially basally:
Relative and ab-olute increase of size of the gut in general, and of

the fore-gut in particular (Figs' 2, 3, 4 etc.).
Thickne— of eiit< -dermal wall (p. 68).
Kiclim--- of entoderm in nuclei (p. 68).
Small size of » ntodermal cells (Herbst; for frog, Bellamy '19).

(c) Rein/in- and Absolute Diminution of the Ectoderm, especially
a pic ally:

Reduction of the ectodermal component of the hour glass forms

(Figs. 2, 3).

K- 'anlation of the apical region in differentiation (p. 69).
!•'• 'ention of radial symmetry (p. 69).

!• sen ess in nuclei (Herbst).
Large si/.e of ectodermal cells (Herbst, Bellamy. Le Plat, '20,

et al.).

I ilure of ectoderm to grow down over the entoderm in the I
(Gurwitsch, '94, Morgan, '(.>3/>, Bellamy. 'i<). Le Plat, '20).

Since the main object of this paper is to deal with the lithium
eiiect physiologically, it will profit to restate the morphological
i just summarized in physiological terms:

Sufficiently strong concentrations of the Li ion will suppress
dc\ t lopment at any early or late stage of cleavage, of' the blast ula,
etc. Slightly lower concentrations also retard development, but
this retarding influence is most marked in the most su-ceptible
• 'derm) of the embryo, distinctly less marked in tin- le--
-n-tepiible entoderm. and least marked in the lea-t -u-( -f|H ililc
mrsi'in -li\ inc. The last tissur undergoes a rel.ilix i-l\- rapid and
rxivssive growth and abundant proliferation (in niaiu \\
resembling malignant tumor growth) to till the blastoccele and
lay down the anlage of an increased number <>| ~kelet.il -pit nle-
(which at once push out the arm- of the plnten- it recovery i~
permitted). The change of position <,f the me-en< h\me ma\ in-
dicate a tendency (active or pa--i\« to or.-upy ell-
of lowest metabolic activity.

The entoderm growth zone extend- it-elf apirally on t«> the
gastrula wall (more and more as the ^r.idiein i- le\dK-<| do\\n or
reversed) by virtue of an enhanced mitoijc ,md m<p\\th activity,
\\hich is one of the most distinctive character- of a lithium embryo

80 JOHN W. MACARTHUR.

and the feature by which the lithium effect may best be quantita-
tively estimated. The greater the lithium effect short of general
inhibition, the greater thelrelative size of the entodermal as
compared with ectodermal parts.

Meanwhile mitosis and growth have slowed down or utterly
ceased in the ectodermal zone, whose development varies in-
versely with the development of the entoderm. In extreme cases
the entodermal growth zone overspreads the whole blastula wall
resorbing and incorporating and changing entirely the prospec-
tive significance of the apical region. This resorption doubtless
proceeds in proportion to the relative rate of metabolic activity of
the entoderm; just as in normal ontogeny the ectoderm feeds
upon and limits the development of the gut, or as the nervous
system of a starving animal maintains itself at the expense of
parts of lower physiological activity. Here the ectoderm, usu-
ally the master germ layer, loses control and the more active
entoderm gains the ascendency and may become the major part
of the embryo. That the fore-gut of the evaginated archenteron

also increases disproportionately may be attributed to the same
process.

The inhibitory effect of lithium appears in the order: ecto-
derm > entoderm > mesenchyme; in the ectoderm itself: api-
cal > basal ;*and in the entoderm: hind-gut > mid-gut > fore-
gut (Fig. 13).

The most substantial direct proof of this reversal of the normal
physiological gradient is found in the more or less complete re-
versal of the usual order of susceptibility; for in extreme cases
the direct susceptibility was seen to be greatest in the fore-gut
region and to decline towards the hind-gut and ectoderm (Fig. 13).

Judged then by the criteria of susceptibility, of developmental
activity in mitosis, growth, rate of differentiation, etc., the meta-
bolic gradient of the lithium larva is diminished, obliterated, or
in the extreme case actually reversed (basi-apical). It is evident,
however, that the reversal of the gradient occurs too slowly or
too late to alter significantly the nature of the differentiated end
products, for the respective germ layers bear recognizable simi-
larity to these of the normal larva. Earlier exposure of the egg
and a different technique of recognizing polarity would be neces-

EXOGASTRULATION AM) RELATED MODIFICATIONS.

81

sary to determine whether lithium can be made so far effective
as to cause ectoderm to appear at the basal pole, etc.

Normal E
apical

a -i

Lithiurn L VTI b *~ i_i o
a p \ c a I

Ect odLerm

hi* "f
i-3at

Ch d o d-e-v

rne

b'

h -L

I/

basal k

Kir.. 13. Diaxrarn showinx tlie changed proportions of parts in a lithium 1-1:

as compared with a normal embryo (left), and the typical reversal oi '
lii lity (the arrows), a-b in normal, b'-a' in lithium embryos. Tl
ill pressing concentration of lithium is to shift the embryo to the right. An ..
riatiiix ax«-rit would alter embryonic proportions in the opposite din . timi.

3. Non-specificity of Lithium Effects.

(a) Changes of Proportions of Parts. — A great vark-t y ( if ( la-mi-
ca 1 agents and physical conditions have been employed in ex-
perimental embryology to cause general death and disintegration,
general inhibition of development, differential (/.«•.. lorali/«<l
inhibition, and, in the minimal effective concentration-, differen-
tial acclimation effects. The striking fund.niient.il -imilarity of
teratological forms produced by di\> • genl ial cmiditions is
too remarkable to be insignificant. Then- can In- no ma tera-
togenic substance or condition \vho>r rt'U-ct dilltr- -h.irpl\- in
kind from others. Though each may have it- <>\vn mode of
attack, the end result is in general injuriou- and inhibitory or
accelerative, according to the p»\\ti of resistance of different
levels along the metabolic ijMdient. The only -peciticity of
lithium effect seems to lie in the higher re^am.il differential
-ii-ivptibility of many organi-m- io it ; lithium ha- a particularly

JOHN \V. MACARTHUR.

marked capacity of inhibiting locally, retarding apical regions
without appreciably hindering basal regions.

(b) Exogastrnlation. In practice quite typical exogastruhe
are obtained in the sand dollar or in sea-urchins by all the lithium
halides, sodium butyrate, KC1, etc. (Herbst), by CaCl2, dilution
or staling of sea water, etc. (this paper), by elevation of tempera-
ture to 30° C. during gastrulation (Driesch); and initial stages of
exogastrulation by main' other agents, as NaCl, CuSO4, neutral
red, etc. As regards exogastrulation specificity appears to reside
chiefly in the biological material rather than in the external agent,
for perhaps only in the pluteus-forming echinoderms is the lithium
effect "typical"; in the starfishes, Crepidula (Conklin), and the
sponge, Oscarella (Maas, '98) the effects are sufficiently marked
and similar as to be recognizable; and in the annelids, ascaris,
amphioxus, frogs (?) only negative results appear. Possibility of
inducing exogastrulation depends, doubtless, on the specific
constitution and organization of the egg, especially on the polar
arrangement of the levels destined to form ectoderm, entoderm
and mesenchyme. The physiological equivalents of exogastrulae
may well appear quite different morphologically in different
organisms.

4. The Mode of Action of the Lithium.

The problems of lithium action resolve into the causes of the
two essential modifications: first, the disproportionate develop-
ment of entoderm and mesenchyme; and second, exogastrula-
tion. These are without doubt closely interrelated processes
proceeding from a common fundamental cause.

Herbst admitted that he found himself in complete darkness
on these matters but gave reasons for his opinion that the site
of action of the salt must be in the egg itself rather than in its
external membranes, and that Li' must produce a specific stimu-
lating effect on vegetal pole and entodermal cells, whose capacity
for taking up and retaining this ion he believed to be greatr-t
('95*7, p. 185). Seeking to put the matter on a physico-chemical
basis one of his students (Spek. '18) gave some evidence that it
Li' did penetrate into certain (e.g., entodermal) cells ii svould
make for a greater imbibition of water by their colloids. Driesch
('95) believed that "the growth process of tin- blast ula wall is

EXOGASTRULATION AND RELATED MODIFIt \llo\-.

the vital fundamental process; the direction towards \vhirh the
\vth pro< eed- in exogastrulation is 'determined by the sur-
roundi; It i- -u restive that both these leading experimen-

tali-t- appealed to -ome more or less differential growth pro.
Inn in neitlu s the evidence more than meager, indirect,

and i-olated.

An..' k'-d t'roin the new viewpoint of differential inhibition,
en .. lithium ariion is seen to be interpretable in a \vay not only
entirely eon-i-ietit and coordinate with a vast body of modern
\\ork in general and special physiology and of physiological form
deiermination and teratomorphy, but also substantiated by
< on~iderable direct proof and many suggestive parallel- (.p. 78.
and Child, '20). Since Li' in the concentrations actually m •
-ary is certainly inhibitory and since the inhibitory ellei •'- arc
felt rhietly at t lit- animal pole and in its differential i< m- and Ica-t
in the mesenchyme and the entoderm, it is unlikely that t la-
Li' act- >o much directly on the basal regions; it may be said
i.it her to release these regions from the control of tin- ectoderm
b\ checking the (usually dominating) activity of the hitter.

I determine the more proximate causes of the inhibitory
effects will require much further detailed observation and e\-
periment (for each separate agent) on sites and rates o! penetra-
tion, water imbibition and withdrawal, alteration ot -in:
ten-ion and chcmotaxis (Rhumbler, '02), precipitation-. Dela-
tion-, -olmions, etc. Into whatever ultimate cau-al compoiieni-
resolved the end rc-ull remains, physiolou' 11\ . .1 dittereni ial
inhibition.

I -irulatioii is evidentK' a secondarx manifestation o| thi-

di Herein ial action alonj; a le\'elled or iv\ er-ed metabolic gradient .
Since the lithium-modified embrvo i~ merely a len;<l heiied and
differentiated e^ in \\hich endogastrulation ha- been |»re\i-nied
Fij ii ' ••-.' v .-irulation mii^ht be con-ii leici I t \ en a direi t
re-nlt of the inhibition of the-i le\elliir^ the normal -radit-m

nece>sar\ \< >v en< h \gas\ rulai ii ui .

But a mechanic.il factor -eein- lo be involved. K\paii-ion and
o\er^rowth of the ectodcrmal i'(Mlhelial la\'er i- checked b\ the
tailure of the u-nal multiplication and enlargement of ii-, cell-,
whik- growth and cell di\i-ion of eiitodcnn are le--. and tho-e of
mc-ciich\me least retarded. Consequently the ecto.lcrm.il cell-

84 JOHN W. MACARTHUR.

are few and large while the entodermal layer is rich in smaller
nuclei. The mesenchyme cells proliferate in enormous excess
and fill and overdistend the small blastocoel. Thus considera-
tions of space and inner pressure alone render almost mechan-
ically impossible the infolding of the entoderm which is, further,
much longer, larger and thicker than usual, and is contiguous
with the ectoderm by a vastly enlarged blastopore ring.

Certain other relevant facts should be noted: that the lumen
of the exogastrular gut is really a part of the blastocoel or coelome
and is not the true gut cavity; that the constrictions of the dif-
ferentiating gut are also therefore reversed in direction; that the
stomadeum also often everts in starfish; that exo- and endo-
gastrulation intergrade (there may be first one and then the
other, or both at once). It is possible that the agent in obliterat-
ing and reversing the primary physiological gradient of the or-
ganism as a whole at the same time obliterates and reverses the
functional polarity of the separate cells of the epithelial layers, so
that insofar as polarity of cells aids in normal endogastrulation it
might act in the opposite way in the lithium larvae (by failure
of the cells to widen and vacuolate basally, change their nuclear
position, etc. cf. Gudernatsch, '13).

Tests of this interpretation of the lit hi um action may be made
by an appropriate modification of the technique on other gradi-
ents (embryological, regenerating, motor). This agent might
tend to reverse the polarity of regenerating pieces (e.g., hydroid
internodes, as an electric current does, Lund, '21), or of motor
gradients (cilia, or muscular peristalsis), etc.

SUMMARY AND CONCLUSIONS.

1. Typical lithium larvae may be produced from a high per-
centage of eggs of the sand-dollar, Echinarachniiis parma, and
of two sea-urchins, Arbacia punctulata, and Strongylocentrotus
francisciana, by addition of LiCl to the sea water. These types
approximate closely those obtained by Herbst in other urchins.

2. Morphological features common to and characteristic of
these larvae are (p. 78): (a) exogastrulation in many cases, and
(b) always increase of mesenchyme and of entoderm and a com-
pensating decrease of ectoderm. Kurt her tin- most apical p.iris
of the ectoderm and its differentiations are most decreased, and

EXOGASTRfLATION AXD RELATED MODIFICATION-. 85

the most basal parts of the gut are most increased, as the lithium
alteration becomes maximal.

3. In the lithium larvae the order of direct susceptibility of
parts to quickly killing agents is essentially the reverse of that
of the normal v;a-trula and larva: the gradient of susceptibility
in the normal e.ui; and blastula is apico-basal, and declines from
tin- e< loderm tn <-nti.<lerm and mesenchyme in the -a-trula and
larva, while in the fully formed exogastrula the susceptibility of
tin- »•( tixli-nn 'LM-trula wall) is least, that of the entodermal
parts more, and that of the mesenchyme probably greate-t of all.
A -iniilar iv\ei-al of the usual susceptibility is met with in pan-
of the differentiated gut of the lithium larva.

4. " Lithium larvae" are not the specific product of the I.i' ion
alone, but are readily obtained in CaCl«, diluted or Male
\\.iter, by Na butyrate (Herbst), by heat (Drieschj, and in less
typieal t'orm by many other salts, basic dyes, etc.

5. The specificity resides rather in the biological material
than in the agent, but Li' appears to possess in high d the
pi.\\er of locali/ed action : i.e., its inhibition is markedly rej i> mal.
>lo\\in- down or stopping development of the most apical pan-,
while hindering little or inappreciably development of the lu-al
parts.

induced, highly differential, rates of growth and
diiii -rentiation lead directly or indirectly to exoga>ti ulatimi in
e\tt( in. . ases; perhaps largely because of the nuvhaniral im;
Ml'ilitx "! invaginating a greatly enlarged entodenn intu an
umiMially Mnall Mastocoel, itself overcrowded \>\ a pn ous
ami exi ess pu'lileration of mesenchyme.

7. I'lieM- ( -liaii^es in proportions and ) n >-it i< ui> of part- in
" lithium rniln \ os" appear to follow as naturally and ne. .---arily
from the altered metabolic relation- ni"iv or 1( inph -te

reversal of the physiological gradients - Hie normal prop,, ni..n-
and positions of part> follo\\ in. in the normal metal xilir relations.

BIBLIOGRAPHY.

Bataillon.

'01 La Pressure Osiiiotit|ii. i-t !. -s grandes Problcint^ dc la B .. f.

Entw. Mrdi; XL. i
'04 Lc- ' rr. ex-

prriiiu-tualfs. Arch. f. Km\v. Mivh.. X\'III.

86 JOHN W. MACARTHUR.

Bellamy, A. W.

'ig Differential Susceptibility as a Basis for Modification and Control of Early

Development in the Frog. BIOL. BI/LL. XXXVII., 312-361.
Child, C. M.

'158 Senescence and Rejuvenescence, Chicago.

'i5b Axial Gradients in the Early Development of the Starfish. Am. Jl.

Physiol., XXXVII., 203-219.
'i6a Axial Susceptibility Gradients in the Early Development of the Sea

Urchin. BIOL. BULL. XXX., 391-405.
*i6b Experimental Control and Modification of Larval Development in the

Sea Urchin in Relation to the Axial Gradients. Jl. Morph., XXVIII.,

65-133.
'17 Differential Susceptibility and Differential Inhibition in the Development

of Polychete Annelids. Jl. Morph., XXX., 1-63.
'20 Some Considerations Concerning the Nature and Origin of Physiological

Gradients. BIOL. BULL., XXIX.. 147-187.
Driesch, H.
'95 Entwickelungsmechanische Studien, VII., Exogastrulae und Anenteria.

Mitth. d. Stat. Neap., XI., 221-226.
Gudernatsch, J. F.

'13 Concerning the Mechanism and Direction of Embryonic Foldings. Anat.

Rec., VII., 417-431.
Gurwitsch, A.
'96 Uber die formative Wirkung des veriinderten chemischen Mediums auf die

Embryonale Entwickelungs. Arch. f. Entw. Mech., III., 219-260.
Herbst, C. Experimented Untersuchungen iiber den Einfluss der veriinderten
chemischen Zusammensetzung des umgebenden mediums auf die En-
twickelung der Thiere.

'92 I. Versuche an Seeigeleiern. Zeit. f. Wiss. Zool., 55: 446-518.

'953 II. Weiteres uber die Morphologische Wirkung der Lithium-salze und
ihre theoretische Bedeutung. Mitth. d. z. Stat. Neap., XI.,
136-221.
*95b III. Uber das Ineinandergreifcn von normaler gastrulation und Lithium-

entwickelung, etc.
IV. Die Formative Wirkung des Lithiums auf befruchtete Eier von

Asterias glacialis.

V. Uber die Unterdriickung von Entwickelungsprocessen. etc.
VI. Uber den Einfluss einiger anderer organischen Salze. Schlussbemer-

kungcn, Arch. f. Entw. Mech., II., 455-516.

(97-'04 Uber die zur Entwickelung der Seeigel Larvcn notwendigcn anorgani-
schen Stoffe, ihre Rolle und ihre Vertretbarkeit. Arch. f. Entw. Mech.,
V., VII., IX.. XL. XVII.,
Leplat, Geo.

'20 Action du Milieu sur le dcvcloppement des Larves d' Amphibians. Arch, de

Biol., 30: 231-322.
Lund, E. J.

'21 Experimental Control of Organic Polarity by the Electric Current. Jl.

Exp. Zool., 34: 471-487.
MacBride, E. W.

'18 Artificial Production of Echinoderm Larva; with Two Water Vascular
Systems, etc. Proc. Roy. Soc. Lond., 90 B, 323-348.

I Mil, \-I KLLATIOX AND RELATED MODIFICATIONS. 87

Maas, O.
'98 Di<- K<-imM. Liter der Spongien und die Metamorphose von Oscarella. /

f. \Yi-- /.,.jl.. LXVIII.. 665-679.
Morgan, T. H.
'038 The (iastrulation of the Partial Embryos of Sphaerechinus. Arch. f.

•v. . M.-i-h.. XVI.. 117-124.

'osb I In i bet\%-een Normal and Abnormal Development in the Embryo

of i as determined by the Effect of Lithium Chloride in Solution.

An h. i. i-.ntw. Mech., XVI.. 691-712.
Przibram, A.

'08 K.\|M-rimi-ntal Zoology. Pt. I., Embryogeny, Chap. 2. pp. 14-15.
Rhumbler

"02 Zur Mechanik des Gastrulationsvorganges insbesondere der Inva;.;iiuuion.

Arch. f. Entw. Mcch.. XI\'., 401-476.
Riinnstrom, J.

'18 Analytischc Studien uber die Seeigel Entxvickelung. Arch. f. Entw. Mech.,

XLIII., 409-447.
Spek, J.

'18 DilTerenzen im Qucllcnzustancl der Plasma Kolloide als eine I
Gastrulainvaginations, etc. Kolloidchem. Beih. IX.

THE EXISTENCE OF A SHORT REPRODUCTIVE
CYCLE IN ANODONTA IMBECILLIS.

EDGAR ALLEN
I". S. BUREAU OF FISHERIES LABORATORY, FAIRPORT, IOWA.

Fresh water mussels of North America have been classed by
Sterki (1895) as summer or winter breeders. He notes that in
winter breeders fertilization of ova occurs in the late summer,
embryonic development is completed and the fully developed
glochidia are carried through the winter in the maternal marsupia
and discharged in the following spring. In the summer breeders
fertilization occurs in the late spring or early summer and the
fully formed glochidia are discharged "as a rule by the end of
August."

Lefevre and Curtis (1912) prefer to designate these two classes
as long and short term breeders. They include Anodonta in the
former group, and although A. imbecillis is not listed in the data
from which their conclusions are drawn, the statement is made
that the breeding season is a generic character, "all species of a
genus having essentially the same period of gravidity."

In a table compiled by Coker, Shira, Clark, and Howard (1919)
gravid Anodonta imbecillis are listed for all months with the ex-
ception of April.

Anodonta imbecillis, although not important commercially, is
of interest for two reasons; first, it is one of the two known species
which carry the young in the marsupia during metamorphosis
from glochidia to juvenile mussels, thus eliminating the parasitic
stage on fishes which other species undergo, and, second, it is
one of the few species which are hermaphroditic (Sterki, 1898).

During the summer of 1922 the writer attempted to work out
a method of aparasitic mussel culture of commercially valuable
mussels. One line of attacking this problem which seemed to
offer promise of results was to simulate conditions pertaining in
metamorphosis during the intra-marsupial life of aparasitic
forms. Therefore a series of observations was begun on Anodonta

imbecillis during the course of which successive examinations

ss

REPRODUCTIVE CYCLE IX ANODONTA IMBECILLIS. 89

\v»-re made at intervals of several days on more than one hundred
individuals. This has furnished the following evidence for a
shorter reproductive cycle than formerly reported for thi- -pecies.

MATERIAL AND METHOD.

Anodnnla imbecillis is commonly called the "paper shell "
because of the extreme thinness of the valves. Pos>ibly corre-
lated with this economy in shell formation, is its rapid rate of
growth. It may begin to reproduce during the second year. The
observations in this paper were made on two- and three-year-old
mussels.

In this species the outer gills serve as marsupia. It i> an easy
matter to insert a capillary pipette between the valve-, mak
HMjJe puncture in the marsupium, and remove some of it- con-
tents with only slight injury to the mussel. This puncture In
rapidly. In order that the least possible modification of reproduc-
tive activity might be produced by this examination, tin- >am
removed were small, containing only from 20 to 40 individual-;.
The puncture was made in a direction parallel to the loni; a\i- <>t
the gill so that the sample removed contained young from several
v;ill compartments.

Relatively slight disturbances may cause "abortion "'of mar-
supial contents in some species. Lefevre and Curtis, however,
note that they have never observed abortion in Anodonta. The
thin shell of .1. imbecillis renders the above described operation
very simple as the valves may be opened for the insertion of the
pipette with the thumb nails. This probably eliminate- the
possibility of abortion as a factor bearing upon the folloxvin-
evidence.

The examinations extended over a period of one and one half
months from the middle of July to early in September.

Most of the mussels studied were taken from the main supply
reservoir of the V. S. Fisherie- Hiolo-jc.il Station at Fairport,
Iowa, which is supplied daily from the Mi — i--ippi River. A few
preliminary observations were made on mus^el> taken from the
river.

Throughout the observations they were kept in running water
in experimental troughs supplied from a pond containing an
abundance of pond life. As opportunity for considerable sedi-

EDGAR ALLEX.

mentation was afforded, this supply was much clearer than water
from the river or reservoir. The fact that very successful yields
were obtained in artificial mussel culture experiments carried on
simultaneously under these conditions by Dr. A. D. Howard and
B. J. Anson excludes any possible intervention of adverse en-
vironmental conditions.

OBSERVATIONS.

Between July 15 and 20 about two dozen mussels were ex-
amined without finding a gravid individual. On July 24, 5 of
33 examined (15 per cent.) contained early embryonic stages.
Nineteen of these 33 were observed repeatedly until September
i. On July 31,8, including the five already noted (42 per cent.),
contained embryos. On August 4, 17 or 89.5 per cent, were
gravid. Larger numbers of this species were examined during
August. From August 8 to September I the percentage of mus-
sels bearing embryos, glochidia, or young in various stages of
metamorphosis never dropped below 77.2 per cent, (see Table
I. and Graph I).

TABLE I.
PERCENTAGE OF MUSSELS BEARING EMBRYOS OR GLOCHIDIA.

Date

7/15-

7/24

8/4

- ij

8/15

8/22

8/28 9/1

20

Number examined. . . .
Percentage

20-24
o

33
I ; I

19

8Qo

69
80.9

IOI

83.2

105
77.2

103 57
80.6 82.5

80

20

20 2V
July

31 4

15

22

28 I

5ef>t

GRAPH i. Percentage of mussels bearing embryos or glochidia.

REPK<>M « I I VI-: CYCLE IX AXODOXTA IMBECILLIS. QI

The com lu-ion drawn is that the reproductive season of
A notion/a inibcrillis in this location begins during the latter part
of July.

As the-e observations were continued the various stages of
embryonic development and glochidial metamorphosis were
ideniilied and data accumulated as to the time required for this
transition. Eleven different stages were identified as follow- : !

1. Karly segmentation of the fertilized ovum.

2. Later segmentation.

3. Karly differentiation of the single anlage of the valves which
at this stage caps the segmenting cell mass.

4. Formation of a groove in this "cap" which di\i<le- ii into
two separate valve anlagen.

5. The developing valves lie on the segmenting cell mass "like
continents in relief on the globe."

6. The two valves resemble a folded leaf which now completely
encloses the future "soft parts" of the young larva.

7. The glochidial shape is attained but the larva is still color-
less and transparent. The single glochidial adductor muscle
can be seen to quiver or "Hash" although the valves are closed.

N. The glochidia are now fully formed, have a rich \ello\v
color, and a single adductor muscle. Main- may be seen to
snap vigorously if freed from the egg membranes in the process
of removal from the marsupia. Those within the nn-inbranes
ha\e lln-ir valves opened. This stage is equivalent to that at
\\liich tin- glochidia of species dependent on para-it i-in on !i-h< •-
for metamorphosis are extruded from the mar-upia.

9. Mid-metamorphosis; rich yellow in color, no adductor
muscle visible.

10. Late metamorphosis; the two delinir.xe adductor nni-< les
are now visible. Snapping of the v,il\<- cannot be induced by
freeing the glochidia from their e^ iiu-inbr.ii

i i . Juvenile mussels; the foot i- m>\v visible, active -napping of
the valves and extension of the foot occui

1 Xo attempt was made to describe fully tin- van pment;

they are merely i haia. t> -ri/rd for greater accuracy in determining tin- duration of
developmental stages. Examination \va- made with a luno. ular mi> : "iag-

nit'yinj; X 50.

KDC.AR ALLEN".

Even juvenile mussels are found within the egg membranes.
A stringy transparent gel forms or is secreted about the egg
membrane enclosing the developing embryos after they reach
the marsupia. It is well formed by the stage at which they begin
to turn yellow and partly disappears in late metamorphosis.

Table II. includes observations on twenty typical individuals.
More than one hundred are included in the data from which the
average time required for the transition between stages is de-
termined.

TABLE II.

STAGES OF DEVELOPMENT OF YOUNG IN MARSUPIA OF .4. imbecillis.-

Date.

No. of Mussel.

7/24

7/27

7/31

8/4

8/9

8/15

8/22

8/28

9/i

i

•7

A

6

8

o

o

i

2

•2 ,

3

T.

4

4

6

7
6-7

8
8

9

o
o

3
i

4

1—2

A

6

7

8

o

o

o

o

o

I

s .

•2

4

6

n

o

I

6

-1

A

6

8

O

o

O

7 . .

7

8

o

o

o

I

T

8

8

i

1

7-8

0 . .

8-0

o

o

I

4

7

10

8-0

o

o

-i

II

6

7

s (J

O

I

3

12

7

8-9

O

o

~i

17 .

1—2

7

8-9

o

i

14 .

I

O

I

4

I?.

•5

4

6

8-9

o

3

16
17 ..

4
i

6-7

A

7-8
6

9
o

0
10— II

I
O

18
19

o
o

o
i

i

s

7-8

8

IO— II

9
o

20

6

8

1 1)

o

I

5

Numbers 13, 14, 17, 19 underwent a complete reproductive
cycle in from 22 to 27 days during August. The calculation of
cycle duration by averaging parts of cycles from all available
data gives an average cycle duration of 3 to 4 weeks. This
period then provides time for fertilization, segmentation, em-
bryonic development to the glochidial or larval stage, and meta-
morphosis of the glochidium into the juvenile mussel. In
Anodonta imbecillis it all occurs in the maternal marsupium, with
the possible exception of fertilization. Metamorphosis, which in

* o in the table indicates empty marsupia during the interval between repro-
ductive cycles. Numbers refer to stages characterized in the text.

RI PRODUCTIVE CYCLE IX AXODOXTA IMBECILLIS. 93

other forms occurs in epithelial "cysts" on the gills of fishes,
requires in this species only 7-10 days, as compared with a mini-
mum of 12 14 in Lampsilis anodontoides and L. litteola. Develop-
ment from tin- fertilized ovum to the fully formed glochidium
may occur in two weeks' time.

The duration of the interval between reproductive cycles
varies from 2 or 3 days to 2 (or rarely 3 )weeks.

It is reali/ed that more extensive observations would provide
material for more accurate determination of the duration of
various stages of this cycle and the possible changes in these time
relations during the fall and winter months.

However, the data at hand does warrant the condition that
Anodonta imbecillis cannot properly be classed as a long term
breeder. Juveniles are discharged during the latter half "I"
August and first half of September and a new reproductive cyele
begins. Whether this cycle is repeated throughout the month-
when this species is listed as gravid is still to be determined.

DISCUSSION.

The possible correlation of the rapid rate of growth and the
early attainment of sexual maturity of Anodonta imbecillis with
the thin "paper shell" has already been mentioned. It is possible
that this short reproductive cycle may be correlated with these
factors.

No attempt was made to determine whether self or cross fer-
tili/ation is the normal process in Anodonta imbecillis. Il >ell"
fertilization is the rule, hennaphroditism may h,i\c been .1 e<m-
tributing factor in the establishment of a shorter reprodm u\ t-
cycle.

Since to date only two species have been reported as de\ eloping
aparasitically, the idea that they may h.i\ r depended .it one time
upon parasitic metamorphosis, but later e\ol\rd .in ability to
develop independently, seems natural. Many iiue-ti^ators
have commented on the existence of -null hook- at the valve
edges as constituting the remains of a former pa i a -i tic adapta-
tion. Since parasitism on fishes which make -ea-onal upstream
migrations furnishes an efficient method of di.-tril union (possibly
(he only method of extensive upstream migration of mussels),
a u ide distribution of aparasitic form- \\< mid lend added weight to

94 EDGAR ALLEX.

this theory. But if the natural course of evolution of reproduc-
tion be followed, parasitism must be considered a secondary
development. It is possible therefore that metamorphosis with-
out parasitism is the primitive method and that Anodonta
imbecillis never has relied upon parasitism for development. It is
also possible that the shorter reproductive cycle is the primitive
one rather than an adaptation from one similar to that of other
species of Unionidae.

SUMMARY.

Anodonta imbecillis cannot be classed with either long or short
term breeders as defined by Lefevre and Curtis. In specimens
obtained at Fairport, Iowa, a complete reproductive cycle (in-
cluding metamorphosis from glochidia to juvenile mussels, for
reproduction in this species does not depend on parasitism) is
completed in from 3 to 4 weeks during the late summer. Another
cycle begins after an interval varying from 2 to 3 days to 2 \veeks.
Gravid A. imbecillis have been found in all months but April.
It is possible therefore that this short reproductive cycle is re-
peated throughout the year.

LITERATURE CITED.

Coker, R. E., Shira, A. F., Clark, H. W., Howard, A. D.

'i9-'2O Natural History and Propagation of Fresh Water Mussels. Bui.

Bureau Fisheries, Vol. 37, No. 893.
Howard, A. D.

'15 Some Exceptional Cases of Breeding among the Unionidea. Nautilus,

Vol. 29, p. 4-11.
Howard, A. D., and Anson, B. J.

'22 Phases in the Parasitism of the Unionidae. Jour. Parasitology, Vol. 22,

p. 68.
Lefevre, G. and Curtis, W. C.

'12 Studies on the Reproduction and Artificial Propagation of Fresh Water

Mussels. Bui. Bureau Fisheries, Vol. 30.
Lillie, F. R.

'95 The Embryology of the Unionida-. Jour. Morph., Vol. 10, p. i-ioo.

RESISTANCE OF RHABDITIS TO ACIDS.

HIKOKL'RO HON'DA.
DEPARTMENT OF ZOOLOGY. UNIVERSITY OF PENNSYLVANIA.

As is well known, Anguillula aceti lives in vineger and is highly
resistant to acetic acid. Henneberg (i) has shown that this
nematode can live in 13.5 {>er cent, acetic acid. Abbott ami
Richards have reported that it can live for more than 24 hours in
Tellyesnicky's fluid, and for three or four hours in GilsonV fluid.

The writer intended to see in the first place whether other
free-living nematodes, which do not come in contact with -u -;.<l-
in their natural life, were resistant particularly to acetic arid as
well as other acids. For this purpose Rhabditis elegans was ch< »ui.
In the course of the work it was found that Rhabditis is ni<>n
resistant than the tadpole, Daphnia, Aehsoma, and Paramecinm
not only to various kinds of acids, but to other toxic substanrr-
as well. The cuticle with which nematodes are covered is known
lo be composed of a very resistant substance; experiments,
therefore, were tried to determine whether or not the cuticle is
responsible for this resistance, the worms, injured and uninjim <l,
being stained in solutions of neutral red and of methylene blur.
'This work w.is begun at the suggestion of Dr. M. II. Jacobs,
whom the writer wishes to thank for valuable advice.

Adult hermaphrodites of Rhabditis elegans and tadpolr> about
io mm. long and with external gills were used. R. dc^nn\ i-
found in decayed matter and is easily cultivated in a prptonr
solution. Every experiment was repeated at lr.i-t twirr. .uul
14 to 30 individuals were used for each experimmt. The tem-
perature at which experiments were dour varied fnnii 21* C. i<>
26° C.

As shown in Table I., Rhabditis is the IDOM rcH-t.int to \.irious
kinds of acids. It might be noted that it can li\r about two hours
in N/3O acetic acid, while Daphnia, the form next in order of
resistance, can live only about an hour and a <|u.irirr in N/ioo
acetic acid. From the results of the experiment- -iven in the
lable we notice, however, that Rhabditis i- not particularly
n>i>tant to acetic acid as comparrd with othrr acids.

96

HIKOKURO HOXDA.

TABLE I.

AVERAGE TIME IN MINUTES REQUIRED TO CAUSE CESSATION OF ALL MOVEMENTS.

Rhab-
tlitis.

Para-
mecium.

Aelo-
soma.

Daphnia.

Tadpole.

HC1, N/ioo

60

<i

<i

2"?

12

HC1, N/so

I 7

<i

<i

o

7

Acetic acid, N/io
Butvric acid, N/io

23

IT

<i
<i

<i
<i

6

<r

Salicylic acid, N/ioo

26

<i

<i

T J

Is this resistance of Rhabditis due to a mechanical reason or to
the ability of the animal to neutralize acids? The results of the
experiments with toxic substances (Table II.) seem to show that
the first is the case, since Rhabditis is the most resistant of the
five forms studied not only to acids, but to the other toxic sub-
stances as well. If this presumption is correct, the cuticle with
which Rhabditis is covered may be supposed to prevent the pene-
tration of the toxic substances.

TABLE II.
AVERAGE TIME IN MINUTES REQUIRED TO CAUSE CESSATION OF ALL MOVEMENTS.

Rhabditis.

Paramecium.

Aelosoma.

Daphnia.

Tadpole.

XaOH N/40
Ether 5 per
cent

38

After 10 all
revived

<i

<i

<i
< i

15
After 10 none
revived

2

After 10 none
revived

HgCh 0.05
per cent. . .

60

<i

<i

25

5

We notice in these tables that Rhabditis is about 2.5 times as
resistant to N/4O NaOH, 0.05 per cent, mercuric chloride and
N/ioo HO than is Daphnia, the next most resistant form. Tad-
poles died more quickly in N/2O NaOH than in N/2O HC1. As
shown in Table II., tadpoles died in 2 minutes in X 40 NaOH,
while they died in about 4 minutes in \ 3oHCl. This is probably
due to the fact that the acid coagulates the body proteins as it
penetrates, \vhile the NaOH dissolves them. Tadpoles, which
were placed in the 5 per cent, solution of ether for 5 minutes,
revived though they were in a very bad condition. None of the
Rhabdites revived after placing them 30 minutes in the solution
of ether.

In order to study the penetration of an intra-vitam stain, some

RESISTANCE OF RHABDITIS TO ACIDS. 97

Rhabdites were placed in a small quantity of neutral red solution
on a slide, and were observed under a microscope. The stain
entered tli rough the mouth, and soaked out through the wall of
the esophagus. In one worm the stain entered through the anus,
but only a Miiall region of the posterior part of the body was
stained.

A solution of neutral red, which stained dead worms deeply,
stained living worms only faintly; a strong solution, then-fore,
was used. Staining for 2.5 hours in such a strong solution,
the anterior part of the body as faras the second bulb was
deeply stained, but in the remaining part the intestine only
was colored. The intestine was stained less deeply than the
anterior regions of the body, and the stain became progressively
lighter toward the posterior end. This seems to show that the
>t.iin entered through the mouth, but not through the cuticle
nor through the anus. After staining for 20 hours the worms were
>till alive, and the anterior regions a little beyond the bulb'of
the esophagus were deeply stained. In the remaining part the
intestine only was colored, and its posterior regions were stained
less deeply than its anterior regions.

Such stained worms were placed in a weak solution of XaOH,
and then the alkali entered through the mouth, and proceeded
towards the interior. In about 20 minutes the portion of the
body as far as the second bulb became yellow, but the change in
color did not go further. Later a rapid diffusion of NaOH from
the posterior end anteriorly was observed, and thus the entire
body of the worm changed to yellow; the alkali seemed to enter
through the anus. In another worm the change in colour pro-
ceeded anteriorly starting from the anus.

To show definitely that the cuticle prevents the entrance of
the stain and the XaOH, the posterior regions of some w< inn-
were cut off. Then the stain entered through the mouth and ilu-
cut end. Such worms were, therefore, Mained in toLo for 4
hours, and at the end of this time they were still alive. Contrary
to the case of uninjured worms the change in color carried by
ViOH began at both ends, and proceeded t,i-ter anteriorly from
the cut end than from the mouth. In about one hour and three
quarters the change in color has been completed in such injured
worms.

98 IIIKOKURO HOXDA.

As in the case of neutral red, methylene blue entered through
the mouth, and in no case penetrated through the cuticle. To
stain the entire body of Rhabditis a strong solution was used
for the same reason which has been stated in the case of neutral
red. After staining 24 hours, some worms were still alive, but
others were dead. The anterior part of the living worms as far
as the second bull) was deeply stained, hut the remaining part
was lighter in colour.

Such stained worms were placed in 83 per cent, alcohol. The
disappearance of the color began at the anterior end. The
worms soon died in the alcohol, and the latter then entered
through the anus. As far as the second bulb of the esophagus or
a little beyond it, the disappearance of the color was due to the
alcohol which entered through the mouth; in the remaining
regions to that which entered through the anus. A small region
of the anterior part of the intestine still remained stained after
one hour in alcohol. In one worm I observed that the posterior
division of the hermaphrodite reproductive organs was still
stained, while the posterior region of the intestine which is found
beside them was already reduced in color.

As in the case of neutral red, in some worms whose posterior
ends had been cut off, the stain and alcohol entered through the
mouth and the cut end.

SUMMARY.

1. Anguillula aceti is highly resistant to acetic acid. To see
what is the case in other free-living nematodes Rhabditis elegans
was used, and it was found that Rhabditis is much more resistant
not only to various kinds of acids, but to other toxic substances
also, as compared with the tadpole, Daphnia, Aelosoma and
Parameciiini.

2. When animals were stained in neutral red and in methylene
blue, no case in which the stain entered through the cuticle was
observed.

3. The entrance of XaOH and alcohol through the cuticle
also could not be observed.

4. The stains, NaOH and alcohol entered through the mouth,
and when the anus and vulva were opened, entered through them
as well.

RESISTANCE OF RHABDITIS TO ACIDS. 99

5. In the case of worms whose posterior ends had been cut off
these substances in question entered through the cut end, as
\\i-ll as the mouth.

6. The conclusion to be drawn from these results seems to U-
that the impermeability of the cuticle is responsible for the re-
-i-^tance of Rlmhditis.

LITERATURE CUED

1. Hcnneberg, W.

"oo Zur Binluxii- -Ics Easigaales (Anguillula aceti). Berlin. Gebr. Unger. >
(Cited from Zool. Cent.. 1900, p. 439).

2. Abbott, J. F. and Richard*, E. L.

'i i The Lethal l-.tlect of Pure L)islille<l Water on the Vineg;r J./M/.J

aceti). BIOL. BULL . Vol. 21.

Vol. XL VI March, 1924 -Vo.

BIOLOGICAL BULLETIN

LUTEAL CELLS IN THE GONAM <>1 THE PHALAROPE.

HARRY B. VOCOM.
DEPARTMENT OK ZOOLOGY. UNIVERSITY OF OREGON.

»

It has been shown by IVurl and Boring (1917, 1918) that there
are in the ovary of the domestic fowl certain interstitial eel,-
which are in many ways similar to, and which they claim are
homologous with the corpus hiteum in mammals. A thorough
study has convinced the authors that these cells are a con-tani
element in the ovary of all fowls but that they are not found in
the testes of normal male fowls.

Boring and Morgan (1918) made a study of the gonads of both
male and female Sebrighl Bantams, a race of fowls in which tin-
males and females are feathered alike. They found in both o\
and testis groups of cells like those described by Pearl and Bnrin^
as luteal cells, and ascribed to these so-called luu-al cells the
function of influencing the type of feathering. Since luteal cell-
are found in tin- ovaries of normal females or hrn-lr.it In-ivd
fowls and not in the testes of normal males or cock-feathered
birds, but in the testes of hen-feathered Sebriijn Bantam-, they
reason that the normal feathering of fo\\l- \\oiild br <>i tin- ni.tlr
type were it not for the suppressing influence «>t tin- Innnmm--
secreted by the luteal cells. Eurthcr proof in f.i\«>r «\ the con-
tention is obtained from the well known fact- "t « .i-iratioii where
hen-feathered females assume typical cm k-tV ather- upon the
removal of the ovary, and that the hen-feathered inalr Si-bright
becomes cock-feathered after the conipk-tr removal i»t" the testes.

It has seemed that further evidence concerning tin- (unction of
luteal cells might be brought to bear if a hi-ti •logical -uidy were
made of the gonads of birds otln-r than fo\\]-. ami especially some
bird in which the sex differences are rever-ed ; t hat is some race of

mi

IO2 HARRY B. YOCOM.

birds in which the females are more brilliantly colored than the
males. Such a sex difference is characteristic of the Phalarope.
During the spring migration when the birds are in full breeding
plumage the females are easily distinguished from the males by
the more brilliant plumage of the back and neck. Bailey (1917)
states the sex differences as follows : ' Male in breeding plumage ;
upper parts dark plumbeous, striped on back with buff and black:
sides of neck rufous, chest gray, upper throat and belly white.
Female in breeding plumage; brighter colored, rufous extending
across throat as well as on sides of neck. " To one not acquainted
with the facts concerning the sex differences of this group of
birds the sexes would undoubtedly be mistaken. During the
fall migration both males and females are lighter in color due to
the loss of the rufous color and more nearly resemble one another.

In one respect the sex differences of the Phalarope do not com-
pare with the sex differences of the fowls and that is in the fact
that the color differences in the Phalarope are not associated
with structural differences in the feathers, as is usually the case
In the fowls. A microscopical examination of the feathers of the
Phalarope showed that the color was due to pigment and not to a
difference in structure. Whether or not this is an important
difference and alters our problem can be determined only by a
study of a large number of birds or different races.

The material writh which this study has to deal consists of the
gonads of both male and female Northern Phalarope, Phalaropus
lobatus, Linn., collected during both fall and spring migrations
along the Oregon coast. The tissues were taken from the birds
as soon as possible after they had been shot, and were put into
Bouin's fluid and left until it was convenient to care for them in
the laboratory. Delafield's li.rmatoxylin was found satis-
factory and was used almost exclusively as a stain.

A studs' of the testes gave only negative results. In no case
were there found any groups of cells resembling the characteristic
packets of luteal cells found in the testes of the Sebright, and not
even single cells which might be considered of that type. The
intertubular material was not abundant and consisted of narrow
spindle shaped cells with oval, more or less shrunken nuclei.

In the ovary, however, characteristic groups of the so-called
luteal cells were found in the theca surrounding the oocyte follicles

I.I II \L CELLS IN GONAD OF PHALAROPE.

103

Figs, i and 2). These groups of cells are present in the material
collected in May and September but the groups of cells seem to

I.e.

^ &) -.' ls-

•'Co.

IB :>•>->•;-••.." v...,A ...-..'..- --

Fig. i. Section of ovary of Phalarope in fall plumage. X 1000.

I.e.

-

"

\^J 9f « * rt

-- ' • -•• . tr --- xf^-'*"' _1 JV - *•

St. , .., - -v^-

f-.c.

^fVV:x:;.,

t. •

V 7- ; ••< '.•

Fig. 2. Section of ovary of Phalarope in -piiny |>lum.ix«-. • i""o. /. c.,
follicle cells, /. c., luteal (• ;;ui.

be somewhat larger and more clearly drfinul in the material
collected in the fall when the sexes were imuv iu-.irl\- alike as
regards feathering. In no case could it ln->ai»l that the groups of

IO4 HARRY H. YOCO.M.

luteal cells were numerous. In many sections there would be no
packets at all while in others there would be one or two. In all
cases where found the packets were clearly defined and easily
demonstrable, being composed of cells having a relatively large
amount of cytoplasm with rounded nuclei. The cytoplasm of the
luteal cells takes very little of the haematoxylin stain. In this
respect as well as in shape and size these cells differ from the cells
of the stroma of the ovary.

The facts cited above do not indicate that luteal cells are in
any way influential in suppressing color development in the
Phalarope. In their consideration of the influence of the luteal
cells on feathering in fowls Boring and Morgan have not separated
the factors for color and structure in the feathers. In the present
study only the color of the feathers is considered. The fact that
luteal cells are found in the ovary while the female is so much more
brilliantly colored seems to eliminate the possibility that the
secretions of the luteal cells in any way suppress color develop-
ment.

The evidence so far obtained is not sufficient to warrant any
statement in regard to the problem of the association of color
differences with structural differences in the feathers.

The recent work of Nonidez (1922) raises another question
the solution of which will in a large measure depend on a more
extensive study of the presence or absence of luteal cells in the
gonads of birds. This author has shown that the so-called luteal
cells as designated by Pearl and Boring and Boring and Morgan
are in reality only the degenerating sex cords of the embryonic
gonad. What have been termed luteal cells and what have been
understood to be homologous with the corpus luteum of mammals,
Nonidez has shown to have an entirely different origin and are
therefore not to be confused in terminology with the mamma-
lian tissues. Whether or not these remains of the sex cords may
have an endocrine function is left l>y the author as an open
question to be determined by a wider study of a variety of mate-
rial. The facts as presented above as regards the presence of
these cells in the gonads of the Phalarope do not seem to bring us
any nearer a solution of the problem. There is no indication,
in the Phalarope, that these groups of cells exert a suppressing
influence on the development of the color of the feathers of the

LUTEAL CELLS IX GOXAD OF PHALAROPE. IO5

two sexes. That they may have some other endocrine function,
in the light of our present knowledge, is to be neither affirmed nor
denied.

CONCLUSIONS.

A histological study of the gonads of the Northern Phalarope
shows that luu-al cells such as are found in the gonads of Sebright
Bantams an- .il>H.-nt in the testes and present in the ovaries.
Since the fcin.ik- is more brilliantly colored than the male there is
no indication that these cells through an internal secretion influ-
ence the development of any differences in the color of feathers
of the two sexes.

LITERATURE.
Bailey, F. M.

"17 Handbook"; Birds of tin- Western L'nilcd States. Pp. lii -f- 574, 601 figures.
Boring. A. M., and Pearl, R.

'17 Sex Studies. II. Interstitial cells in the reproductive organs of the

chicken. Anal. Rec.. 13. 253-268. 6 figs, in text.
Boring and Morgan, T. H.

"18 l.uteal cells and hen-feathering. Journ. Gen. Phys. I, 127-131, 4 figs, in

text.
Nonidez, Jose F.

'22 Studies on the Gonads of Fowls. III. The origin of the so-called luteal
cells in the tcslis of hen-feathered cocks. Am. Jour. Anat.. 31. 109
7 figs, in text.
Pearl, R. and Boring. A. M.

'18 Sex Studies. X. The corpus luteurn in the ovary of the domestic fowl.
Am. Journ. Anat. 23, 1-36. o pis.. 6 figs, in text.

INTERACTIONS OF PROTOPLASMIC MASSES IX RE-
LATION TO THE STUDY OF HEREDITY AND EN-
VIRONMENT IN ARCELLA POLYPORA.1

BRUCE D. REYNOLDS,

FROM THE DEPARTMENT OF MEDICAL ZOOLOGY OF THE SCHOOL OF HYGIENE AND
PUBLIC HEALTH. JOHNS HOPKINS UNIVERSITY, BALTIMORE, MARYLAND.

INTRODUCTION.

In the first article of this series, Kepner and Reynolds ('23)
have shown that under favorable conditions Difflugia usually
reappropriates, by fusion, fragments of protoplasm which have
been severed from the cell-body. In the last paragraph of their
summary these authors make the following statement: "Obser-
vations have been made in which individuals of the same species
obtained from the same wild culture, showed a decidedly nega-
tive response towards each other's fragments, yet in other
instances such cross-fusions did occur." Elsewhere in their
paper two suggestions are made as to the probable explanation
of why some specimens did fuse \vith fragments belonging to
other individuals, viz: "(a) The individuals were closely related
by having a recent, common ancestor, (b] By living in the same
surroundings, the environmental influences have acted upon both
organisms in such a way as to cause an identical physiological
The question thus raised seemed to be of sufficient im-
portance to justify further work on the subject; consequently
the experiments incorporated in this paper were conducted for
the purpose of clarifying the situation.

The previous publication was largely concerned with describing
the process of protoplasmic fusion. Although a number of
observations were made upon the reactions between cell -bodies
and fragments from other individuals of the same species, no
attempt was made to culture the organisms and then test for
reactions between members of a clone. As a result, only two

1 This is the second article of a scries dealing with the phenomenon of fusion
between protoplasmic masses. The first article was written by Win. A. Kepner
and B. D. Reynolds ('23) from the University of Virginia. Further work on the
subject is now in progress.

1 06

INTERACTIONS OF PROTOPLASMIC MASSES. IOJ

types of reactions were observed: (a) fusion and (6) failure to
attract, or if contact was made the fragment would be pushed
along for a while, without causing any apparent disturbance, and
then abandoned.

During tin- course of this investigation, involving much more
extensive experiments, it was found that in some cases the
masses would readily attract l each other, but upon making con-
tact there would be a sudden contraction of the involved regions
as if violently shocked, while the protoplasm constituting both
the fragment and the organism's pseudopod (5) which had made
contact with it would be shattered into bead-like masses. See
Fig. i : A, B, and C. This type of reaction was observed so fre-
quently that it is referred to throughout this paper as the "shat-
tering" reaction.

As the investigation progressed it seemed advisable to divide
the problem into three sections; and to study the reactions be-
tween organisms and fragments from related specimens under the
influence of three different conditions as follows:

I . Similar environments.

II. I >ilferent environments.

III. Identical environments.2

Numerous methods have been employed in studying variations
among the proto/.oa, such as: rate of reproduction, measurements
of size, counting the number of spines, teeth, tentacles, etc., resis-
tance to poisons and to hiiji temperatures. In tin- Miuly, for
the first time, advantage has been taken of the property possessed
by certain organisms of fusing with their own or frav;inrin- lr»m
closely related specimens. The superiority of this method lie- in
the fact that it affords a quick and definite means of di tci iniuin-

1 In tliis, as in the previous paper, it was n ;mi.-.l why :

were attracted to each other. When pn>i»pl.i-n. t present the

organisms would move about at random; but \\lu-n \- m wliii h liad

been detached from themselves or from related individual- \\n«- \\ ithin two or three
hundred micra of them, they would invariably ] irds the frag-

ments. This statement is based on over liiu-rii liundn-d ob us. A series

of experiments are being planned to ascertain, it ; nulus involved in

these positive reactions.

- The organisms comprising the two lines undi-r ol n were placed in the

same concavity, or else, portions of the culture nn-diu v. uently exchanged

tween the concavities containing them.

108 BRUCE D. REYNOLDS.

minute physiological differences — even before these changes have
become expressed morphologically. Furthermore, the physio-
logical difference between two organisms, as indicated by this
method, appears to be more permanent and less subject to fluc-
tuations than certain morphological variations which have been
utilized by others. In some instances cross-fusion was observed
to take place readily between specimens belonging to the same
clone — though they appeared quite different to the eye. In other
words: the physiological variations, as determined by means of
the ability or inability of an individual to fuse with a fragment
of protoplasm belonging to another, seem to be independent of
differences in size or shape of the two animals concerned.

Fusion between cell-bodies and severed fragments has been
observed by the author in ten species of the genus Difflugia, in
Arcella poly pom, Centra pyxis aculeata, and also other forms of
protozoa. The phenomenon seems to be fairly common among
the thecamceba-. Perhaps it would be well to emphasize at this
point the fact that throughout these experiments any individual
Arcella would fuse with its own fragments (provided they were
not too far removed, or had not been separated for too long a
period) under any environmental conditions to which they were

subjected.

ACKNOWLEDGMENTS.

This problem suggested itself to me while I was discussing
with Professor Wm. A. Kepner, the results of our work ('23)
Reactions between Cell-bodies and Pseudopodial Fragments of
Difflugia, I wish to express my deepest gratitude to Professor
R. \V. Hegner and Dr. Wm. H. Taliaferro, under whose direction
the work was done, for their valuable advice and criticisms. I
am also deeply indebted to Professor H. S. Jennings for reading
the manuscript and making some important suggestions.

THE ORGANISM.

The experimental animal used in this investigation was a mul-
tinucleated Arcella closely resembling the species described by
Penard ('85) as A . poly pom. Hegner ('20) worked with the same
organism, calling it A. polypora. However, in the same paper he
also gave measurements of a larger multinucleated form which

J\IIR\( IIONS OF PROTOPLASMIC MASSES. IO9

agrees with the amended description of .4. polypora given by
Penard ('92; vix: "/r£s grande, a bouche tres largement ouverte,
et toujoiirs ponrvne de noyaux en nombre considerable.'1 Appar-
ently Hegner i- the only investigator in this country who has re-
ported the existence of Arcellae containing many nuclei. Leidy
('79) states that he was able to detect two, occasionally one. and
rarely three, though admitting that he had possibly overlooked
the presence of some nuclei in the larger forms of Arcella. Occa-
sionally large, polynucleated, wide-mouthed specimens were found
in my collections. The differences between these two multinu-
cleated forms seem to be great enough to justify the creation of
a new species. However, if this is done, the law of priority makes
it necessary to retain the name .-1. polypora for the smaller torm
which was used in these experiments.

Because of the existing confusion it is deemed advisable to say
a few words regarding certain characteristics possessed by the
organism studied: shape, structure, and coloration of shell simi-
lar to A. discoides; diameter of shell from 80 to 120 micra; height
of shell from 25 to 40 micra; oral orifice approximately equiva-
lent to one third the diameter of shell; number of nuclei usually
eight or ten. Careful scrutiny under the oil immersion failed to
reveal any minute pores around the mouth of the shell, but the
shadows at the outer margins of the large cancelli lining the
buccal aperture might readily be misinterpreted as such under a
low magnification.

Arcella polypora was selected as the object for study because it
reproduces rapidly, is easy to culture, and is capable ot with-
standing the influences of various stimuli. This rhi/op.xl may
be found in either large or small bodies of water, li\ iir^ e-|>t;rially
on the under surfaces of aquatic plants, in deca\in^ vegetable
matter, and ooze. On April loth. 1922 a < ..IK-i ii<>n, runiaining
main" -1. polypora, was obtained from an artificial pool in the
botanical garden near the Johns Hopkin- Mi< "logical Laboratory.
All of the organisms used in these experiment- \\rre descendants
of this lot. A mass culture was kept going, and a- new individuals
were needed they were isolated and new (lone- \\viv started from
them.

110 BRUCE D. REYNOLDS.

METHODS.
(a) Apparatus and Technique.

In carrying out these experiments more standardized and exact
methods were employed than those used in the first article of this
series, though they were essentially the same. In every case
(unless specified) distilled water (pH •• -- 0.7) was used, and each
animal was washed once or twice before being transferred to the
slide on which the amputations were made. Furthermore, any
detritus clinging to the shell was removed before the operation.
In all of the cross-fusion experiments the two organisms to be
tested were placed on the same slide, and when a fragment was
severed from one, that individual was moved away and the other
placed near the fragment of protoplasm. The glass needles,
with which the protoplasmic masses were detached, were drawn
from standard Pyrex tubing \ OD furnished by the Corning Glass
Company, Corning, N. Y. All operations were performed under
the compound microscope — 32 mm. objective and No. 10 eye-
piece (B. & L.) — while the 16 mm. objective was used in making
the observations. A camera lucida was attached to the micro-
scope and at intervals (usually one minute) the outlines of the
objects were traced with a pencil. These sketches were 170
times the size of the objects.

The pseudopods of A. polypora are usually digitate and rarely
extend far beyond the periphery of the shell, making it very
difficult to sever masses of protopiasm by cutting. them off with a
needle as was done in the case of Difflugia. This difficulty was
overcome by taking advantage of two things — (a) the shape of
the shell, and (b) the fact that if undisturbed for a while the
animal ordinarily attaches itself to the substratum. The or-
ganism is shaped somewhat like a shallow bowl, the outer rim of
the shell forming a rather sharp edge, Fig. 2. If, after a pseudo-
pod has been extended beyond the outer margin, pressure is
applied to the top of the shell the pseudopocl will be caught be-
tween the rim and the sub-stratum and its distal end will thus be
pinched off. Severance is aided by the animal's sudden reaction
to this stimulus. Sometimes a specimen would fail to protrude
pseudopods of sufficient length to permit use of the method just
described; consequently it was left undisturbed for a few min-

INTERACTIONS OF PROTOPLASMIC MASSES. Ill

utes and then suddenly displaced by means of a glass filament.
This procedure would frequently result in a large mass of proto-
plasm being torn from the organism.

(b) Culture.

Hollow-ground >lides, each containing two concavities, were
used as receptacle- in cultivating the organisms. Seven drops of
the desired culture medium were put into each concavity and the
animals were transferred to them. After this the slides were
labelled and placed in transparent glass moist chambers which
were kept before a window, unless otherwise stated.

At first the method described by Jennings ('16) for Dirfln^ia
was used in preparing the culture medium, i.e., a quantity of
ooze from ;i pond in \\hich the organisms were found to be living
was shaken up in water, this was allowed to settle, and then the
water was decanted oft and used. Although other investigators
have obtained successful results by using culture media prepared
in tlii> way, several objections to its use in these experiments
were evident: first, fresh material was hard to obtain, since no
ponds were near the laboratory in which the work was done;
secondly, a constant culture medium was not insured, for the
plant and animal life in a given j>ond vary constantly; and
finally, it was necessary ever to be on guard against possible con-
tamination with wild specimens and injurious chemical agents.
After encountering these difficulties for two months the- above
method was discarded in favor of a hay infusion culture medium
which dispensed with all of the objections mentioned, and in
which the organisms thrived even better. It was prepared as
follows: 10 grams of clean timothy hay was placed in a clean
beaker containing 250 c.c. of neutral distilled water. This was
boiled slowly for five minutes, then strained through two thick-
nesses of cheese cloth, after which it was .-tc.ivd. in quantities of
about 3 cc., in small sterile test tubes. Tin- tube- were then
plugged with cotton and placed in boiling water for fifteen min-
utes. Two days later they were subject i-d to boiling water again
for the same period of time, in order to kill any }>.« u-ria which
might have resisted the first sterili/ation by being in the spore .
Mage. The liquid in the tubes constituted what wa< known as
stock solution; it would keep for month^ wit IK mt deteriorating

112 BRUCE D. REYNOLDS.

provided evaporation didn't take place. In making up the cul-
ture medium one part of the stock solution was taken and to it
were added nine parfs of distilled water — giving a 10 per cent hay
infusion solution. When other ingredients such as sugar, alcohol,
etc., were used, they were mixed with distilled water in the desired
proportions and then nine parts of this were added to one part of
the stock solution. Thus, in every case the percentage of hay
infusion remained the same. After a tube containing some of
the stock solution had been opened and a part of its contents
used the remainder was discarded. The Arcellcr grown in this
medium subsisted largely on bacteria, the progenitors of which
were introduced with the organisms.

(c) Pedigrees.

"Pure" lines, or clones, were obtained by isolating specimens
from the wild culture and designating them, for example, as the
progenitors of clone A, clone B, etc. Upon dividing, the daugh-
ter-cell remaining in the old shell, and all of its descendents, were
referred to as clone A line a, while the daughter-cell occupying
the new shell, and all of its descendents, were denominated clone
A line ad. Fig. 3 expresses graphically the procedure followed.

Generations: ist.

2d. 3d. 4th. 5th. 6th. etc.

Clone A
(Isolated

cultures)

FlG. 3. Diagram illustrating method employed in recording the relationship of
experimental animals. Those marked with an asterisk were either used for obser-
vations, or else discarded.

When a different environment was imposed on one line of a
clone that line was designated, for instance, as clone A line SL-SU
(sn indicating that sucrose was added to the culture medium in

IVII.KAt I lo\S OF PROTOPLASMIC MASSES. I IJ>

which this line wa~ nourished). Ordinarily lines were perpetua-
ted through the individuals remaining in the old shells, but this
rule was not strictly adhered to since experience showed no differ-
ence in tin- results obtained when other members were used.
From two to four individuals of each line were kept in culture as
a precaution again-t possible fatalities.

I SIMILAR ENVIRONMENTS.

These experiments are discussed under the heading "similar*
environments because the external conditions surrounding the
two lines of the clones involved were as nearly alike as it was
possible for the experimenter lo make them. Though kept apart
from the first division of the clonal parent, the two diverging
lines were treated in exactly the same manner; for instance: t In-
same pipette was used in transferring members to fresh concavi-
ties; these concavities contained culture medium taken from the
same bottle; afterwards the depression slides containing the
organisms were placed alongside each other in the same moist
chamber. Frequent tests were made, according to the method
described above, between individuals of one line and fragments of
protoplasm from individuals of the other. It was found that
daughter-cells (members of the second generation) would readily
fuse with each other's severed fragments — and that such reactions
would occur regardless of whether or not the organism taking up
the fragment had just previously lost a part of its own protoplasm.
Furthermore, in every case when both lines were kept under
similar conditions, cross-fusions continued to be exhibited for
fifteen or more generations; consequently in later experiments
tests for interactions were not begun until several generations
had elapsed.

A series of experiments conducted between the two lines of
clone I are given in extenso in Table I. It \\ill be seen by re-
ferring to this table that cross-fusion between line i and line id
persisted until the 22d day and 24th generation. After this no
further inter-fusions occurred, but in-tead the prot..pla-ms usu-
ally were shattered upon making contact. Attention -hoiild be
called to the fact that in some of the experiment li-ted in Table I.
the organism was on its back when placed near the fragment ; in
other cases the fragment would become detached from the sub-

BRUCE D. REYNOLDS.

TABLE I.

INTERACTIONS OF CELL-BODIES FROM ONE LINE AND PSEUUOPODIAL FRAGMENTS
OF SPECIMENS FROM THE OTHER IN CLONE I, LINES i AND id. BOTH LINES
WERE EXPOSED TO SIMILAR ENVIRONMENTS. 10 PER CENT. HAY IN-
FUSION USED AS CULTURE MEDIUM. EXPERIMENTS WERE STARTED
ON JUNE 25, AND ENDED ON JULY 31, 1922.

0

s

e

00

e n

^

1*4

O> I

8 ^

£u .

^ f

O

s >•

0 .

— *i~

.E a

£>,S ^

.E c "o

rt ~ p

h

.2 E

- s

~^'Z 2

•°j8|J

'2"2%

Reaction after

Q._ u

£ i

2S

" ° &> —

o — ,^^

= u in

Contact.

-- " _i

O *T*

O 13

o § B "

^ r^\ c

"... J? rt

* 51

So

Si

2 o

^avJ<g.M

P"~ ^

° ti

o

u

§"

5

H*

16

1-18

id -i 7

9x25

50

2

Fusion within i'.

1 6

id- 1 7

i-i8

IOXI8

IOO

3

Immediate fusion.

16

i-i8

id- 1 7

1 5 x 40

no

3

Fusion after 3'.

18

i-2O

id-19

12x15

90

-

No contact.

18

id- 1 9

i-2O

10 X48

85

i-5

Fusion within 30".

22

i-24

id-24

30 x 40

1 60

4

No fusion.

22

id-24

1-24

9x30

150

3

Violent contraction.

22

id-24

i-24

12 X 25

55

2

Shattering reaction.

25

1-28

id-27

8 x 38

95

-

No contact.

25

i-28

id-27

7 X45

48

-

No contact.

25

id-27

i-28

45 x 55

32

I

Shattering reaction.

27

id-28

i-3i

10 x 40

10

0-5

Shattering reaction.

29

i-34

id-32

14 x 20

82

2

No fusion.

29

id-32

i-34

12 X 30

15

3

Shattering reaction.

29

id-32

i-34

25 x 35

6

3

Shattering reaction.

32

i-38

id-36

20 x 30

62

-

No contact.

32

i-38

id-36

20 x 30

o

o

Slight shatteiing.

32

id-36

i-38

24 x 26

40

2

Slight shattering.

32

i-38

id-36

1 0 X 6O

JO

6

No fusion.

32

id-36

i-38

15 X45

47

i-5

Shattering reaction.

32

i-38

id-36

8x25

o

o

No fusion.

36

i-43

id-4i

7x 70

78

2

Shattering reaction.

36

i-43

id-4i

6 x 40

105

4

Shattering reaction.

36

id-4i

i-43

20 x 60

53

I

Shattering reaction.

stratum and float away, or else be pushed along by the organism's
pseudopods \vithout making intimate contact with them. This
accounts, to a great extent, for discrepancies observed in the time
required to make contact, and also for the ca>c> marked No
fusion and No contact.

A similar table might have been arranged for any of the other
clones studied in this manner, but to take up so much space with
tables would be both impracticable and unnecessary; therefore
a single table has been prepared which gives the number of days
and generations elapsing before the intraclonal reactions became
negative (as indicated by shattering instead of fusing of the pro-

INTI-.kA' 1 ION'S OF PROTOPLASMIC MASSES.

115

toplasms). This table (Table II.) shows a range of from 18 to
27 generations, and from 15 to 32 days, in the time of the first
appearance of negative reactions between two lines of a clone, or
an average of 22.25 generations and 21.58 days. After two lines
had once become negative to each other no further cross-fusions
took place between them unless certain modifications were made,
the nature of \vhirh will be discussed later.

TABLE 1 1 .

TABLE SHOWING DATE AND GENERATION OF FIRST EXHIBITION OF SHATTERING
REACTION IN TWELVE CLONES OF ARCELL.C. BOTH LINES OF THESE CLONES
WERE EXPOSED TO SIMILAR CONDITIONS.

rst Xcg. Reaction Observed.

Experi-
ment!
Begun.

Clone.

No. o-
Made.

Culture Mrdium
used. N<

>. of Generations.

Days

C|n«M

Line. zd. Line.

uUtvC

May 5

.-I

Pond water

24

IS

5. •

/<

. . ..

21 21

21

38

10. .

O

. . . •

3

5- •

£>>

»« ..

16

June i . .

E

35'., hay iiiin.

23 27

32

25

i . .

P

I'.w. and 11. i.

23 23

30

37

" 25..

/»

ioc'c hay infu.

24 24

22

24

25- •

J

* * ' * * *

18 18

l.S

39

25- •

K

" " '•

25 25

18

37

25 •

1.

. . ** • •

2O 20

17

30

25- •

0

ii < • •*

18 18

16

16

25- •

r

i* .. ..

i<) 20

18

21

AUK. 29. .

Q

4i I* ••

26 20

23

8

29- •

R

25 24

23

18

Average: 22.

2.S gen.

21.58 clays.

He-fore these cultures were discarded the culture medium, in
which the organisms were living, was tested for its pi I \.iliu- by
means of the Phenolsulphonephthulein indicator. In r\n\ •
it was found to be neutral; i.e., no detectable change had taken
place because of the presence of the org. mi-ins.

SUMMARY OF SECTION I.

In summing up the results obtained I'nun tlu-e experiments,
the evidence indicates that physiological difleremv- do de\elop

'Cross-fusion obtained on the i6th day. Cultuir< >lin| l :;h.-r obser-

vations were made.

1 Cross-fusion obtained on the xoth day — cultun - ill-
I'liis set of experiments is given inexu-nsn in T.iMr I

Il6 BRUCE D. REYNOLDS.

among the members of a clone, without necessarily being accom-
panied by corresponding morphological differences, and that
these changes can be demonstrated by the cross-fusion method
within relatively short periods of time — approximately 22 days.
Since both lines of a given clone were subjected to the same exter-
nal influences, the inference might be drawn that such proto-
plasmic changes are probably due to inherited variations and not
to the action of environmental factors. However, later experi-
ments show that this position is untenable.

II. DIFFERENT ENVIRONMENTS.

Having determined the time required for intra-clonal differen-
tiation under similar environments, it was now possible to subject
the two lines of a clone to different environments and observe
the relative effects such treatment would have on their behavior
towards each other's fragments. In carrying out these experi-
ments such factors were employed as might occur in nature, viz:
(a) 25 per cent hay infusion, (/?) sucrose, (c) saline, (d) acetic
acid, (e) sodium carbonate, (/) alcohol, (g) darkness, and (h)
temperature. The organisms seemed to thrive best in 10 per
cent hay infusion, and since this could be prepared by a given
formula, it was used in every case (except one) as the standard
culture medium: i.e., one line of a clone was grown in it at room
temperature in front of a window, while the other was altered
as desired.

(a) 25 Per Cent Hay Infusion.

An organism was isolated from a wild culture and placed in a
concavity containing pond water. Upon dividing, the new cell
thus formed was transferred to a receptacle containing 25 per
cent hay infusion and designated as clone H line h-Jiay, while the
individual remaining in the old shell was placed in a concavity
containing fresh pond water and now became known as clone
H line h. These two daughter-cells were then placed in the same
moist chamber, so that the only difference in their environments
was represented in the culture media. The usual tests for cross-
fusion between the lines were made with the following results:
on the second day after the beginning of the experiments +, 4th
+ ,6lh +,8th--o,()th -o--.1

1 Tin.- significance of these symbols is as follows: + fusion occurred; — shatter-

INI1.KX' 1IOXS OF PROTOPLASMIC MASSES. IIJ

A similar -eries of experiments were conducted between two
diverging lines of clone \Y. In this case seven observations were
made, extending through the eleventh generation, with the result
that cross-fu-ion took place in every instance. Further obser-
vations were ni.i'lt impossible on account of death in one of the
lines. In one case then, the time required for two related lines
to exhibit the negative reaction was greatly reduced, while in
the other death interfered with the determination of this point.
If standing alone, further evidence would certainly be necessary
before much emphasis could be placed on the significance of these
observations; but, since the conclusions drawn, from the experi-
ments dealing with different environments, were largely deducted
from later observations, it was decided to include these as they
stand without attempting to add more convincing proof.

(b) w Per Cent, liny Infusion Containing I Per Cent. Sucrose.

In these, and all succeeding experiments, the standard culture
medium was 10 per cent, hay infusion. After the first division
of the clonal parent, one of the daughter-cells was transferred to
a concavity containing 10 per cent, hay infusion plus I per cent.
sucrose, while the other was kept in the standard culture medium.
In every other respect the two lines arising from these individuals
were subjected to similar influences. In order to avoid possible
contaminations, a different pipette was used in transferring mem-
bers of these two lines to fresh culture media. The density of
the medium was, of course, increased ujxm the addition of sugar,
which resulted in the newly formed shells becoming smaller ami
smaller, so that it was considered advisable (perhaps necessary)
to remove the osmotic influence after six or seven days; conse-
quently, after having remained in the medium containing I per
cent, sucrose for approximately that length of time, tin organisms
were put back into 10 per cent, hay infu-i<>n. From thi- time on
both lines of a clone were living amid>t the -aim -urnuindings.
It was found that cross-fusion between the two line- continued
for five or six days, then for a period , ,|" -i\ , ,r -«-\vn days the
interactions were negative (even though the n-motic agent had

ing reaction; o contact made but neither fusion M<>I -li.itii im^ !»lln\viMl. The
number of symbols placed after a figure iiiilii.it<-- tin- munlx-i vatiohs made

on that day.

BRUCE D. REYNOLDS.

been removed) after which another short period of inter-fusion
ensued. This is clearly shown in Table III.

TABLE III.

SHOWING INTERACTIONS OF RELATED LINES, k AND k-s«. THE ENVIRONMENT
WAS THE SAME FOR BOTH LINES, WITH THE EXCEPTION THAT FOR 7 DAYS LINE
k-ju WAS KEPT IN A i PER CENT. SUCROSE SOLUTION. THE EXPERI-
MENTS WERE BEGUN ON JUNE 27, AND DISCONTINUED ON JULY 22,

1922.

k.
c

IM

g c

c

CO

o •*-*

l2g

o .

ii

c c •
Q " 2

|if _.

ll|

c-*2 .«

ll

'? E

?'*"'•"

— -o S '-

S(3s

Reaction after

rt ,
- _-.

Orr

£3

- rt

C — ^

"•? u=

.S v v>

Contact.

O D X

e,4!
o-J

•- i-.

eta

S° e
._ »"•

5^ 2

%<£ >

^fflW

'O

O

o o

.S^^1

So""

•- J2

12

Q

H

2

k-jH-3

k-2

9 X34

75

10

Immediate fusion.

2

k-2

k-^w-3

12 X 18

40

2

Fusion within 15".

3

k-4

k-5/t-4

12 x 40

93

4

Fusion after n'.

3

k-su-4

k-4

8 x 24

63

2

Fusion within 15".

3

k-sz<-4

k-4

9X33

67

2-5

Fusion within 15".

3

k-sz<-4

k-4

8 X 20

112

3

Fusion within 30".

4

k-6

k-su-5

15x80

45

3

Immediate fusion.

4

k-6

k-5M-5

8x35

60

3-5

Fusion within 30".

4

k-Stt-5

k-6

8x45

o

o

Fusion after 2'.

6

k-8

k-sw-61

35X75

95

3

Shattering reaction.

8

k-9

k-s/<-72

12 X 3O

9

i

Shattering reaction.

9

k-io

k-5zt-93

27 x 80

95

2

Shattering reaction.

9

k-sw-9

k-io

12 X 15

55

-

No contact.

II

k-su-i i

k-i3

9X25

85

I

Shattering reaction.

II

k-i3

k-su-i i

8x 14

32

0-5

No fusion.

13

k-i6

k-5?<-i4

24 x 26

26

2

No fusion.

13

k-sz<-i4

k-i6

10x45

13

I

Immediate fusion.

13

k-sw-14

k-i6

8 x 40

90

2

Immediate fusion.

13

k-i6

k-5M-I4

9x32

63

1

Immediate fusion.

13

k-i6

k-5M-i4

8 X30

70

I

Immediate fusion.

13

k-sM-14

k-i6

25 X30

40

2

Fusion within i'.

13

k-i6

k-su-14

10 X 56

85

2

Fusion within i'.

14

k-I7

k-su-is

12 X 3O

140

4

No fusion.

15

k-19

k-su-i6

IO X 2O

115

2

Immediate fusion.

15

k-5«-i6

k-ig

8 x 40

90

-

No contact.

20

k-5Z/-2 I

k-24

25 x 60

37

4-5

Shattering reaction.

23

k-2 8

k-su-25

9 x 50

IO

i

Shattering reaction.

23

k-5Z<-25

k-2 8

8xio

33

i

Shattering reaction.

25

k-SM-28

k-3i

12 X 30

62

i

Shattering reaction.

25

k-3i

k-xu-28

8 x 25

35

i

Shattering reaction.

1 Later specimen k-sit-6 fused with its own fragment which had caused a shat-
tering of specimen k-8's cytoplasm.

• The young individuals of line k-sw, having been subjected to the influence of
i per cent, sucrose for 7 days, were reduced in size to a diameter of 65 micra (the
normal diameter being 100 micra). On this day line k-su was put back into 10 per
cent, hay infusion, where it remained during the rest of the experiments.

3 Later specimen k-sw-Q fused with its own fragment after it had been abandoned
by specimen k-io.

I\ I 1 kACTIONS OF PROTOPLASMIC MASSES. I IQ

By referring t<> Table III. it will be seen that cross-fusion be-
tween tin- two lines persisted through generations k-s«-5 and
k-6, while two d.iys later (k-sw-6 and k-8) their potency to fuse
with the otlu T'- fragments had teased. On the 2d day after the
first negative reaction was observed. linek-5/f was taken out of the
sugar solution and placed in the same medium used for line k;
nevertheless, tin- reactions between the two lines continued to be
of the shattering type for 4 more days; i.e., cross-fusion was not
observed again until the fifth day after the two lines had been
restored to a common environment (k-s«-i4 and k-i6). For
the next few days the inter-reactions were positive, but from
generations k-5M-2l and k 24 until the experiments were dis-
continued, shattering of the protoplasms t<x>k place in every
instance. Compare the second shattering periwl with re>ulis
shown in Table 1 1

An interesting jx)int brought out by this study, is the fact that
though the normal si/c was regained two days after the individu-
als of line k sn were removed from the sucrose solution, the
inability to cross-fuse with fragments of specimens from line k
persisted for five days. In other similar experiments the stability
of this physiological variation as compared to the morphological
dillerence in si/e was shown even more strikingly.

Three other series of experiments, in which I per cent, sucrose
was used as the modifying factor, were conducted, and all of them
gave results parallel to those shown in Table III. For instance,
on June Jth the progenitor of clone M was isolated from a wild
culture and upon dividing the next day, lines m and m-sn were,
started. Cross-fusion between these two lines was observed on
the third day (generations m-4 and m-SM-4), shattering reaction
exhibited on the oth (m ~ and m .v//-6), line m su put back in
standard culture medium on the 7th, ne^athe inter-. n-iions per-
sisted until the i.sth (m-i8 and m-s//-i5 I, then . To^-fu-ii.ns took
place during the next five days, all of the interaction- after this
were of the shattering type. Twenty observations \\ere made in
this series of experiments.

Observations in clone J, lines j and j -sn were U^im on August
27th. Cross-fusion between line j and line j •>/< on the 4th day
(j-4 and J-S//-5), negative on tin- <>th. ( j <> .md j vw-j), line
j VK removed from I per cent, surro-e on oth, ne-.iiive inter-

I2O BRUCE D. REYNOLDS.

actions persisted until the I5th (j-i6 and j-s«-i6); this was
followed by a brief period of cross-fusion, after which the inter-
actions became permanently negative — number of observations
in this case was 35.

A similar series of experiments between lines n and n-su of
clone N were started on August 2yth. In this case line n-su was
removed from the sugar solution before the negative condition
had been attained; consequently cross-fusion continued until
the 1 5th day. From this time on all of the interactions were
negative. Forty observations were made in this series of experi-
ments.

(c) 10 Per Cent. Hay Infusion Containing N/io Saline.

Two series of experiments were conducted in which one line of a
clone was placed in a culture medium of 10 per cent, hay infusion
plus one tenth normal saline solution, while the other line was
kept in the standard medium — the only difference between the
environments of the two lines was that the culture medium used
for one contained 0.09 per cent. salt. During the first three days
the organisms in the saline solution seemed to grow and repro-
duce normally; after this, three days elapsed before another
division occurred ; then further growth ceased. In the meantime
tests showed that cross-fusion between the two lines took place
readily. After keeping the organisms in saline solution for eight
days they were put back into standard culture medium. The
interactions continued to be positive for two more days, then the
experiments were discontinued.

The evidence obtained from these observations indicates that
N/io saline is an unfavorable environment for A. polypora; but
regardless of this, no physiological change that can be detected
by the cross-fusion method was thus produced.

10 Per Cent. Hay Infusion Containing Acetic Acid.
Numerous attempts were made to culture the organisms in 10
per cent, hay infusion to which had been added various amounts
of acetic acid, but after repeated efforts it was found impossible
to keep the animals alive in solutions containing only o.oi per
cent, acetic acid. Other acids, or weaker concentrations of
acetic acid, were not tried.

INTERACTIONS OF PROTOPLASMIC MASSES. I -1 I

(e) JO Per Cent Hay Infusion Containing 0.025 Per Cent Sodium

Carbonate.

By beginning with one per cent, sodium carbonate and each
time decreasing the proportion, it was determined that .4. />o/v-
pora could be successfully grown in a medium containing 0.025
per cent, of this substance. Only one set of experiments was
conducted, but the results obtained were very definite. On
June 2Qth an individual was taken from clone K and set aside to
become the progenitor of a new clone S. The next day it had
divided, so one of the daughter-cells resulting from this division
was placed in 10 per cent, hay infusion containing 0.025 Per cent,
sodium carbonate, while the other was kept in standard medium.
The size, activity, and rate of reproduction of the organisms sub-
jected to the influences of sodium carbonate (a weak base) seemed
to be perfectly normal. Results of the reactions between line s
and line s-s.c. are given by means of symbols (See footnote page
1 1 6 for meanings). On the second day after the experiments were
begun the interactions were H — K 4th o o +, 6th line S s.c. per-
manently transferred to standard medium, 8th - -, Qth H — (-,
nth + + , 1 4th - o, 1 9th — , 2 ist 23d - -. No further

observations were made after this date. These results strongly
suggest that, if one line of a clone be placed in 10 per cent, hay in-
fusion and the other line in a similar medium to which 0.025 Per
cent, sodium carbonate has been added, the physiological consti-
•tutions of the members comprising the two lines will quickly
become so differentiated that cross-fusion will not take place
between individuals belonging to one line and fragments of speci-
mens from the other. Such changes are shown to persist for some
time after the causative agent has been removed.

(/) 10 Per Cent. Hay Infusion Containing 1 Per Cent.

Alcohol.

On August 28th specimens were taken from two (loin--, which
had been under observation since June 251)1, and \\ith tla-in two
sub-clones, J and K, were started. These Mib-rloiu-; v
divided into two lines each; one of which w.i^ krpt in M.md.inl
medium, while the other was placed in 10 pi-r rmt. h.iy inl'ii-imi
plus I per cent, alcohol — the latter bun- iir-i;;n.iti-o! MI!>-<IMIH-

122 BRUCE D. REYNOLDS.

J line j-a, and sub-clone K line k-a respectively. The organisms
subjected to alcohol seemed to thrive quite well — the rate of
fission, in both lines being slightly greater than that of those kept
in standard medium. It was found, by testing the interactions
of related specimens grown in these different environments, that
the results were parallel in both cases; yet different from those
obtained when substances other than alcohol were used for alter-
ing the environmental conditions. These observations were
extended over a period of 65 days, representing 66 generations
in sub-clone J and 63 in sub-clone K. The results are indicated
by symbols (see footnote on page 116 for meanings). For
sub-clone J, reactions between line j and line j-a were as
follows: on the 3d day + o, 5th + o, '8th - -, loth

+ +, 14111 - - o — , 1 6th -- o, i8th -- o o, 2ist + + , 23d +
+ +, 29th oo-, 33d + + +, 35th- -, 461)1 --o-
63d - -,65th -- o.

For cub-clone K, the reactions between line k and line k-a were
as follows: on the 3d day o +, 5th + o, 7th + , 8th + + + +,
9th - -, loth + + +, nth + + +, I4th - -, i6th

+ 0 + 0, i8th + +, 2ist + + + + +, 23d + +, 28th --oo
o, 30th + +, 35th + +, 46th o + +, 52d + +, 63d - -,
65th -

As can be seen by glancing at these data, the interactions were
variable and inconsistent. This suggests, perhaps, a fluctuating
change of the protoplasm in contradistinction to the uniform
conditions found to prevail in other experiments included in this'
section. No explanation of this is offered, since time did not per-
mit an exhaustive investigation of the subject to be made; but
it seems that one would hardly be justified in attributing the
cause to a detrimental environment, for, as has already been
dointed out, growth was not diminished by the presence ot
alcohol. Hegner ('19), found that Arcella dentata would grow
and reproduce in a medium containing from 0.25 to I per cent,
alcohol. However, the fission rate was retarded and irregulari-
ties in the shells of the offsprings were observed.

INTERACTION'S OF PROTOPLASMIC MASSES. 12.}

(g) One Line of a Clone Kept in Darkness, the Other Exposed

to Light.

Having funnel that related specimens could readily be made
negative to each other's protoplasmic fragments by adding
various substances to the culture medium in which one line was
grown, the following question arose: Are the changes thus brought
about due to the direct action of the added material, or to some other
causes? With this in mind, a series of experiments was begun
in which the same culture medium was used for both lines of a
clone, and in every other way they were treated alike, except one
line was placed in the dark while the other was kept before a
window. Thus, in one line the organisms were exposed to light
during the day, while light was entirely excluded from those
comprising the other, except for the short time required to trans-
fer them to new depression slides (a period of about two minutes
every other day). Four experiments of this type were conducted,
involving clones G, T, I , and V. In three out of the four cases,
the rate of reproduction was greater among the organisms kept
in daikness than among their relatives which were exposed to
light. However, it is possible that this was due to a greater
abundance of certain forms of bacteria upon which the Arccllce
fed; for it was observed that the bacterial growth was usually
very vigorous in the cultures which had been kept in the dark.

The results obtained in these experiments may be briefly given
as follows: In clone G, line g and line g-d continued to inter-fuse
with each other's protoplasm until the I5th day (i6th genera-
tion); in clone T. the first negative reaction was observed on the
I2th day (12th generation); in clone U, shattering reactions
occurred between line u and line u-d on the i^th day (i5th gen-
eration); in clone V, cross-fusion between line v and line v-d
ceased on the I3th day (i5th generation). After two related
lines had once become negative, no further cross-fn-ion- occurred.

In these experiments the average time required for two line-
of a clone to become so differentiated as to shatu -r each other'-
detached fragments of protoplasm was 13.25 day-, and 14.5
generations. While this is a longer period than was required
for the shattering reaction to be exhibited when sucrose, for
instance, was used as the differentiating factor, it i- s.;v^ d.iy-

124 BRUCE D. REYNOLDS.

(and 7.75 generations) shorter than was the case when no change
was made in the environments. It is thus shown that the
absence of light does affect, either directly or indirectly, the
physiological constitution of A. polypora.

(/;.) The Two Lines of a Clone Exposed to Different Temperatures.

Specimens were exposed for twenty-four hours to temperatures
of 36, 35, 30, 28, and 27 degrees centigrade, but in every case
death occurred during the period of incubation. However, if
incubated for only four hours, the organisms were able to resist a
temperature of 36 degrees C. Tests were made to determine if
such short exposures to heat would affect the protoplasm enough
for it to be detected by the cross-fusion method. No changes of
this sort could be revealed; regardless of whether the observa-
tions were made immediately after incubation, or twenty-four
hours later.

Attempts to culture A. polypora at low temperatures proved
to be more successful. Fifteen degrees centigrade was arbi-
trarily chosen as the low temperature environment, while room
temperature (21 degrees C.) was used as the norm. A lower
temperature was not selected because it was desirable that the
organisms reproduce at a fairly rapid rate; though at 15 degrees
C. the fission rate was reduced approximately one halt.

On Nov. 7 two temperature experiments were started, involv-
ing Clone X line x and line x-t, and clone Y lines y and y-/. The
same culture medium was used for both lines of a clone, and the
lines which were not kept in the low temperature incubator were
put in a darkened place to counteract any effect lack of light
might have had on the organisms in the refrigerator. As usual,
tests were made at frequent intervals to determine the interac-
tions of two diverging lines. The results obtained are shown
below by means of symbols.1 In clone X — reactions between
line x and line x-t on the 6th day + i + o, loth + i, I4th -
i i, i6th o o, I7th i - - , igth - - it.

In clone Y — reactions between line y and y-/ on the 6th day
+ , 8th + + +, iolh -f -f, i2th + + +, I4th +'+ +,
i6th + o +, I7th o -, igih - -.

+ signifies that fusion took place; •- shattering reaction; o contact was made
but neither fusion nor shattering followed ; £ contact was not made. The number of
symbols placed after a figure indicates the number of observations made on that day.

INTERACTIONS OF PROTOPLASMIC MASSES. 125

If the one shattering reaction observed on the 6th day of the
>eriment between the two lines of clone Y be disregarded, the
above d.it a will show that negative interactions first occurred <»n
the 141(1 day in clone X, and on the iyth day in clone Y. Taking
tin- hi^hi T figure 17 it is 4.58 earlier than the average time re-
quired when the environments in the two lines of a clone were
similar. However, if we consider the number of generations,
a greater difference is shown. On the day the first negative
reaction was observed in clone X, line x had produced 15 genera-
tions and line x-/ 8, which if added and the result divided by 2
yields 11.5 generations, while in clone Y, line y represented 18
generations and line y-/ 10, or an average of 14 generations,
in, if we take the higher number it gives a decrease of 8.25
generations. It seems then, that if two lines of a given clone
are subjected to different temperatures, they tend to i!i\<
physiologically more rapidly than when kept at the same tem-
perature.

SUMMARY OF KXPKRIMKNTS INVOLVING DIFFERENT KNVIRON-

MBNTS.

As previously shown, when two lines arising from daughter-
cells are subjected to similar environmental influences the
property of fusing with each other's detached fragments of pro-
toplasm ceases after approximately 21.58 days and 22.25 genera-
tions. By keeping one line in 10 per cent, hay infusion at room
trni| u-raturc in front of a window, such intra-clonal different ia-
tii'ii could be greatly hastened by altering (in only one resp«
t In- other, as follows:

Differonti.it in
Factor imposed. Days. I'-rations.

25 per cent.1 hay infusion used 8

i per cent, sucrose added 6 7

N/io saline added (unfavorable)

o.oi per cent, acetic acid added (killed niv-mi-in

0.025 sodium carbonate added 7 7

i per cent, alcohol added (fluctuated between + ami

-)

Kept in darkness i-Jo

Kept at 15 degrees C.! 1^.25

1 The other line kept in pond water instead of 10 per cent, hay infusion.

* The other line kept in darkened place to counteract darkness of refrigerator.

126 BRUCE D. REYNOLDS.

III. IDENTICAL ENVIRONMENTS.

When the experiments involving similar environments had
been completed, and before the experiments described in section
II had been performed, it seemed that the inability of an organism
to fuse with a protoplasmic fragment which had been severed
from a related individual was due to inherited physiological
variations. However, it was found, as shown in section II, that
by employing different environments the negative state was
reached in a much shorter time than was the case when the con-
ditions were unaltered. These results indicate either (i) that
the different environments have brought about heritable physio-
logical changes which influence the character of the reactions,
or (2) that the different environments have brought about tem-
porary physiological changes which do not persist after the organ-
isms are returned to identical surroundings. In other words:
the so-called "similar" environments may have, in reality, been
different. As Jennings ('16) pointed out, and Hargitt and Fray
('17) and Phillips ('22) have shown by feeding Paramecia on pure
cultures of bacteria, food is an important factor in bringing about
modifications. There seems to be no doubt that certain forms of
bacteria prevail in one concavity while in an adjoining concavity
the flora may be entirely different. The Arcellce in hay infusion
cultures subsist principally on the bacteria by which they are
surrounded ; therefore, one should not be surprised if physiological
differences are exhibited between organisms kept in different
containers.

Several methods might be employed in determining whether or
not the bacterial flora available to the organisms of two contrast-
ing lines are similar:

1. By making cultures of the medium in which they are living,
and determining the kinds and relative proportions of bacteria
present through the use of differential culture media. However,
since little is known about hay infusion bacteria, this would be a
difficult undertaking.

2. By isolating a pure strain of suitable bacteria and feeding
them to the organisms under aseptic conditions.

3. (a) By frequently transferring small quantities <>f tlu-
medium from one concavity to the other, and (b) by placing both

INTERACTIONS OF PROTOPLASMIC MASSES. 12J

organ! -in- in the same concavity and removing the new individuals
as they appear.

Method-, i and 2 were not undertaken, but both (a) and (6) of
method .> \\ere utilized.

Four clones, J, K. Q, and R were started and each of these was
divided into two lines upon the first division. Individuals from
these lines were transferred to new concavities every other day,
and upon each transferral of the organisms a quantity of the
culture medium from the old concavity in which one line had
been growing was placed in the new concavity in which a specimen
from the other line was to be confined, and vice versa. Thus, each
time an organism was placed into fresh culture medium it brought
with it some of the bacteria from the old culture in which it had
been living; but in addition to this, bacteria from the old culture
in which its sister line had been growing were also introduced, so
that any difference in the flora existing in the two old concavities
were distributed to the new concavities alike. Tests were made
I >e i \veen representatives of the two lines as had been done in
former experiments, and it was found that cross-fusion did not
cease after twenty or twenty-five generations, but persisted as
l»n^ as the experiments were continued; or in clone J until the
) jd generation, clone K until the 47th, clone Q until the 65th, and
clone K until the 5Nth. After obtaining cross-fusion in clones J
and K for 42 and 47 generations respectively, Petri dish cultures
wen- made of the two lines in each clone by putting four indi-
\idti.ils in a IVtri dish containing 25 c.c. of to per cent, hay
infusion. These were placed before a window when- they re-
in.lined undisturbed from July 31 until August 25.

Another series of experiments was begun in \\hich attempts
\\ere made to bring related specimens, that had been reacting
.itively towards each other's fragments, to a state in \\hii-h
cross-fusion would take place between them, 1>\ frequently ex-
changing small quantities of culture medium between the con-
cavities in which they were living. Three \ariatioii- of ihi-
system \vere employed: (a) by cro^-tran-ferring a small quan-
tity of culture medium from the old retvptarle "I "lie line to the
new container of the other each time the organi-m- \\ere con-
veyed to fresh culture medium— thus bringing their en\ in mments

128 BRUCE D. REYNOLDS.

to a common state; (6) by transferring medium only from one
line to the other — thus making the conditions in the latter like
those in the former; (c) the environments of both lines were
changed through the use of a different culture medium; these
were then brought to a common identity by exchanging small
quantities of culture medium between them.

Members from the Petri dish cultures of clones J and K were
used for this purpose. When tested on August 25 it was found
that the two lines of these clones reacted negatively towards each
other's fragments. By estimating the number of individuals in
each Petri dish, and knowing the number of days the cultures had
been running and their usual rate of reproduction, a fairly accu-
rate determination could be made of the number of generations
represented in each line. In every case the number used was
slightly under the estimated number. By following this pro-
cedure the figure obtained on August 25 were: for clone J, line
j-66 generations and line jd-6o; for clone K, line k-yo and line
kd 68. The same day three types of experiments were begun
in which both of these clones were used.

(a) Two small drops of culture media were interchanged be-
tween the old receptacles containing specimens from one line
and the new concavities into which a specimen from the other
line was to be placed each time the organisms were conveyed to
fresh cultures (every two days). At intervals of two or three
days observations were made of the reactions between individuals
from the two lines. In clone J the interactions of lines j and jd
continued to be of the shattering type for 18 days (until Septem-
ber 12), when cross-fusion was exhibited (line j-84 and line jd-
s<». l-'rnm this time until the experiments were discontinued on
November I ( j-124 and jd-i 19), individuals from one line would
readily fuse with fragments that had been severed from members
of a sister line. In clone K the interactions between lines k and
kd were of the shattering type for 20 days (until September 14).
From this time on, cross-fusion between the two lines was exhib-
ited, though several shattering reactions were also observed.
The experiment was discontinued on November i, at which time
line k had undergone 135, and line kd 132 divisions since clone K
was started on June 25.

In the following experiments attempts were made to bring

INTERACTIONS OF PROTOPLASMIC MASSES. IJ'l

lines jd and kd to the conditions of lines j and k respectively, by
conveying some of the culture media in which the latter had been
living to the concavities containing the former. Under this
treatment individuals from lines j and jd reacted negatively
towards each other's fragments during the succeeding fifteen
generations (until September 9). After this cross-fusions were
the usual occurrences, though occasional shattering reactions were
encountered. When the experiment was discontinued on Novem-
ber i line j represented 129 generations and line jd 120 genera-
tions since the beginning of clone J. The reactions between
lines k and kd continued to be of the shattering type until
September 16 (for 24 generations). From then until the experi-
ment was discontinued on November I (line k-1^5 anc^ h'ne
kd-133) most of the interactions were positive.

(c) In these experiments specimens from clone J, lines j and jd
and clone K, lines k and kd were taken from the Petri dish cul-
tures and placed in concavities containing 10 per cent, hay
infusion plus i per cent, sucrose; but after remaining in this
solution for six days, the organisms had decreased so much in
si/e that they were put back in standard medium. Hvery other
day small quantities of culture media were exchanged between
the receptacles in which sister lines were kept. The organisms
subjected to i per cent, sucrose not only decreased in size, but
their physiological states were so altered that they no longer
fused with fragments from closely related individuals which had
been kept in standard culture medium — see section II, (6). Thus,
it is evident that an environment different from what existed
before had been established.

Upon testing by the usual method, it was found that the
shattering type of reactions persisted between specimens of line
j and line jd for 18 days and 19 generations, but from then on,
cross-fusion was exhibited. The experiment was discontinued
on Nov, 2d, at which time line j represented 125 generations
and line jd 122. In clone K shattering reactions were ob>er\r. 1
between line k and line kd until the 23d d.iy ,m<l i;tli o -aera-
tion after the experiment vvas started. Later ol»ei -\ .itions
yielded a much larger proportion of shattering re.u ti<m> than
were encountered in any of the other experiment-, but a mmiln-r
of cross-fusions did take place between the two lines. When the

I3O BRUCE D. REYNOLDS.

experiment was stopped on Nov. 1st. line k had undergone 119
divisions and line kd 122.

The results obtained in the above experiments indicate that
the changes produced in the protoplasm of related individuals
reproducing vegetatively, are at least greatly influenced by the
environment — in fact to such an extent that two lines which have
been negative to each other's protoplasmic fragments for several
weeks can be brought back to a common condition in which cross-
fusion will take place between them within approximately 20
days. However convincing this evidence seems, it was con-
sidered advisable to obtain further proof. To that end, a series
of experiments was planned which wouki extend the observa-
tions over a longer period, and at the same time remove any
question as to the identity of the environments in which the two
lines of a clone were existing. It is a matter of common knowl-
edge, that when Arcella divides the newly formed shell is lighter
in color than the old shell, and that this difference in shell colora-
tion continues for some time after division takes place. There-
fore an individual possessing a dark shell was taken from each, of
the two diverging lines of a given clone and placed in the same
concavity. Then daily examinations were made, and when two
specimens with light-colored shells appeared, they were removed
and tested for inter-protoplasmic reactions; while the two dark-
shelled organisms were transferred to fresh culture medium. It
might be said in this connection, that the morphological differ-
ences between specimens from the two lines, were usually also
great enough to enable a careful observer to distinguish one from
the other.

On November I, 1922, four experiments of this type were
started, involving clone J, line j-124 and line jd-ii9; clone K,
line k-135 and line kd-132; clone Q, line q-65 and line qd-65;
and clone R, line 1-58 and line rd-58. The two lines of each of
these clones had previously been brought to a common condition
1>\ exchanging small quantities of culture media from one to the
other, as described elsewhere; so that individuals from one line
would fuse with fragments of members of the other at the begin-
ning of these experiments. So then, it was to be expected that
the immediate progeny of two such organisms would also exhibit
the cross-fusion phenomenon.

INTERACTIONS OF PROTOPLASMIC MA.-SES. 1 U

Individuals from the two lines of these clones were confined in
the same concavities for ten days. During this time frequent
tests showed cross-fusion in every instance: consequently on
\o\emlier 10 the related lines were separated again. by placing
the individuals in different Petri dishes containing 25 cc. of 10
PIT 1 1 nt. hay infusion. At intervals of two week> the number ot
indi\ idti.ils in the Petri dish cultures were estimated, and the
approximate number of generations represented determined.
Then fresh Petri dish cultures were made, and thus the lines were
perpetuated until January 5, 1923. l On January 5 tests were
made for cross-fusion between the two lines of these clones, with
the result that all of them gave the shattering reaction. Then an
individual from line j was put in a concavity with an individual
from line jd. In a like manner the two lines of the other clones
were brought together. During the next five days the inter-
protoplasmic reactions were universally negative, while on the
sixth day they were all positive. Cross-fusion between the re-
lated lines continued until the experiments were stopped on
January 15. At that time clone J had been observed for 188
generations, clone K for 200, clone Q for 126, and clone R for 122.

These results present a strong argument in favor of environ-
ments as the principal modifying agent; but the skeptic miijit
say that perhaps the two lines were confused while being kept
in the same container, and what seemed to be cross-fusion be-
tween very distantly related individuals was really positive inter-
actions of very closely related specimens. That such was i he
case is very improbable; for in the first place, especial care was
taken tc^avoid that error, and secondly, the results obtained were
too uniform. The same mistake would hardly have been made
in every experiment at the same time. However, the most con-
\incing evidence obtained was made po— il>le by the Midden
appearance of a double form in a Petri di>h culture <>1 rl«me O,
line qd. This abnormal specimen was di-< -«\ ered <>n \n\rinU T
21, 1922, and seems to have appeared l>e..uiM- i.t" the failure of
the shells to completely separate at the time of ii--.it m. It is

1 An objection might be raised to mass cultun-s mi tln-gri>mid- that i"ii

illicit take place between some of the organism-, thus disturbing the germinal

.cncy of the individuals concerned. ThU ]>• .— ibiliiy

linn and guarded against, and it can be said with .1 :i ili>l nut

place among the specimens used in this invi--ti.iMtiiin.

132 BRUCE D. REYXOLDS.

shown contrasted with a normal individual in Fig. 4. (See
Reynolds '23.)

As can readily be seen, the morphological differences between
this abnormal specimen and a normal individual were quite
marked; consequently they could be kept in the same concavity
without any danger of confusion. The abnormal form could be
reared, and it was found to breed true to type. By January 12
34 generations had been obtained from the abnormal Arcella,
while the normal individuals of the line from which it sprung had
given rise to 37 generations. During this time, a period of 52
days, the normal and abnormal organisms had been kept in
different receptacles, and no exchange of medium between them
had been made. In accordance with other findings, therefore,
they should have reacted negatively towards each other's frag-
ments of protoplasm — even though both had possessed the same
morphological characteristics. In testing this out on January 12
it was found that they exhibited a violent shattering reaction.
On the same day an abnormal specimen Ab-34 was Put m the
same concavity with a normal organism qd-123, both of which
belonged to the same line (qd) of clone Q. By making daily
tests "• it was found that the interactions continued to be negative
for three days. From the fourth day on, cross-fusion took place
between them readily.

Another experiment involving the abnormal form was also
started on January 12. In this case a normal individual j-i85,
which was descended from a specimen taken from a wild culture
on June 26, IQ22,2 was placed in a concavity with an abnormal
organism Ab~34, after it had been demonstrated experimentally
that one would not fuse with fragments of the other's protoplasm.
Then the usual procedure of daily testing the interactions be-
tween the progeny of these two organisms was carefully practiced.
During the first five days the involved protoplasms were shat-

1 In this and the following experiment all amputations, and subsequent obser-
vations of protoplasmic reactions, were made without removing the organisms from
their containers.

2 While the two organisms used in this experiment belonged to the same species,
and probably at some time in the past had a common ancestor (since their progeni-
tors were isolated from the same wild culture); their relationship was certainly re-
mote, for the two clones to which they belonged had been under observation for
nearly seven months.

INTERACTIONS OF PROTOPLASMIC MASSES.

tered to pieces upon making contact, after which the animal
\\<>uld move off, leaving the remnants of its own and the frag-
ment's protoplasm behind. On the 6th, 7th, and 8th days con-
tact was followed by a slight shattering of protoplasm, and later
some or all of the remnants were appropriated by tlu- reacting
nism. On the 9th, roth, nth, and I2th days of coexistence
(/DC organism would fuse with a fragment of the other practically,
if not quite, as readily as with one of its own fragment >. No
Inn her observations were made after the twelfth day.

SUMMARY or Kxr ERIMENTS INVOLVING IDENTICAL ENVIRON-
MENTS.

By continually interchanging small quantities of media from
old cultures of one line of a given clone to new cultures of the
oilier line, cross-fusion between the two lines can be prolonged
for sixty-five generations (perhaps indefinitely?). After two
lines of a clone have ceased to fuse with each other's fragment-,
the property of cross-fusion can be restored by: (a) frequently
exchanging small quantities of media between the cultures — thus
bringing them to a common state; (6) frequently transferring
small quantities of medium from only one culture to the other—
thus bringing one line to the condition of the other; (c) creating
a different environment in both lines by adding I percent. MU i
to the culture medium and then exchanging media between them
frequently. Quicker and more satisfactory results can be ob-
tained, if members of the two lines to be tested are rultivated in
the same container. By this method, two lines of a clone that
have been negative to each other's protoplasm for month-, can
be brought back to a state of cross-fusibility — even though they
are removed from the clonal parent by as many a^ Jo<> genera-
tions. A normal specimen was put in the same concavit) with
an abnormal relative. Though reacting negatively in the be-
ginning, they readily fused with each other'- ents <>t proto-
plasm after living together for four day-. An unrelated (?) nor-
mal specimen was placed in the same container with an al>n< .rmal
Arcella. In this case the shattering reaction \\.(- exhibited until
the ninth day of coexistence ; from then on, tu-ioii occurred readily
between one organism and pseudopodial fragment- of the other

134 BRUCE D. REYNOLDS.

DISCUSSION.

Before entering into a discussion of the results obtained in
these investigations, it is perhaps advisable to show that the
method employed is sound, and that it offers some advantages
over the methods which have been utilized by others. As stated
in the introduction, this is the first time that the interactions of
protoplasmic masses have been used in studying the subject of
Heredity and Environment. It was not until the recent appear-
ance of the first article of this series that anything was published
concerning the fact that certain of the shelled rhizopods are
capable of reappropriating, by fusion, fragments of protoplasm
which have recently been detached from their cell-bodies. In
this paper it is clearly pointed put that "the severed fragments
are not recovered as food, but enter again immediately into the
protoplasmic structure of the cell-body." Later work has
strengthened the evidence for such a contention. In fact, the
findings in these experiments with A. polypora substantiate all
of the conclusions drawn from the work on Difflngia, except para-
graphs 3 and 5 of the general summary.

In regard to paragraph 3: In Arcella, fusion has been observed
to occur at the ends of pseudopods, as well as along an extended
mid-region. This difference in behavior of the two genera may
be due to the difference in the nature of their pseudopodial forma-
tion. As to paragraph 5 : In six cases out of approximately 1,000
two enucleated fragments have been observed to fuse with each
other; though in such instances they were severed from the same
cell-body at the same time, and were almost in contact in the
beginning.

In addition to the data on protoplasmic fusion given in Article
No. i : the "shattering" reaction is described in the introduction
of the present paper. This is the kind of reaction usually ob-
served when an organism made contact with a fragment severed
from a distantly related individual. It was found that in such
cases there was a mutual attraction between the organism and
the fragment — even though their protoplasms were shattered
upon making contact. If the fragment came from a closely
related specimen fusion occurred. When a fragment of proto-
plasm was severed from a different genus (Dijflngia) and placed

INTERACTIONS OF PROTOPLASMIC MASSES. I x>

near an Arcella the two bodies did not attract eachother. This
\\.is done on ten different occasions; each time the fragment w.i-
placed within loomicra of the organism, but neither body -ho\\v<l
any response to the other's presence. Chance wandering -h<>uld
have brought the two masses together in a certain percentage of
es. This did happen on two occasions out of the ten tried,
but nothing resulted from the contact. In both instances the
animal moved on as if the fragment had been merely a mechanical
obstruction.

Another observation which is probably worthy of mention in
this connection, is that under ordinary circumstances two A rcella-,
\\hich would re-act negatively towards each other's fragments,
may come in contact with each other's pseudopods without
causing any disturbance; but when a protoplasmic fragment of
one of these organisms is in their midst, it was frequently observed
that their pseudopods would be shattered upon making contact.

With these various facts established, it seems evident that the
inability of an organism to fuse with a pseudopodial fragment is
due to a difference in the physiological constitution of the pro-
toplasm involved. It is generally considered that morphological
dillerences are merely physical expressions of physiological
changes — a kind of cause and effect relationship. If this be true,
then physiological tests offer finer distinctions between minute
variations than can be jxrceived in the structure of organisms.
Furthermore, as shown by the experiments in which one line of a
cl.me was placed in a culture medium containing I per cent, su-

-e, physiological changes may persist for a long time ai
morphological differences have disappeared.

Most investigators in the past have, of necessity, studied
\.iriations among protozoa by making use of structural differ-
ences. While these arc undoubtedly of great importance, they
d<> not offer as delicate a method of determining variation-, or as
-ure a means of detecting them. Anyone who ha- oli-i-r\ed a
"-hattering" reaction can but be impressed with tin- fa< t that a
marked difference exists between the reacting ma-ses — a few
generations back they were parts of the -ame individual; t hex-
have diverged by means of binary ti--ion until they not only fail
to unite when brought together, but are shattered into in. my
pieces. Why does this happen? We can postulate many pos>il>le.

136 BRUCE D. REYNOLDS.

explanations. It seems most reasonable to think of it as being
due to a difference in molecular structure, or ionization; at least
the sudden violent shattering is suggestive of an electrical phen-
omenon. The inability of one organism to fuse with a proto-
plasmic fragment of another, does not imply that the two organ-
isms concerned differ structurally, nor does cross-fusion indicate
that no morphological differences exist. In either case the
animals involved may or may not be identical in size, structure
and rate of fission.

Maupas ('88, '89), in studying the effect of temperature on the
fission rate of infusoria, arrived at the conclusion that under
given conditions the fission rate remains the same, and that
inherited variations do not exist. Regardless of the important
bearing this statement has on the theory of evolution, there was
but very little interest exhibited until some time after the publi-
cation of Johannsen's ('03) results, in which he enunciated his
concept of genotypes. Since then workers in various fields have
investigated the subject of Heredity and Environment. Like
Maupas, most of them have obtained evidence which seems to
emphasize the "fixity" of species. Among these might be
mentioned: Johannsen ('09, 'u) working with beans, Hanel
('08) and Lashley ('15) with Hydra, Jennings ('08, '09, '10) with
infusoria, Ackert ('16) with Paramecium, and Jollos ('21) with
infusoria. However, recently Barber ('07) working with bac-
teria, Jennings ('16) with Difflugia, Hegner ('19) with Arcella,
Root ('18) with Centropyxis, Middleton ('15, '18) with Stylon-
ychia, Erdmann ('20) with Paramecium, and others, have appar-
ently been able to demonstrate that variations do take place
among organisms reproducing asexually — and as Erdmann ('20)
says: 'The rigid conception of the genotype does not hold true
for protozoa."

To enter a field filled with so much conflicting evidence, armed
with a new method, was an especially attractive undertaking.
The experiments were planned with the idea of duplicating, so
far as possible, the various conditions to which protozoa have
been subjected by others. The only new factors employed were
the method of determining variations which might arise, and a
little different way of preparing the culture medium.

The results obtained seem to combine the two diverse opin-

INTERACTIONS OF PROTOPLASMIC MASSES. l.>7

ions, and thus answer some of the hitherto perplexing questions.
In the first place, they show that both physiological and morpho-
logical variations do appear among protozoa reproducing \ege-
t, nively, and that such variations can be hastened by alurin-
ihr environments; and secondly, that these changes are not due
to inherited variations which persist for very long periods of time.
for they could be reversed within a few days. Unquestionably.
certain heritable variations often arise in cultures, and seem to
persist for some time afterwards; but, in our experiments at
least, if the organisms in which these changes are manifested are
left in the same surroundings, undisturbed, they tend to revert
to the normal type, while if placed in a different environment
where selection would be affective, or if artificial selection is
practiced, the aberrant characteristics may be continued, or even
emphasized.

In a similar manner the physiological status of organisms is
subject to modifications. Adjustments are constantly being
made as a result of the ever changing surroundings. The ques-
tion seems to resolve itself into one of environmental factors, of
which food is by no means the least important. The only diffi-
cult factor to control in these experiments was the bacU-ri.il
growth in the culture media. In fact, the rapid differentiation
observed when unlike culture media were used, was probably due
to a change in the character of the bacterial flora produced by the
action of the environment; for if the action had been direct, in
some cases at least, physiological differences should have occurred
\\ iihin a few hours after the organisms were exposed to it.

SUMMARY AND CONCLUSIONS.

i. Like Dijjlugia, Arcclla polypora will, under favorable con-
ditions, reappropriate its detached fragments of protopla-m by
lu-ion.

j. .1. polypora is not attracted by a pseudopodial fragment of

'Jiigiu. If accidental contact is made the fra^im-m
as a mechanical obstruction.

,v Fusion will take place between one individual and a proto-
plasmic fragment of a closely related -]>t •> -mien.

4. Two distantly related individuals, which ha\e not been
kept in the same receptacle, will be attracted by each other'-

1 V"1 BRUCE D. KKVXOLDS.

severed pseudopods, but upon making contact the involved pro-
toplasms will be shattered into bead-like masses.

5. When two lines of a clone are kept under similar environ-
ments, cross-fusions between them cease after approximately 22
days; while if the two diverging lines are subjected to different
environments, the time required for cross-fusions to cease is
greatly shortened (varying from 6 to 16.5 days depending on the
factor employed).

6. When small quantities of culture media are frequently
exchanged between the concavities in which members of two
diverging lines are confined, cross-fusions between them appar-
ently continue indefinitely

7. After two related specimens have become negative to each
other's protoplasm, cross-fusions between their progeny may be
induced by (a) exchanging small quantities of culture media
between the receptacles in which they are kept (time required was
20 days), or (6) by placing both of them in the same concavity
(time required was 6 days).

8. Morphological differences are not necessarily associated
with different physiological states, as evidenced by the fact that
abnormal specimens could be induced to fuse with fragments of
normal individuals — and vice versa.

9. Physiological changes do occur among the descendants of a
single A. polypora reproducing vegetatively, but such differences
are probably due to environmental influences rather than to
hereditary variations.

BIBLIOGRAPHY.
Ackert, J.

'16 On the Effect of Selection in Paramecium. Genetics, Vol. 1.387-405.
Barber, M. A.

'07 On Heredity in Certain Microorganisms. Kansas University Science, Bui.

4. 3-47-
Calkins, G. N.

'n Regeneration and Cell Division in Uronychia. Jour. Exp. Zool., Vol. 10,

95-116.
Calkins, G. N., and Gregory, L. H.

'13 Variations in the Progeny of a Single Ex-conjugant of Paramecium canda-

Intn. Jour. Exp. Zool., Vol. 15, 467-525.
Castle, W. E.

'14 Pure Lines and Selection. Jour. Heredity, Vol. 5, 93-97.
Erdmann, Rh.

'20 Endomixis and Size Variations in Pure-bred Lines of Paramecium aitrelia.
Arch. f. Entwicklungsmech., Bd. 46, 85-146.

INTERACTIONS OF PROTOPLASMIC MASSES. 139

Hanel, Elise.

'08 Vererbung bei ungeschlechtlicher Fortpflanzung von Hydra grisfa. Jrn-

aische Zeitschr., Bd. 43, 321-372.
Hargitt, G. T. and Fray, W. W.

'17 The Growth of Paramecium in pure Cultures of Bacteria. Jour. Exp.

Zool.. Vol. 22. 421-455.
Hegner, R. W.

'19 Heredity, Variation and the Appearance of Diversities during the Vegeta-
tive Reproduction of Artella denlata. Genetics, Vol. 4. 95-150.
'198 The Effects of Environmental Factors upon the Heritable Characteristics
of Arcella dentata and .1. Polypora. Jour. Exp. Zool.. Vol. 29: 427-441.
'20 The Relations between Nuclear Number, Chromatin Mass, Cytoplasmic
Mass, and Shell Characteristics in Four Species of the Genus Arcella.
Jour. Exp. Zool.. Vol. 30. 1-95.
Jennings, H. S.

'08 Heredity. Variation and Evolution in Protozoa. II. Heredity and Varia-
tion in size and form in Paramtcium. with Studies of Growth, Environ-
mental Action and Selection. Proc. Amer. Philos. Soc.. Vol. 47. 393-
546.

'16 Heredity. Variation and the Results of Selection in the L'niparentnl Repro-
duction of nijfliiRia corona. Genetics. Vol. I. 407-534.
Johannsen, W.

'03 L'eber Erblichkeit in Populationen und in reinen IJnien. v -)- 68 pp.

Jena: Gustav Fischer.
Jollos, Victor.

'21 Experimeiitelle Protistenstudien. Arch, f. Protistenk., Bd. 43, 1-222.
Kepner, W. A., and Reynolds, B. D.
'23 Reactions of Cell-bodies and Psoudopodial Fragments of Di fflttgia. BIOL.

B< I... Vol. 44. 22 47.
Lashley. K. S.

'15 Inheritance in the Asexual Reproduction of Hydra. Jour. Exp. Zool..

Vol. 19, 157-210.
Leidy, Joseph.

'79 Fresh-water Rhizopods of North America. Report U. S. Geol. Survey of
the Territories, Vol. XII. xi -f- 324 pp. Washington: Government Print ini;
Office.
Middleton, A. B.

"15 Heritable Variations and the Result of Selection in the Fission rate of

Stylonychia pustulata. Jour. Exp. Zool., Vol. 19, 451-502.
'18 Heritable effects of Temperature Differences in the Fission rate of Slylony-

thia piistnlata. Genetics, Vol. 3, 535-572.
Penard, E.

'02 Faune Rhizopodique du Bassin du I-cman. 714 pp. Geneva: \V. Kuii'liu

et fils.
Phillips, R. L.

'22 The Growth of Paramecium in Infusions of Known '-nt.

Jour. Exp. Zool.. Vol. 36, 135-183.
Popoflf, M. .

'08 Experimented Zellstudien I. Arch. f. Zellforsch., IM i. 245- ^79.
'09 II.. Bd 3, 125-178.

140 BRUCE D. REYXOLDS.

Reynolds, Bruce D.

'23 Inheritance of Double Characteristics in Arcella Polypora Penard; Genet-
ics. Vol. 8, 477-493-
Root, F. M.

'18 Inheritance in the Asexual Reproduction of Centropyxis aculeata. Genetics,

Vol. 3, 174-206.
Stout, A. B.

"15 The Establishment of Varieties in Coleus by the Selection of Somatic Varia-
tions. Carnegie Institution of Washington, Pub. No. 218: +80 pp.
Whitney, D. D.

'16 The Transformation of Brachionits pala into B. amphiceros by Sodium
vSilicate. BIOL. BUL., Vol. 31, 1x3-120.

142 BRUCE D. REYNOLDS.

EXPLANATION OF PLATE I.

FIG. i. Camera lucida sketches illustrating shattering reaction. A, Arcella
poly fora approaching a fragment from another individual of the same species, frag-
ment shown below. X i~o. B, The same bodies just after making contact. C,
A portion of B enlarged.

FIG. 2. Diagram of vertical section illustrating shape of shell and position of
pseudopods.

FIG. 4. An abnormal Arcella which appeared suddenly in a hay infusion culture
of clone Q, line qd. .4, Lateral view; B, Ventral aspect, showing ellipsoidal mouth;
C, normal individual of line from which the abnormal specimen was obtained. X
170 — camera lucida.

BIOLOGICAL BULLETIN, VOL. XLVI.

PLATE i.

£

<*/*

Fig-1

Fig. 2

c

Pig. A.

BRUCE D. REYNOLDS.

THE ATTACHMENT OF OYSTER LAR\VE.

•

THfRLOW C. XELSOX.1

Knowledge of the life history and ecology of the American
oyster, Ostrea virginica Gmelin, is more extensive than in the case
of any other species of lamellibranch. The researches of num-
erous investigators, notably of Brooks, J. Nelson, and Stafford,
give us an almost unbroken history of this mollusc from the egg
to the* adult. As in most marine bivalves, the eggs of the oyster
are shed into tin- water where fertilization occurs. Subsequent
development is rapid, the larva forming a bivalve shell within
two days or less. Then there follows a pelagic period of approxi-
mately two weeks during which the larva swims about with the
aid of the velum, and increases considerably in si/e. At the
close of the free-swimming period the larva settles upon soim-
solid object and remains firmly attached at that point for the
remainder of its life, t'p to the time of attachment the history
of most other marine lamellibranchs is similar to that of the
oyster, but once the free-swimming period is ended we find u
variety of behavior; the larva- of Pecten, Modiolus, and MytUns
attach to seaweeds or to other objects, the larva- of the Tered mi-
da? and of the I'holads bore within a solid substratum, while those
of ]'enns and ,\fya burrow into the bottom.

Concerning the actual attachment of the oyster larva at the
close of its pelagic life we have little information, the nature of
the process having been deduced in the main from a study of the
stages just preceding and just following fixation. Ryder, '82,
Huxley, '83, Jackson, '90 and others, including the writer (Nel-
son, '21 A), believed that the secreting border of the mantle was
used to cement the larva fast to the substratum. Stafford, '13,
on the other hand, found in his youngest oyster "spat" that the
left valve bore a layer about five times as thick as the >hrll it-til,
composed apparently of a different material. This thick l.iyrr

'From the Zoological Laboratory of Rutgers College ami th. I >• partment of
Biology, Xew Jersey Agricultural Experiment Station. Paper X<>. 1 1 7 of the
Journal Series, New Jersey Agricultural Experiment Suiti"ii-, 1 >' partment of
Biology. This paper \vill appear in Rutgers College Studies Vol. ->.

143

144 TIIURLOW C. NELSON.

covers more than one half of the outer surface of the left valve,
extending from close to the anterior ventral edge almost to the
umbo, but falling short posteriorly. From his study of the
byssus gland in the base of the foot of the larval oyster, Stafford
concluded that this organ must furnish the substance used to
attach the larva, since the large amount of this material and its
position far under the left valve preclude the use of the mantle
for this purpose.

It has been my good fortune to observe the actual attachment
of oyster larvae under semi-natural conditions, and thus to be
able to fill the gap in the information about this process. -Large
numbers of full grown oyster larvae were found at 8 P.M., July
23, in the water surrounding our laboratory houseboat on Barne-
gat Bay. A glass plate measuring 3x4 inches was lowered ver-
tically into the water and removed half an hour later, and imme-
diately suspended horizontally in a dish of sea water under the
binocular.1

Six oyster larvae were moving about over the glass, holding
on by means of the very active and highly adhesive ciliated foot.
Some were on the upper, others on the lower side of the glass,
thus permitting observation of the attachment process both from
above and from below. With the velum withdrawn, and slowly

B

FIG. i. Illustration of the method of creeping of the oyster larva over a glass
surface. -4, larva with foot fully extended; B, larva with foot contracted and
body pulled forward. Larvae shown from the right side.

rolling the shell from side to side, the larvae extended the foot to a
distance of about 0.25 mm., attached the distal end and then a
contraction of the foot pulled the body forward. The "heel"
of the foot next being attached, the body was swung part way
around; the tip of the foot was again extended in the new direc-
tion and the process repeated (Fig. i). When first observed at

1 Jackson, '88, J. Nelson, '08, and Stafford, '10, used this method of obtaining
oyster spat; but no one of these authors records having witnessed the actual attach-
ment of the larva.

THE ATTACHMENT OF OYSTER LARY/E. 145

8.35 P.M. the larvae were describing circles a little over i mm. in
diameter. As they continued moving around the radius of the
circles was gradually diminished, until at 8.46 the first larva
came to rest With the foot extended to about one half its full
length, its distal end flattened in contact with the glass, the
larva swung the body until the foot occupied the median posi-
tion, with the left valve against the glass and inclined to it at
about 30 degrees (Fig. 2). In this position the ventral edge of the
It It valve almost touches the substratum.
The ventral border of the mantle was ex- /^Z^*vr |v

tended until it came in contact with the
glass \\here it remained for 2 minutes, and
then was \\ithtlrawn. The foot, with its
(ilia beating very feebly, was then slowly
drawn in, and II minutes after the larva
ceased circling the f(KJt had entirely dis-
appeared between the valves, and fixation

FIG. 2. The PC

was accomplished. The method of attach- taken by an oyster
merit of the larva? on the upper and on the larva du»"K fixation,
lower side of the glass was the same; it

/, foot; /. r., umbo of
illustrates, in the case of those larva' attach- Jcft Vai\-c; m. mantle.

ing from below, the power of the foot in

holding the left valve firmly against the substratum while fixation

is being effected.

The great importance of the foot to the ful n oyster larva

was recognixed by Stafford '10, who described it- • n t pin.^ move-
ments. The manner in which the larva cement- it -elf fast was
deduced by him ('13, p. 65) solely from hi-toloijr.il evidence.
He concluded that the mantle could not furni-h material in
sufficient quantity nor rapidly enough t" IT of u-e .1- the organ of
fixation. As he says: 'The usefulness of the foot .is an organ
of locomotion, as a clinging organ, as an or^.m of hvition, had
appealed to me for some time but I li.nl no dim t » -\ idence to
support the view that it was really the ot;<an of final attach-
ment."

Stafford found that a large part of the ki-e of the foot of a
lull-grown oyster larva is occupied by the byssus gland which
consists of three lobes communicating with a duct which opens

I46

THURLOW C. NELSON.

to the outside at the heel of the foot. The cells of this gland are
gorged with a transparent secretion which is not found in the
much shrunken gland which characterizes the larva immediately
after attachment. He believed that fixation was accomplished
l>y the larva extending the foot until the heel came to a position
near the upper anterior edge of the valve, and that the byssus
gland then discharged its secretion, which, flowing between the
left valve and the substratum, soon hardened and held the larva
fast. Stafford believed that the mantle played no part in the
fixation process.

My own observations show that Stafford was right in consid-
ering the foot as the organ of final attachment. It is not thrust
forward, however, as he held, but is brought to the median posi-
tion. The extrusion of the mantle for a short period evidently
aids in the quick and economical distribution of the cementing

fluid as it is poured out of the byssus gland at the ventral edge
of the left valve. This secretion hardens in less than 10 min-
utes. In Fig. 3 are shown a number of newly attached larvae

I'll',. 3. Oyster larvae removed just after attachment and photographed from
the left side to show the area covered by the cementing substance from the byssus
gland. Magnified about 40 diameters.

photographed from the left side to show the area covered by, and
the distribution of, (he cementing material.

The use of the foot in creeping over surfaces has been described
in a number of larval bivalves, and it is the chief means by which
the young molluscs obtain foothold in a favorable locality.

THE ATTACHMENT OF OYSTER LARV.E. 147

KVIlo-g, '99, showed the importance of the foot in aiding the
larva <>f Mya to obtain attachment. Sigerfoos, '07, noted that
larval shipworms possess a powerful foot by means of which they

••p rapidly over surfaces and select spots favorable for fixation
to the exclusion of those which are not suitable. Field, '09,

•es that the larvae of .}fytilits creep from unfavorable situa-
tions upon seaweeds to locations which are better adapted to
them. Belding, '10, describes the foot of the larval Pecten, a
most important organ, since it is used for swimming as well as in
crawling. Brlding, '12, showed that in the full-grown larva of
Venus the foot likewise is used in swimming as well as for crawling,
and for procuring attachment.

Tli.it the use of the foot enables oyster larvae to exercise sonu-
M-K-ction in seeking a place of attachment is evident from the
\\idr <lillrrenres in the number of spat procured when \.uioii-
kinds of cultch are placed together in the water. I h.i\ e n •• cntly
shown (Nelson, '23,) that oyster larva; will not attach to >ln-!U
\\hich are extensively pitted by the boring sponge, Clione, or
\\hiili are badly corroded and which present suit.ni-, ih.it an-
nit i« iscopically rough.

For some hours before oyster larva.- set they may In- oli-rrved
iiio\ing about over various surfaces with the \rluin e\ten<led, its
cilia beating actively; and with the foot protruded .unl ln-nt at
right angles, thus bringing the distal half in contact with the
Mil 'stratum. The foot remains quiescent in this portion while
the larva slowly "skates" along, as it \\i-n-, pn-prllrd by the
velum, but aided somewhat by the cili.t on the looi. Thru,
swimming upwards, the larva progresses a short dist.inci- thnm-li
the water and drops down to the Milisir.itnm to coniiniu- as
before. In this manner the larva m.iy w.nult -r over .^n .q»pr»'ci-
able area before it eventually attaches i Nelson, '21 \

In this connection it is well to correct tin- idr.t ln-l<l \>\ »\ -ter
growers and by some scientists also, that attachment of tin-
oyster larva occurs when its shell becomes so he.ivy that the
animal sinks to the bottom, unable longer to -\\ im. 1'or e\ani|)U-,
Stafford ('13, 34) says: "when presumably the lai\.r bcconu-
too hea\y to swim with ease, settle towards the bottom, creep
about, and select some clean solid surface upon which they fix
themselves.1

148 THURLOW C. NELSON.

That such is not the case must be evident to anyone who has
observed extensive oyster sets on the bottoms of boats, or upon
the bottoms of floats kept at the very surface of the water
(Nelson, '17,). The full-grown oyster larva, so far from being
helpless, is capable of more powerful swimming than at any time
during its pelagic life. During the course of the development
and growth of the larva the increase in size of the velum is more
than commensurate with the increase in the weight of the shell.
Plankton samples taken during the calm of early morning show
large numbers of full-grown oyster larvae swimming at the sur-
face. At times the larva may project the foot into the surface
film and, withdrawing the velum, may hang suspended by the
foot and slowly rock the shell from side to side, much as in the
familiar habit of pond snails hanging from the surface film.

Full-grown oyster larvae are found mostly at the bottom be-
cause they develop during the last two days of their pelagic life a
strong positive stereotropism, and since fixed objects are most
abundant on the bottom the larvae mainly congregate there.
The intensity of attachment which occurs at the very surface of
the water when suitable cultch is available, precludes the idea
that increased weight of the shell or even positive geotropism are
factors exclusively governing the time of fixation.

The use of the foot by oyster larvae in crawling over surfaces is
similar to that reported by Crozier, '21, who found that the adult
Lima oriented itself away from the light by attaching the tip of
the foot to the substratum after bending the free portion away
from the source of illumination. The creeping of oyster larvae
is also strikingly similar to that observed in the fresh-water
Sphacridae, and especially in the young of Corneocyclas when first
liberated from the brood pouches of the parent (Nelson, '2I-B).

•

In sessile forms, such as the oyster, which remain fixed on the
spot where the larvae set, keen competition follows when large
numbers settle at one time. During heavy sets 1,000 larvae may
attach to one oyster shell 8 cm. long, upon an area which will
permit not more than 6 oysters to grow to brtrding age. It is
significant that at such times of abundant settlement, when the
water over the oyster beds may show as high as 250 full-grown
larva; per liter, the frequency of attachment per unit area ol"
available cultch does not exceed a certain maximum.

THK ATTACHMENT OF OYSTER LARV.-E. 141)

I-'ig. 4 >how- in a circle 6 mm. in diameter a group of 40 o\
larva- ju>t attached to an experimental shell. Unfortunately I
have never witnessed the behavior of two oyster larvae coming
intu contact with each other while circling about over a surface,
Inn an examination of shells taken at times of maximum settle-
ment indicates that the movements of the larva? prior to final
fixation not only aid them in finding a spot suitable for attach-
ment, Inn that this behavior also tends to keep the larvae sep-
arated from one another. The proximity of a neighboring larva
could presumably be recognized only through touch, hence it is
conceivable that here and there individuals failing to come in
contact while circling about might attach close to one another,
.1- ci « urred in six instances in Fig. 4. Kxamination of the nn>-t

FIG. 4. Newly attached oyster larvae on the - !i< -11. Within

tliis circle, 6 mm. in diameter, are 40 larva?. Mai;iiiti<-.| about i 5

heavily set shells in our collection -lx>\\- «.in ».| v>o newly at-
tached larva1 only 57, or 11.4 per cent., \\liicli are within .5 mm.
of another larva. This distance is taken since ii repiv-ents the
approximate radius of the circles described l>v the larva- when
observed. Studies of the behavior of other bival\e larva?

I5O T1IURLOW C. XELSOX.

during the attachment period should give interesting information
as to the prevalence of this type of behavior.

SUMMARY.

Full-grown larva? of the oyster shortly before attachment move
over an appreciable area of solid surface testing it out with the
foot.

Actual attachment is preceded by circling movements, the
larva finally coming to rest with the left valve held in contact
with the substratum by the foot.

The suggestion of Stafford, 13, that the foot is the organ of
attachment has been confirmed by direct observation. The use
of the foot and of the mantle during fixation is here described.

The circling movements of the larva while crawling over a
substratum not only aid it in obtaining a favorable foothold,
but probably are also instrumental in producing a fairly even
distribution of the spat.

CITATIONS.
Belding, D. L.

'10 A Report upon the Scallop Fishery of Massachusetts. The Commonwealth

of Massachusetts, Boston.
'12 A Report upon the Oyster and Quahog Fisheries of Massachusetts. The

Commonwealth of Massachusetts, Boston.
Brooks, W. K.

'79 Abstract of Observations upon the Artificial Fertilization of Oyster Eggs
and the Embryology of the American Oyster. Am. Jour. Sci., New
Haven, Vol. XVIII. pp. 425-427.

'80 The Development of the Oyster. Studies from the Biol. Lab. Johns Hop-
kins Univ. No. IV., pp. 1-106. Baltimore.
Crozier, W. J.

'21 Notes on Some Problems of Adaptation. 5. The Phototropism of Lima.

BIOL. BUL., Vol. XVI., No. 2, p. 102.
Field, I. A.

'09 The Food Value of Sea Mussels. Bull. U. S. F. C., Vol. X XI X., p. 87.
Huxley, T. H.

'83 Oysters and the Oyster Question. Eng. Illus. Mag., Oct. and Nov., pp.

47-55, 112-121.
Jackson, R. T.

"88 The Development of the Oyster with Remarks on Allied Genera. Proc.

Bost. Soc. Nat. Hist., Vol. X XIII., p. 531.

'90 The Phylogeny of the Pelecypoda. Mem. Bost. Soc. Nat. Hist., IV., p. .277.
Kellogg, J. L.

'99 ( )bservations of the Life History of the Common Clam, Mya arcnaria.
Bull. U. S. F. C.. 1899.

THE ATTACHMENT OF OYSTER LARV.E. 151

Nelson, J.

Annual Reports of the Department of Biology. N. J. Agr. Experiment Sta-
tion, 1888-1893, 1900-1915. New Brunswick, N. J.
Nelson, T. C.
'17 K'-|>ort of the Department of Biology for 1916.

'21 A Aids to Successful Oyster Culture. Part I., Procuring the Seed. Bulle-
tin 351.

'21 B R«-|)ort of the Department ol Biology for 1920.

'23 B Report of the Department of Biology for 1922. N. J. Agr. Experiment
Station, New Brunswick, N. J.

"23 A The Attachment of Oyster I «arvse. Abstr in Anat. Rec.. Vol. 24, No. 6,

P- 395-
Ryder, J. A.

'82 Notes on tlu- Brooding, Foot!, and Green Color of the Oyster. Bull.

U. S. F. ( '.. 1.. p. 403.
Sigerfoos, C. P.

'07 Natural History. Organization, and Late Development of the Teredinidse.

or Shipworms. Bull. T. S. F. C.. Vol. X XVII.. p. 193-
Stafford, J.

'10 The Larva and Spat of the Canadian Oyster. Am. Nat.. Vol. XL1V.. p.

343-
'13 The i .in. i'li. in Oyster. Commission of Conservation. Ottawa.

Vol. XL VI April, IQ24 No. 4

BIOLOGICAL BULLETIN

THI-: m.HAVInR OF THE NUCLEUS AND CHROMO-

SOMI.s DURING SPERMATOGEXESIS IX THE

ROBBER FLY LASIOPOGOX BIVITTATUS.1

< HAS, \V. METZ AND JOSE F. NONIDEZ.
CARM'.ii [NST11 WASHINGTON, DEPARTMENT OP GENETICS, AND CORN

avERsiTY MEDICAL COLLEGE.

INTRODUCTION.

tin- growth period of the primary sperm. itocytes in
Ln\ii'(>"['ini hii'nttulns L\v.,2 nuclear changes occur which trans-
form tlu- -pherical nucleus into a series of ramifying ch.nnl
endo-in^ i he chromosome threads. Conditions in these nucK-i
gesl th.ii tin- -taiiuible threads are axes or cores of cr >n-iderably
IT liodit •-, which are mainly hyaline. The pre-ent account U
com niicd primarily with these features, although it include- a
ill -' ripti'iii «>t' >|)ermatogenesis up to the second s|>«-nn.it<.cyu-
di\ i-imi.

The nl.-ervations are based on two specimen-, collected ne.ir
Slit i id. in. Wyoming. Both were fixed in strong l-'leinmin^ and
>t. lined v> ith iron ha?matoxylin. One (_'"-- was counterstained
\\ith lij;ht screen. The technique employed w.is the -.nne .1- that
refill. irly used in preparing dipteran material, and the fixation is
i^ood in luith specimens. So far as we know the peculiar nuclear
behavior is characteristic of the species and does not represent
an accidental abnormality.3

\ pn-liminary account of tli; .<-n U-lnn.- t; ;iu nt.il

Biology and Medicine and published in tlir Socift;.'- I'r. ici-i-dini; and

Nonidcz, '22).

* \Y. li-liti-d to Mr. Jus. R. Hine for thr specie-; 'l.-tcrniinatioti.

' This is supported by the fact that conditions sonii-uliat similar have been found
in tun t'thri < 'f Mies (.discussed below). In these our material : ntly

extensive to prove that the conditions are normal.

153

154 CHAS. W. METZ AXD JOSE F. XOXIDEZ.

The testes of Lasiopogon resemble those of Asilus and many
other members of the Asilidae, in being a pair of long, coiled tubes,
with the spermatogonia near the apex, and successive stages in
spermatocyte development progressing in serial order toward the
base. This serial orientation of stages makes it relatively easy to
follow the succession of events during spermatogenesis.

SPERMATOGONIA.

At the extreme tip of the testis lies a small mass of disorgan-
ized, nutritive material containing only a few cells. Adjacent
to this lie the spermatogonia, many of which are connected with
the nutritive mass by long protoplasmic processes. This con-
dition differs from that in many other asilids. In the latter the
nutritive mass lies in the center of the tube, a short distance from
the end, and is surrounded on all sides by spermatogonia.

The spermatogonia present no unusual features except in late
prophase. There the nuclei appear to undergo a change similar
to that observed in the primary spermatocytes. The nuclear
membrane breaks down or becomes irregular in outline and the
chromosomes appear to lie free in the cytoplasm, each surrounded
by a hyaline zone of about the thickness of the chromosome. The
other stages in spermatogonial development appear to be essen-
tially the same as those in Asilus sericeus (Metz and Nonidez, '21).

The spermatogonial chromosome group probably consists of
six pairs, or possibly of five pairs of autosomes and an unpaired X-
chromosome. Fig. I only shows five pairs, but the nucleus may
not be entire. One or two other figures (not suitable for drawing)
seem to include six pairs, and spermatocyte metaphases almost
certainly include six chromosomes (e.g., Fig. 12).

THE GROWTH PERIOD.

Owing to the scarcity of final spermatogonial division figures
and to the fact that the final spermatogonia are not readily iden-
tified by size differences, as they are in some asilids, it is difficult
to detect the boundary between spermatogonia and spermato-
cytes. However, the earliest growth stages seem to repeat the
spermatogonial prophase changes, as in the case of Asilus sericeus
(1. c., p. 171); and so far as we can determine they show essen-

BEHAVIOR OF NUCLEUS AXD CHROMOSOMES. 155

tially the same features as does Asilns. Hence they \vill be
passed over rapidly.

The pairs of chromosomes appear soon after the final spermato-
gonial telophase as granular aggregates (Fig. 2). They are less
druse and clear cut than in Asihis, and more nearly resemble
those in Dasyllis (Metz, '22). These aggregates soon loosen up,
overlap, and spin out into long, granulated threads, intermingled
in a net-like structure. The nucleolus, a dense, spherical body,
is prominent from the earliest observed stages (see special account
of the nucleolus below).

Soon after the growth period has begun, the chromatin threads
become drawn out so finely that they cannot be traced far, indi-
vidually. They are heavily granulated, with prominent chro-
momeres. From their history and from their subsequent be-
havior, as well as from analogy with other forms, they are assumed
to be bivalent, although no duality is visible in their structure.
Then- is no conspicuous polarization of the threads at this time,
nor is there any evidence of a synizesis or dense contraction
stage. The nucleus is clear cut in outline and is of the usual more
or less spherical shape, as is also the nucleolus.

At this point changes set in which accompany one another as
growth progresses. The nucleolus elongates and eventually be-
comes double; the chromatin threads undergo a gradual con-
densation; and the smooth contour of the nucleus becomes
broken by a series of imaginations and evaginations, at first more
or less irregular, but soon conforming to the outline of the thrr.nl-
liki- chromosomes. It is difficult to determine what phy-iral
changes occur in the cell; but the nucleus acts as if its tension
were released, so that the nuclear membrane offered no n-i-t.mce
to movements of the cytoplasm or of the chn>m"-'>im--. The
former appears to flow in around the chromatin thread-, while
tin- latter become extended out into tlir < -\ -topla-m. Si. mi tin-
nucleus is converted into a series of lonu l<>U-s or jx.rkrt- rami-
fying throughout the cell. Each pocket, .1- .1 rule, encloses only
one (bivalent) chromatin thread. These changes are n -|>n •-( nted
by the cross sections shown in Figs. 2 to 4. The -urf.i-v of the
nucleus is tremendously increased by this proce», bin tin- volume
is so difficult to estimate that we are unable to determine whether
it is affected or not.

156 CHAS. W. METZ AND JOSE F. XOXIDEZ.

Little or no regularity of plan is to be seen in the configuration
of the nucleus. Its shape is varied and often bizarre. Appar-
ently the lobes ramify at random, following the contour of the
long, more or less twisted threads. The nucleolus usually lies in
a lobe or pocket by itself, or in a chamber at the junction of two
or more lobes.

The lobes themselves are of various shapes, flat, pocket-like
sheath-like, cylindrical, or of intermediate sorts. It is to be
noted, however, that all of the nuclei undergo the same series of
changes and show essentially the same features. There is no
irregularity in this respect.

The maximum of distention and distortion of the nucleus is
reached soon after the process begins, and while the chromatin
threads are long and slender. At this time the lobes are so
attenuated and slender that the nucleus is almost lost to view.
For the most part it can only be seen as narrow, winding channels,
ramifying through the cytoplasm. These channels often extend
almost or quite to the periphery of the cell. Since each channel
usually contains a single (bivalent) thread, the threads thus be-
come almost completely isolated from one another.

Subsequently the threads undergo a gradual condensation, and
the lobes a coincident contraction, as growth progresses (Figs.

5-9)-

When the detailed structure of the lobes is examined, at any

stage, it is seen that almost invariably the deeply staining chro-
matin thread (chromosome), extends through the center of the
hyaline lobe like a core, if the lobe is cylindrical, or lies midway
between the sides, if the lobe is flat. It is separated from the
membrane by a transparent layer of approximately uniform thick-
ness. The structure might be likened to an insulated wire, in
which the chromosome represents the wire and the hya|ine region
the insulation.

As the lobe is followed beyond the chromosome it is seen to
become narrow or closed. The maximum thickness is almost
always in the region occupied by the chromosome. This suggests
that the chromosome is surrounded by a cortex of dense or gela-
tinous material which holds off the nuclear membrane.

NO . --! Mil ,! ( haiige in tlu-sr conditions is I" l>r "1 >-IT\ vl as
the chromosomes condense and the lobes contract. The two

BEHAVIOR OF NUCLEUS AND CHROMOSOMES. 157

processes apparently go together. As the chromosome thickens,
the surrounding layer thickens, increasing the diameter of the
lobe in that region. This process gradually progresses up to the
prophase, when, as the chromosomes approach their maximum
condensation the lobes become decreased in size. In the late
growth period there is a suggestion of a secondary lengthening
out of the chromosomes and the lobes into a more ramifying con-
••1 it ion such as exists at an earlier stage, but the change is not
extreme, and may not be constant. It suggests, ho\vever, the
stage in Asilus sericens (I. c.) in which the threads spread apart
after undergoing contraction. This also occurs relatively late
in the growth period.

In Lasiopozon there is no "contraction stage," such as occurs
in Asilus; but throughout most of the growth period there are
some indications of a polarization of the threads toward the
nucleolus; and often faint connections are to be seen between tin-
two (Figs. 7 and 8).

I Hiring the prophase, and up to a late stage in metaphase, the
hyaline layer around the chromosomes remains conspicuous, as
shown in Figs. 10 i.v The spindle fiber may often be traced t
projection of this layer corresponding to that of the chromosome,
and thence through the layer to the chromosome.

Not enough nnaphase figures are present in our material to
permit us to follow the chromosomes through this st.

During the late growth period and the first sperm. iu>. \ u- divi-
MIUI the mitochondria become very prominent, and by darkening
tin- r> toplasm, aid in bringing the outline of the chromosomal
sheath into relief. In metaphase large mitochondri.il n><l- are.
1 1 < •< iiiently present which are almost as dense as the chn>m< >s< m
These are seen longitudinally in Figs. loand i;v .iml in trail-verse
section in Figs. II and 12. It is to be noted that they are not
surrounded by a hyaline cortex comparable to that around the

chromosomes.

THE XUCLEOLUS.

The nucleolar behavior during the growth period is somewhat
different in our two specimens. In one (2056) there i> a single
nucleolus in the early growth stages.while in the other (2055 1
there are two. In the following description the two specimens
are considered separately.

158 CHAS. W. METZ AND JOSE F. XOXIDEZ.

In number 2056 the nucleolus is prominent from the beginning
of the growth period. It grows gradually and serves as an addi-
tional criterion for seriating the growth stages. When the
nucleus begins to become irregular in outline, the nucleolus elong-
ates and becomes bilobed, then double. One portion is large,
dense and irregularly shaped; the other is smaller, less dense and
nearly spherical — resembling chromatin in texture. The two
parts are often connected by a dark, thin thread. In the late
growth stages the large, irregular portion frequently lies near the
periphery of the cell, suggesting that it may be cast off. We
have not been able to determine its fate, on account of the scarcity
of prophase figures, but it seems to disappear before the chromo-
somes go on the spindle. The smaller portion appears on the
spindle in its characteristic spherical form, but reduced in size, and
presumably represents the sex chromosomes.

In specimen 2055 there are normally two, equal sized, spherical
nucleoli in the early growth stages. Just before the nucleus
begins to become lobulated the twro nucleoli come close together
and apparently unite, although it is possible that one may de-
generate after they become approximated. Following this the
nucleolus becomes bilobed as in specimen 2056, and subsequently
its history is the same as that of the latter.

In both specimens the nucleolar structures are surrounded by a
hyaline cortical region similar to that of the autosomes, although
it may be noted that the cortex of the large, irregular portion is
usually thinner than that of the small, spherical part which prob-
ably represents the sex chromosomes.

DISCUSSION.

The constancy of the hyaline zone around the chromatic
threads in the nuclear lobes we interpret as indicating the presence
of a relatively soiid (gelatinous) cortical layer enveloping the
respective chromosomes. The presence of the layer is revealed,
of course, by the structure of the lobes. If the nuclei were
spherical it would not be detected. This raises the question as to
whether the condition is restricted to the present species, or is of
more general occurrence but usually invisible.

We have noted above that evidence of the hyaline zone is to be
seen in spermatogonia as well as in spermatocytes of Lasiopogon.

BEHAVIOR OF NUCLEUS AND CHROMOSOMES. 159

In the other Diptera known to us,1 conditions are not usually
such as to reveal the presence of a chromosomal envelope if one
its. But indications of one are to be seen here and there,
« --penally in late prophases or metaphases. And in one genus,
owing to the unusual nuclear behavior — somewhat like that in
I.-! i«j)ogon — the evidence is strong. This genus is Ptecticus, a
number of the Stratiomyidie. It will be considered more fully in
another paper, but we may note here that in P. trivittatits and
P. sackenii, the two species we have studied, there is a distinct
hyaline zone around the chromosomes of the prophase and meta-
ph.ise of the spermatocyte divisions, and there are indications of
such a structure extending back into the growth period. These
•ures are shown in part by Figs. 14 and 15, from prim.iry
spermatocytes of P. trivitlatns.

This condition, in a family distinct from that to which Lasi-
'>n belongs, suggests that we are dealing with something which
is widespread — at least among the Diptera.

In other organisms the nearest approach to conditions like
those described above, so far as we are aware, is found in the
Orthoptera. Professor Mc(Mung informs us that in this group
prophase and metaphasc chromosomes (of primary spermato-
cytes) are frequently bordered by clear spaces, differentiated
from the surrounding cytoplasm, and that the condition is suffi-
ciently characteristic to suggest the constant presence of a
hyaline zone around the chromosomes in these stages. As
examples may be cited McClung's published photographs of the
in-t spermatocyte chromosomes of Mermeria (Mr('hmg, '14,
I igs. 129-132) and of Ilesperotettix (McClung '17, plate 8, I
c}. In all of these the hyaline zone is visible around the chro-
mosomes.2 We have observed this condition in fn-t -pcrmato-
i-yte anaphases of Rhomaleitm and find that the hyaline l.i\«-r
agrees in appearance with that in Lasiopogon, alt hough we have
not as yet tried to trace its history.

The presence of this condition in the < )rthopu-r.i .md I )ij>tera
It. i- prompted us to make a brief comparison with other structures
of a somewhat similar nature.

No close relatives of Lasiopogon have been examim-il a- y«-t.
- It is probable that other cases of this sort have escaped our notice, although we
li.ive examined numerous published photographs of chromosomes with ttiis point in
mind.

l6O CHAS. W. METZ AND JOSE F. XONIDEZ.

These include, for the most part, the "chromosomal vesicles"
of Conklin ('01, '02, etc.), Wenrich (16), Richards (17) and
others. A resume of the observations on these structures up to
1917 has been given by Richards (/. r.), who cites examples in
various organisms from molluscs to vertebrates (fish eggs). There
is considerable difference in the structure of the vesicles in differ-
ent forms, but at certain stages in several, if not all, of them it
consists of a hyaline outer layer surrounding a more or less defi-
nite chromatic axis, suggestive of the lobes in Lasiopogon. This
is clearly the case in spermatogonial telophases of the Orthoptera
and in Fund ill us eggs (cf. Wenrich, Richards). The stages in
which such a condition exists (prophases in the case of Fundulus)
may be preceded by others in which the chromatic material is
scattered through the vesicle, or even largely applied to the sur-
face (Crepidula, Conklin '02). In Fundulus eggs and in the
Orthoptera the anaphase chromosome swells into a vesicle in
which the chromatin appears as a granulated network. This
persists through the resting stage and then the network condenses
into a chromatic thread or rod, extending more or less axially
through the vesicle.

If we assume that the hyaline portion of the vesicle is more
dense or gelatinous than the surrounding cytoplasm, as is sug-
gested by the appearance and behavior of the vesicles, then the
structure bears a close resemblance to that in Lasiopogon. On
this basis we might conceive of the structures in Lasiopogon as
arising in a similar fashion — i.e., by the swelling of telophase
chromosomes into vesicles, with a diffusion or dispersal of the
chromatin (as seen in the early growth stages) and then its re-
condensation within the vesicle. The term "vesicle" is perhaps
inexact in the case of Lasiopogon for no limiting membrane, ex-
cept the nuclear membrane, is visible around the hyaline layer.
But the periphery of the latter would, on this view, be consid-
ered as analogous to the periphery of the vesicle, whether en-
veloped in a membrane or not.

The only other description of a structure similar to these, so
far as we know, is that of Lee's "chromosomal sheath. " l In the
case of the Orthoptera cited by Lee his sheath evidently corre-

1 We are indebted to Professor L. W. Sharp for calling our attention to Lee's
paper and to de Litardicre's criticism of it mentioned below.

BEHAVIOR OF NUCLEUS AXD CHROMOSOMES. l6l

spends to the hyaline portion of the vesicle considered above, and
as such is similar to that in Lasiopogon. But his conception of a
-heath as characteristic of animal chromosomes and possibly also
of plant chromosomes, seems to us to be open to serious question,
at least in the form stated. It has been considered at length in a
recent paper by de Litardiere ('21), who interprets the "sheath"
as an artifact — probably a fixation product. The latter author
maintains that the same sort of zone is to be seen around the
mitochondria, cytoplasmic inclusions, etc., and hence has no
special connection with chromosomes. We may omit further
consideration of these structures, therefore, until the evidence
becomes clearer.

Possibly de Litardiere's criticism may apply to the cases we
have cited here of "indications" of a hyaline zone around chro-
mosomes in the Diptera, including spennatogonial chromosomes
of Lasiopogon; and it may apply to the zone seen in Mr( "lung's
figures referred to above. Further evidence is required to settle
this point. But we may note that in these cases the zone seems
to be found only around the chromosomes, not around other
structures in the cell.

It may be added that in the case of the spcrmatocytes of Lasio-
pogon described here the structures are clearly not artifacts, and
are confined to the chromosomes (including the nucleolus which,
from analogy with other Diptera, is assumed to be chromosomal
in nature). Numerous large mitochondria! rods are present, but
are not surrounded by sheaths comparable to those enveloping
the chromosomes.

If we omit the questionable cases, such as those just mentioned
as possibly due to artifacts, it is clear that few forms exhibit
structures comparable to those in Lasiopogon, although the possi-
bility still remains that the structures may be present in other-
ii when not visible.

In the foregoing discussion little emphasis ha- l>r< n laid on the
fact that some of the cases considered involve bivalent chromo-
somes (maturation stages) and others univaknt chromo-omc-.
It is not yet clear that any distinction should be made, so fai
the present subject is concerned. But it may l-e noted in clo-iu;c
that any view which considers the chromosome .1- rncli.-cd in a
-heath or envelope must take account of the approximation of

1 62 CHAS. \V. METZ AXD JOSE F. NOXIDEZ.

homologous chromosomes during synapsis, and, in the case of the
Diptera, must allow for such an approximation in every cell
generation (cf. Metz, '16) This would seem to necessitate the
assumption that the envelopes of homologous chromosomes could
unite.

LITERATURE CITED.
Conklin, E. G.

'02 Karyokinesis and Cytokinesis. Jour. Acad. Nat. Sci. Phila., 12.
Lee, A. Bolles.

'20 The Structure of Certain Chromosomes and the Mechanism of their Divi-
sion. Quar. Jour. Micro. Sci., 65 : 1-32.
Litardiere, R. de.

Recherches sur 1'eloment chromosomique dans la caryocinese somatique

des Filicinecs. La Cellule 31 : 255-473.
McClung, C. E.

'14 A Comparative Study of the Chromosomes in Orthopteran Spermatogenesis.

Jour. Morph. 25 : 651-749.
McClung, C. E.

'17 The Multiple Chromosomes of Hesperotettix and Menneria (Orthoptera).

Jour. Morph. 29 : 519-605.
Metz, C. W.

'16 Chromoiome Studies on the Diptera. II. The Paired Association of Chrom-
osomes in the Diptera, and its Significance. Jour, of Exp. Zool., Vol. 21,
No. 2.
Metz, C. W.

'22 Chromosome Studies on the Diptera. IV. Incomplete Synapsis of Chromo-
somes in Dasyllis grossa Fabr. BIOL. BULL., 43, No. 4, October, 1922.
Metz, C. W., and Jose F. Nonidez.

'21 Spermatogenesis in the fly Asilus sericetis Say. Jour. Exp. Zool., Vol. 32, No.

i.
Metz, C. W., and Jose F. Nonidez.

'22 Observations on the Behavior of the Nucleus and Chromosomes in Spcr-
matocytesof Lasiopogon (Diptera). Proc. Soc. Exp. Biol. and Med.. 19 :

373-374-
Richards, A.

'17 The History of the Chromosomal Vesicles in Fiindulus and the Theory of
Genetic Continuity of Chromosomes. BIOL. BULL., Vol. 32, No. 4,
April, 1917.
Wenrich, D. H.

'16 The Spermatogenesis of Phyrnotettix ma gnus with Special Reference to
Synapsis and the Individuality of the Chromosomes. Bull. Mus. Comp.
Zocil. Harvard, Vol. 60, No. 3.

164 CHAS. W. METZ AXD JOSE F. XOXIDEZ

EXPLANATION OF FIGURES.

The figures were drawn with the aid of a camera lucida, using a Zeiss 2 mm. oil
immersion objective and number 12 ocular. Magnification about 1600 diameters.

All figures except number 4 are from specimen number 2056.

It should be emphasized that the figures represent cross sections at one level
in most cases, and do not represent entire nuclei.

FIG. i. Spermatogonial metaphase, probably not entire.

FIGS. 2-10. First spermatocyte growth period.

FIG. 2. Very early growth stage.

FIG. 3. Slightly later, nucleus becoming tabulated.

FIG. 4. Later, but still in the early part of the growth period; nucleus com-
pletely transformed into narrow ramifying chambers.

FIGS. 5-10. Various aspects of the nuclei during the successively later growth
stages, showing the lobes and chromosomes cut longitudinally, transversely and at
intermediate angles. The isolated portions in figures 8 and 9 are sections of lobes,
not separate vesicles.

FIGS. 11-13. First spermatocyte metaphase in polar and lateral views, showing
the persistence of the hyaline cortical zone.

FIGS. 14-15. Ptecticus (see text, p. 159).

BIOLOGICAL BULLETIN, VOL. XLVI.

PLATE I.

1

-*

/f

« .

1 ^

_ - ,

.

—^

• -•

7

. 5

v .

8

* . •

10

i

«

F

• x .

11

12

^

13

15

C. W. METZ AND J. F. NONIOEZ

THE SUSCEPTIBILITY OF CELLS TO RADIUM RA-
DIATIONS.

CHARLES PACKARD.

PEKING UNION MEDICAL COLLEGE.
DEPARTMENT OF BIOLOGY.

1 t has frequently been observed that cells may show at different
periods of their existence marked variations in susceptibility to
radium radiations. An embryonic cell is more quickly injured
than the same cell in the adult condition: at the metaphase,
cells are more sensitive than they are immediately before or
after that brief period (i). Bohn (2) first suggested that the un-
derlying cause of such differences in response must be sought in the
physiological condition of the cells at the time of radiation. That
they cannot be due to changes in the absorptive power of proto-
plasm is obvious: whether a cell is sensitive or not, the rays are
absorbed to the same extent. The actual changes produced by
them in protoplasm must therefore be the same. But the reac-
tion of the cell to such changes differs immensely. \\e may
therefore say that a cell is susceptible when it is in such a physio-
logical condition that a modification produced by the rays re-
sults in a greater or less injury, and that it is resistant when the
same modification is not followed by injurious effects.

The object of the present paper is to show that among the con-
ditions which affect the susceptibility of cells to radium nidia-
tions are (i) the temperature of the cells at the time of exposure,
and (2) the relative permeability of the surface layer of the cell.

The experiments to be described were carried out on certain
Protozoa, for these cells are better adapted to this kind of experi-
mentation than any others. They can live in both high .md low
temperatures without injury: and different genera show in. irked
differences in permeability. Cells of the same species, even
descendants of the same individual, vary widely in their re.ictic.n
to radium radiations at different periods of their life cycle,
is perhaps the best cell for experimental purposes
it is more susceptible than any other common type.

1 66 CHARLES PACKARD.

I will not here review the results obtained by other investi-
gators who have tested the effects of radium radiations and X-
rays on Protozoa. Different methods of exposing the cells lead
to such wide variations in results that comparisons cannot safely
be made. In general, however, cells which are sensitive to the one
are also sensitive to the other type of radiation. Great differ-
ences in the susceptibility of cells of the same species have been
reported. Thus Zuelzer (3) states that Pelomyxa palustris when
exposed to 6 mg. of radium element dies sometimes within one
hour and sometimes only after four hours of continuous exposure.
Such differences make any conclusion as to the length of the
lethal dose out of the question. But with appropriate methods
the lethal dose is found to be constant.

METHODS.

The method used in the following experiments is this. The
radium was enclosed in a glass capsule which prevented the alpha
rays from escaping. In the experiments on the relation of tem-
perature and permeability to susceptibility, the strength equalled
13.4 mg. of element. In the third series, on the change in per-
meability induced by the radiations, the strength was 25 mg. of
element. The radium tube was used unscreened and was sup-
ported above the drop of culture medium at a distance of 2 mm.
Thus all of the rays which emerged from the lower side of the
tube could reach the cells. The whole preparation was kept in a
moist chamber at the desired temperature.

In order to determine what type of rays produced the effects
which are to be described, I interposed between the radium tube
and the Paramcecia lead sheets of various thicknesses, thus filter-
ing out the more penetrating rays. All of the beta rays are
stopped by 2 mm. of lead: the gamma rays are not affected. It
became apparent at once that the changes produced in the
Paramcecia were due to the action of the slowest beta rays, for
when a lead screen of 0.12 mm. was interposed the Paramcecia
were affected hardly at all. This is to be expected, for the sur-
face layer of the cells is very thin and can absorb only those rays
which have a low velocity: it offers almost no resistance to rays
having considerable powers of penetration.

In conducting these experiments it was found necessary to use

-I -CEPTIBILITY OF CELLS TO RADIUM RADIATION-. 167

only Paramcecia from a pure culture, for unrelated wild cells show
great variations in their susceptibility to the rays. Another
necessary condition has been mentioned by Jacobs (4), namely,
that in each test the same amount of liquid must be used.

THE REACTION OF PARAMCECIUM TO RADIUM RADIATION -

When a Paramceciiim is exposed to radium radiations under the
conditions described, it quickens its movements at first and then
gradually slows down and ceases to swim unless thedi-h is -hakeii.
Later the contractile vacuoles pulsate more and more -lowly ami
finally stop, usually in the expanded condition. If radiation i-
longer continued, a typical cytolysis ensues. The cells imbibe
water, swelling considerably in consequence, and the ectopla-m
bulges out in the form of clear vesicles which later run together.
Then the pellicle separates from the rest of the cell carrying with
it the cilia. The protoplasm is now highly fluid. At thi- time
the macronucleus, in stained preparations, is seen to be divided
into several parts. Not infrequently the cells burst violently.
These phenomena are in every point similar to those which are
observed when Parania'cium is treated with a variety of cytolyiic
agents, as described by Biulgett (5), Harvey (6), and Jacobs

THE RELATION OF TEMPERATURE TO SUSCEPTIBILITY.

Rhodenburg and Prime (7) first pointed out that there is a defi-
nite correlation between temperature and the susceptibility of
cells when treated with X-rays. In their experiment- tlic\
posed mouse sarcoma in vitro at a temperature of 42° (".
definite dose of X-rays, and then inoculated healthy mice \\ ith tin-
radiated cells. At this temperature 10 per cent, of the inocula-
tions failed to take. When the cells were radiated at 4.V° ('.. 76
percent, of the inoculations failed. Control experiment- pn.\ed
that the-e temperatures alone are not sufficient to produce thi-
eiiect. The combination of high but sublethal temperature-
\\ith radiation was five times as effective as radiation alone.

Mammalian tissue cannot be subjected to wide variation- in
temperature, but the Protozoa can live normally at temperature-
as low as 15° C. and as high as 37° C. In the following experi-
ment- these were the limits employed.

When Paranicecia are radiated at hi-Ji temperatures, they sue-

1 68

CHARLES PACKARD.

cumb much more quickly than at low temperatures. The results
are shown in the accompanying figure. An analysis of the curve
shows that for each increase of about 8° C. the length of the
lethal dose is halved. The curve is thus similar to that which
expresses the relation between temperature and the velocity of a
great number of reactions both inorganic and physiological.

Hours of

345

Radiation

678

10

37
36
35
34-
33
3Z
3 I
Ci) 30
329
b*6
o5 2,7

0-9 A

Ezs

.0 2-5
"24
^o 23
cj ^^

8^'
£20

19
18

17
16

I 5
14

i

I

r

do

T

A

°N

\

\

\«

\,

>

\

\

\

\,

N

Xk

%

\

\

xs

^

X

"N

ND

TEXT-FIGURE i. The effect of temperature on susceptibility.

Snyder (8) cites more than fifty physiological reactions which con-
form to this type of curve : Woodruff and Baitsell (9) show also that
the division rate of Paramcecium aurelia varies similarly with
changing temperature.

The cells were not injured by these temperatures alone. While

SUSCEPTIBILITY OF CELLS TO RADIUM RADIATION-. 169

the experiment- were in progress the room temperature varied
from 30° to 36° C. and the cultures flourished. Single lines of cells
showed a steady divi>ion rate of about 2^ divisions per day. In
the coolest temperatures the cells remained normal, and regained
the usual division rate on being brought back into a warmer place.
This increase in -u-ceptibility at high temperatures is not due to
any in< Tea-ed a< ti\ itv of the radiations, nor to any change in the
power of pron.pl.i-m to absorb the rays. The amount of radiation
.il.-orl/ed i- <li t. •riiiiiu •<! by the atomic constitution of protopla-m
and thi- dot-- not vary materially during changes in temperature.
< >\\c of the condition- which varies with the temperature i- the
permeability of the cell membrane. Hober (10) stair- that the per-
meability of plant ceil- is doubled with each increa-e «\ io ( "., the
(!•!!- I" Jit time- as permeable at 30° as they a IT at O° C.

That r<irnnnri i<i are more permeable in warm than in cold solu-
tions can be demonstrated by staining them at 30° and 14° C.
in m-utral red. In the higher temperature they stain deeply:
.it tin- lower, they do not stain at all. Whether a change in per-
meability is the only cause of increased sensiti\ene— to radia-
tions remains to be demonstrated. That it is an important
factor i- -hown in the next section.

I in Ki I.ATIOX OF PERMEABILITY TO Sust i i-i 11:11.11 v.

Tin- fart that cells are most sensitive to radiation- when their
permeability is increased by heat suggests that the two phenom-
ena are rcLted. I lere again the Protozoa are admirably adapted
te-tiiii; thi- point, for the permeability of the cell membrane
can l>e nu-a-ured in the living condition. The method employed
in these experiments is that followed by Harvey (> . Anumberof
I\ininiti-< i>! from a pure culture are drawn up into a capillary
(lipette which is calibrated so that exactly the same amount of
liquid i- taken in each experiment. The cells are stained for ten
minute- in a solution of 0.02 per cent, neutral red mixed with
[O CC. "f tap water. At the end of this time the vacuoU> at the
P->-terior end of the cells are a bright pink. The -urrounding
l>rotopla-m is also colored. The neutral red in this dilution is
not toxic, although in more concentrated solutions it produces
cytolysis.

The cells are now drawn up in a calibrated pipette and added

I/O CHARLES PACKARD.

to exactly 2 cc. of w/i28o XH4OH solution. The amount of
liquid thus added reduces the strength of the ammonia to W/I3OO.
The stained Paramcecia when put into this solution give first
the avoiding reaction and quickly begin to lose color. The pro-
toplasm, and later the gastric vacuoles, turn yellow, and finally
become colorless. There is a wide variation in the rate at which
the color fades in individual cells, some destaining in three or
four minutes while others retain some pink color as long as ten
minutes. In order to determine the average time for destaining
the usual number of cells (about 40), I used the following method.
The cells were observed, during their destaining, under a bin-
ocular microscope and each one, as soon as it lost color, \vas re-
moved and the time which had elapsed since its first entrance into
the ammonia solution recorded. The following measurements,
typical of many, indicate the rate of destaining.

PARAMCECIUM DESTAINED IN n/i28o NHjOH.
Minutes Elapsed. No. of Cells Destained.

i o

2 o

3 0

4 3

5 2

6 4

7 5

8 8

9 6

10 6

ii 6

12 5

Total 45 Ave. 8.5 min.

By this method the personal equation is greatly reduced since
the observer cannot form any idea of what the average will be
until the entire number of cells has been destained. Many tests
on the same pure culture of Paramcecium gave very constant
results, the average time of destaining in tests carried out on the
same day varying less than one half minute. Other lines of
Paramcecia showed somewhat different averages but even here
they differed from each other by not more than one and one half
minutes.

A study of Paramcecium cells at different phases of their life
cycle shows that the permeability varies, being much greater at
the time of conjugation than at any other period. Thus among
four lines of cells in which conjugation did not occur, the de-

SUSCEPTIBILITY OF CELLS TO RADIUM RADIATIONS. i;i

staining time varied from 7.8 minutes to 9 minutes. In a culture
which was undergoing an epidemic of conjugation the pairs
showed an average destaining time of 4.5 minutes. This culture
was presumably not pure, and the pairs in their reaction to the
ammonia showed a wider variation than was found in any homo-
geneous group. But the difference in permeability between con-
jugant- and non-con jugants was in every case large enough to be
!iiti« .nit.

Other Protr>/na diitcr from Paramcecinm in permeability. A-
Har\ey ha- puinted out, Stylonichia and Oxytricha are compara-
ti\elv impermeable. Indeed, after ten minutes in neutral red
:ie ii-u il <Min duration they have taken up almo-t no color at
all. The -lain must act for twenty-five to thirty-five minutes
IK hui ilic-« <-i-lls are stained sufficiently for experimental pur-
poses, and e\en after this time they are not as hiyjily coloiv:

•iniii'innt \> after ten minutes. The time required fur <K-
stainini; Stylonichia varies somewhat in different cultures, but
tin- average is approximately forty minutes.

\\ hen the relative permeability of these cells is compared with
the length of their lethal dose of radium radiation-, we find a
di.-e tuMvlation between the two measurement-: that i-. cells
\\hich are relatively permeable are quickly killed by the i
\\hile those which are less so are more resistant. The following
table indicates these relations.

< "MI'AKISON OF THE DESTAINING TlME AND LETHAL I> R\D|iM R \IHA-

TIOSS AT 27° C.

Destaining Time. Lethal I'

/'.;» am ' >>n single cells 8.6 min. i \\

••: KUijiiK.itintJ 4.5 min. i ' _, li<

nia 40. min. 15!.

It is evident therefore that the susceptibility of thc-e 1 Y. itozoa
to radium radiations varies directly with the permeability of the
surface layer of the cell.

Tin-: EFFECT OF RADIUM RADIATION- ON mi-: CELL MKMHR\NE.
The question naturalU arises, what is the reason for this corre-
lation? The an-\\er is to be found in the fact that the rays which
are absorbed produce in the cell membrane changes which lead
to increased permeability. The experiments described below
indicate the rate at which these changes take place.

172

CHARLES PACKARD.

In the following experiments the radium amounted to 25 mg. of
element. The cells were exposed in the same manner as before,
then stained in neutral red and destained in 11/1280 NH4OH.
The accompanying figure shows the results of experiments per-

Hours
2

of Radiation
34

O

-7

T>

OJ

^6

3

1 54

L

5
4*

Control

Cells

TEXT-FIGURE 2. The effect of radiations on permeability.

formed in one day. It is necessary to make many tests within
the limits of a few hours for the cells vary somewhat in their
reactions to these manipulations with changing conditions of
food, etc. By using two tubes, each amounting to 25 mg. of
element it was possible to perform two experiments simultane-
ously and thus make seven or eight determinations in a single
day. The temperature in these tests remained uniformly .it
22° C.

The control cells showed a slight increase in permeability after
remaining in a small drop of culture medium for some hours.
This however is not significant, for cells left as long as eight
hours show practically the same reaction to ammonia as those

SUSCEPTIBILITY OF CELLS TO RADIUM RADIATIONS. 173

left for only two hours in a small drop. But the radiated cells
display a marked shortening in the time required for destaining,
or, in other words, they show a considerable increase in permea-
bility, the change beginning very soon after the exposure to
radium commences. At the end of four hours the cells are still
alive but are for the most part motionless. If the exposure is
longer continued the cells die, showing the characteristic signs of

cytolysis.

The second figure ~ho\vs the result of a larger number of tests.

hours of Radiation

1234

-glOO

C

S95

-o

£80

o

87S

7°

Wl

a>
P

50

c o

unlrc

Rolaiatc

Cell'

1 i \ i-rii.i'Ki-. .

I li-- • !!• . t of i.i.li.iti..ii .

Since they were made during the course of ni,m\- \\vt/k- the con-
trols varied somewh.it. although on any one day they weiv \»r\
uniform. The time required for destaining is here given in per-
centages of the control time which is reckoned as 100 per cent.
The curve clearly indicates that the normal semi-permeability of

I~4 CHARLES PACKARD.

Paramceciiim increases progressively as the exposure is more and
more prolonged.

Some of the radiated cells were removed to separate drops of
culture medium before treatment with the stain, and their
division rate observed. Those which were exposed for one hour
or a little less frequently showed a higher rate than the controls.
This result has been described by Markowits (i i) who studied the
effect of mesothorium rays on Paramceciiim. The acceleration
can be observed for five or six generations, after which the cells
return to normal and show no evidences of injury. The same
phenomenon I have observed in sea urchin eggs when lightly
radiated (12). Indeed, a stimulation in the rate of growth is of
general occurrence, having been noticed in the case of growing
plants, embryos, tissue cultures, and abnormal tissue growths.

From these facts we may conclude that the slowest beta rays
increase the permeability of the surface layer of Paramcecium
cells. If the exposure is brief, this change in permeability is
accompanied by an acceleration in the rate of cell division. If
it is more prolonged, a destructive cytolysis ensues, and the cells

die.

DISCUSSION.

Radiations which are absorbed at the surface of the cell pro-
duce definite changes which lead to an increase in permeability,
and, if the exposure is sufficiently prolonged, to complete cytol-
ysis. For a definite dose, such changes must be the same in
extent, regardless of the physiological condition of the cell, for
the absorptive power of protoplasm remains constant. Yet it is
quite apparent that a given dose of radiation may result in no
appreciable injury in certain instances, while in others it is
followed by the death of the cells. This is clearly shown in the
text-figures. For example, at 22° C. radiated Pammcecia under-
go cytolysis, under the conditions described , in five hours. From
the beginning of the exposure to the death of the cell there is a
steady increase in permeability. An exposure of half this dura-
tion is followed by no permanent injury. But when evils are
radiated at 30° C. they are cytolyzed after only two and oin- half
hours. Cytolysis occurs when the permeability of the cell has
been raised above a definite limit. It it is already high, due to
high temperature- or to other conditions, the cyt«]yiir action of

SUSCEPTIBILITY OF CELLS TO RADIUM RADIATION-. 175

the rays quickly raises the permeability above the limit and the
cell dies. But if it is low, the lethal point is reached only after a
prolonged exposure.

A similar phenomenon is observed in eggs exposed during dif-
ferent ph.i-t - "f mitosis. At the metaphase their permeability,
as shown by I. yon iv and others, is notably greater than at any
other period : so also is their susceptibility . Ihaveshown 12 'hat
an amount « .f radiation \vhich will induce a quickened cell di\ i-ion
in sea urchin when applied just before or after the meta-

pha-e. ha- .i n 'tarding effect when applied at that period. That
i-. the ( i -11- are mop- -<-n-itive then than they are during the pn>-
pha-e or tclopha-e of mitosis.

The -.tine re-uh- are obtained when other cytolytic agent- are
1 in pi. n e . .1 radium radiations. Lillie ( 14) finds that if freshly
li/eil sea urchin e^g-, which arc highly impermeable,
t re. ited •uith hypotonic sea water they resist its cytolytic action
fur thirty minutes. But if they are placed in thi- -olution when
the clea\agi- furrow appears, they rapidly undergo rytoly-i-.

It appears possible therefore that the susceptibility of cell-
may be rai-ed by the simple expedient of increasing their per-
bility by heat or by some other means. Thi- ha- been done
l>v Rhodenburg and Prime (7) in the experiments already citrd.
ll'.u far thi- method can be used in the treatment of abnormal
ti— ue growths remains to be demonstrated.

The fact that agents which differ so widely as do radium radia-
tion- and h\ po tonic sea water produce the -ana- effects under
-imilar i «> IK lit ions suggests that the action of the-e rays, and al-o
other t \pe-of radiant energy, such as* ultra-violel light and alpha

3, i- not peculiar to themselves. Indeed il may be -aid that
any of tin- rays which are absorbed at the surface of ihe <vll will
cause changes differing in no way from tho-e prodixvd by a great
variety of chemical cytolytic agents.

The mori- penetrating beta ra\ -. tin- gamma, and X-ray-
appe.:r to be so -lightly absorbed at the -urf.n e oi , dl- \vhich are
freel\- iApo-i'd to them that they can produce little or no eff<
Ki. haul- i - ic-ti-d the permeabilit \ of \ariou- eggs and of Arcni-
cola lar\a- after an exposure to X-rays and found no evidence of
any increa-e. I h,i\e-hown that the rapid beta and the gamma
ray- of radium do not act on the surface layer of r<irti»ia-< in»i.

Ij6 CHARLES PACKARD.

But it has been demonstrated that very definite proportions of
these rays are actually absorbed by cells of deep-lying tissues.
And histological study shows that such cells after radiation im-
bibe water, swell considerably, and undergo degenerative changes
which have all the appearance of a typical cytolysis. It is appar-
ent, then, that any kind of rays which are absorbed produce the
same effects. But these more penetrating rays differ from the
less penetrating types in this respect, that they are also able to
bring about degenerative changes in the interior of the cell, par-
ticularly in the nucleus, which the latter, because of their slight
penetrating power, are unable to produce.

I have mentioned the fact that a brief radiation with the slow
beta rays produces an acceleration in the division rate of Paramas-
cium. This effect is not restricted to the action of these par-
ticular rays, for all the radiations of radium, as well as X-rays,
can produce this result if the proper exposure is made. This has
been demonstrated in a great variety of cases, some of which have
already been cited. According to Lillie (16) such acceleration is
the direct outcome of increased permeability, for this condition
allows a freer interchange of CO2 and 0 through the surface layer
and a consequent hastening of all metabolic activities. It may
be possible, therefore, that all cases of stimulation following
exposure to radium radiations and to X-rays can be explained on
this basis.

SUMMARY.

1. The susceptibility of Parafticecium to radium radiations
(chiefly the slowest beta rays) varies with the temperature at the
same rate as do physiological reactions of various kinds.

2. The susceptibility also varies directly with the degree of
permeability of the surface layer of the cell.

3. The slow beta rays act on the surface layer of the cell, in-
creasing its permeability, and if allowed to act long enough,
causing a typical cytolysis. In this respect they resemble other
types of radiant energy, and diverse chemical cytolytic agents.

4. Cells which have a relatively high permeability are more
susceptible than those having low permeability, for the cytolytic
action of the rays is quickly followed in the former by a cytolysis
which is irreversible, while in the latter it is reversible.

SUSCEPTIBILITY OF CELLS TO RADIUM RADIATION-. 177

5. It is suggested that this increase in permeability, following
brief exposures, is the cause of the acceleration in division rate
seen in Paramcecinni and in other cells and tissues.

LITERATURE CITED.

1. Mottram, J. C.

'13 On tin. Aui'-n ima Rays of Radium on the CY11 in Different

•«.•»< if Nuclear Dk i-ion. Arch. Middlesex Hosp. I2th Cancer Ri.-;

2. Bohn, G.

'03 Inline: lium sur lesanimaux en voie . • I'ornpt.

[., V.il. is'i, p. :

3. Zuelzer, K.

"05 I ivadiumstrahlen auf Protozoen. Arch. f. I'nnist..

4. Jacobs, M. H.

'14 .n Various Protozoa. Journ Vol. u.

5. Budgett. S. P.

'99 < tn tli-- Mmil.i -ural Changes Produced by I .u m,l

Am. Journ. Physiol.. Vol. I. p. -
Harvey. E. H.

'ii ility of Cells. Journ. Exp. Zoiil., V..!

~. Rhodenburg and Prime.
'21 I ibincd Radiation and Heat on Neo; Ai. h. Sur-

. p. i 16.

x. Snyder, C. D.
'06 T<-iii|" :il3 of \*arious Physiological Action-.. Am. Journ.

l'h> -lol., \ i,l. 22. p. 309.

9. Woodruff and Baitsell.

'ii Tempei ticients of the Rate of Rcprodm-ti •nae^ium

u. Physiol., \'ol. 29, p. 147.
10. Hober, R.

'14 I' ' li.-inie der Zelle und Gcwebe. 21! 1-M .. p. i j.

i i . Markowits, E.

'22 / \'eranderungen von Paramcecium n.uh Hi-~tr.ihluni; "'it

iiii->otlioiium. Arch. f. Zellforsch.. \'ol. 16. p. 338.
i .•. Packard, C.

"16 1 In- I Radium Radiations on the Rate of < ••!! Di\i-ion. Journ.

1 • \|>. /• ' 1 . \'ol. 21, p. 193.

13. Lyon, E. P.

'04 Khvt! -usceptibility and of COi Production. Am. Journ. 1'liysiol.,

\'ol. II. |>. 52.

14. Lillie, R. S.

'16 1 h.- Physiology of Cell Division. VI. Journ. Exp. Zool., \"ol. 21, p. 371.

15. Richards, A.

'15 K\|><-iimi-nts on X-Radiation as the Cause of Permeability Clia: Am.

Journ. Physiol., \'ol. 36, p. 400.

16. Lillie, R. S.

'09 On the Connection between Changes in Permeability and Stimulation.
Am. Journ. Physiol., Vol. 24, p. 14.

THE PHYSIOLOGICAL AND SYMBIOTIC RELATION-
SHIPS BETWEEN THE INTESTINAL PROTOZOA
OF TERMITES AND THEIR HOST,
WITH SPECIAL REFERENCE TO
RETICULITERMES FLA VIPES
KOLLAR.*

L. R. CLEVELAND.

CONTENTS.

Page

Introduction 1 79

The Problem 180

Material 181

Acknowledgments 181

General Considerations:

1. The Termites:

a. Castes 182

b. Food — Direct 184

c. Food — Indirect 185

2. The Protozoa 185

3. Biochemical Considerations 186

Relation of the Protozoa to their Host:

A. Historical:

1. True Parasites 188

2. Commensals 190

3. Symbionts 191

B. Original Observations:

Attempts to free termites of their protozoa 194

The incubation method 195

Table 1 197

Death of termites after incubation and the removal of the protozoa. 196

Cause of death 196

Incubation per se 196

Loss of protozoa (defaunation) 199

Table II 198

Effect of incubation on other insects 199

Sugars, starches, etc., are fed defaunatod termites in an effort to pre-
vent their death 200

* From the Department of Medical Zoology, School of Hygiene and Public
Health, Johns Hopkins University, Baltimore, Maryland.

This is the second of a series of papers on this subject. The first papri <>t th •
scries, "Correlation between the Food and Morphology of Termites aii'l the 1
ence of Intestinal Protozoa," deals with the more .ym'i.il aspe< ts oi thr subject,
and will appear in the July (1923) number of the Am,-ri,<in Jmtmnl «f

178

INTESTINAL PROTOZOA OF TERMITES. 1 79

When digested wood (humus) is fed defaunated termites, they do not

die ................................................... 200

Table III — : i various attempts to prevent the death of the

defaunated termites .................................... 201

INTRODUCTION*.

The mo-t unusual and peculiarly interesting faunal association
known to para-itolo-v i- that represented by the teemin- mena-
gerie of inte-tinal pmto/oa harbored by termites. These pro-
tozoa are favorite objects for study and have attracted the at-
tention "i many in\ ' •-! u ators. The initial microscopic \ iew ot
thou-aiid- of these organisms, but recently removed from the
termite's inte-tine, wrinkling, undulating, and twi-iin^ in a dmp
•Intion. ne\er tails to till the observer with enthusiasm,
and brin^ forth volleys of questions. Perhaps the tir-t thin_
atti i-ntion i- the abundance, both in form and number, ot

jiroto/o.i pn-eiit in a single host. The large abdomen is alnio-i
. ompletely tilled by the greatly distended intestine, the lumen of
which i cd by a vast horde of protozoa.

The protozoa of the termites, for the most part, differ
eraMy from those found in other animals, and this

• :ioi fail to attract the attention of a proto/oolo-i-t . Their
mo-t di-iiiH live features are: the ingestion of solid particle- of
\\ ..... 1 for food; the diversity of organization among the \arioii-
form-; tin- degree of structural specialization ' attained by many
of tin -, resulting in the formation of the nio-i com|»le\

il ^ell.iie oi-.iiielles yet known.

No observer lias ever failed to raise the question of the relation
of the proto/,,,1 to their host, but most investigator- in the p
ha\e regard* d the protozoa either as true parasites or commen-al-.
though, in almost every instance, no proof whate\er i- oft.
in support of either contention.

At piv-eiit no termite of the most highly >pe« iali/cd family,
the Termitid Metatermitida? Holmgren , i- known to harbor
inte-tinal pmto/oa; but all the genera and species of the oilier
familie-. Kalotermitida? (Pretermit ida- Holmgren . Khinoter-
mitid.r , Me-oiermitida? Holmgren) and Ma-tot, rmitida-, that
have been examined2 have been found to harbor enormous

1 N Lilly tin- papers of Kofoid ami S\\v/y (iQig >in« the iu.-un>-
niotnr -y-ti-i: :mf>iniula and Triilt«niitu<; tfrmit.

2 See tin- lu-t papri ni tl, latioii l>rt\vr<-n tin- I-' ..... 1 and Mor-

ISO L. R. CLEVELAND.

numbers of protozoa. It is not definitely known why protozoa
are never found in the intestine of any of the Termitidae; but a
difference in the feeding habit has been suggested as a very prob-
able explanation, since few, if any, of the Termitidae feed solely
on wood.

THE PROBLEM.

To determine the relation of the intestinal protozoa of termites
to their host.

Experimental methods were the paramount mode of attack;
and it was necessary, in order to solve the large problem stated
above, to solve several problems, which may be briefly stated as
follows:

1. What is the principal compound in wood, the sole article of
diet of most termites, which is used for food.

2. How do termites utilize cellulose, the principal compound
in wood, as their chief article of food?

3. How may all of the intestinal protozoa be killed without at
the same time injuring their host? To develop a rapid and suc-
cessful method for removing the protozoa from the termites was
a very difficult proposition.

4. Why do termites, when subjected to a temperature of 36°
C. for 24 hours, lose their ability to maintain themselves on a
wood diet?

5. Is clefaunation (the removal of the protozoan fauna),
which occurs when termites are incubated at 36° C. for 24 hours,
in any way responsible for the inability of the termites to main-
tain themselves on a wood diet?

6. Why do the incubated and defaunated termites live indef-
initely when fed the decomposition products of wood, or the
products of fungus-digested cellulose; but die within 10—20 days
when fed wood, their normal diet, or when fed cellulose?

7. Why do the incubated and defaunated termites regain
their ability to make use of wood or pure cellulose as food when
reinfected (refaunated) with protozoa?

8. Do the unincubated and faunated termites harbor any in-

phology of Termites and the Presence of Intestinal Protozoa," for a list of the
genera and species that have been examined. Since writing this more than 75
species of the Termitidae have been examined and three of this number have been
found to harbor protozoa and feed on wood.

INTESTINAL PROTOZOA OF TERMITES. l8l

testinal bacteria or fungi, capable of digesting cellulose, which
are killed off during incubation?

9. Do the protozoa aid their host mechanically to digest wood,
or do the protozoa digest the wood particles which they take
into their bodies and convert them into substances which their
host uses for food?

10. Sin<v the termite- lose their ability to digest wood when
the protozoa arc removed from their intestine, that is when they
are <1< -faunaied. but regain this ability when reinfecu-d with
protozoa, .md -ince t lie protozoa themselves can be -hown t<>
di-e-i tin- \\ood particles which they take into their bodies, does
a true -\mliiotic relationship exist between termites and their

inte-iinal proto/.

MATERIAL.

Ri .'/• nlili-rnn"i es Kollar, collected in Maryland, near

Baltim 1 for all the experimental work on this genus,

though main- colonies from other states have been u-ed for
-tnd\in^ tin- proio/oa of this species. Termopsis sp. (the species
nuild not )„• di it -i mined because there were no winded adults
in an\ of i lie (..lonies collected, but it is either a ngitsticollis II i.« n
or ni--tn/i->: . II .'ii) from Ashland, Oregon, owinu t<> it- 1
H/e. \\.t- used in many of the feeding experiment-, though
R< :i(-ulitc> cs was also used in all of the feeding e\peri-

nieni-. Kiitolcrmes schu'arsi Banks, K. jouteli Bank- and Pro-
rhinotfrnifs .v/w/>/r.v Hagen, from southern Florida, were used for
ineubation e\]iei iinents, but the material of these genera was not
-uffu ient in in. ike |>ossible as extensive a study as wa- i-arried out
\\itli the material from Maryland. Aniitermes sp., from I 'valde,
I :as, an.l .\,!Militertnes nwrio Latreille, from Port" Kii o, both
of the family Termitidio, were studied.

ACKNOWLEDGMENTS.

Manx thanks are due Drs. R. \Y. Hegner and \\'. 1 1. Taliaferro,
of the Pe]>artment of Medical Zo6log>', School of Hygiene .md
Public Health, Johns Hopkins Univer-ity, Bultinior. . M ryland,
for valuable -n^gestions and criticisms. I am al-o \t-r\ deepl\-
indebteil to Dr. T. E. Snyder, Specialist in Fore-t Insects, Bureau
of Kntomology, U. S. Department of Agriculture, Washington,
D. C., who identified all the termites studied and furnished all
the material except ReticnUtcrmcs flavipes and Nasutitennes morio.

182 L. R. C! EVELAXD.

GENERAL CONSIDERATIONS.

i. THE TERMITES.
a. Castes.

In order that the reader, unfamiliar with the nature of a ter-
mite colony, may know something of the castes present, the fol-
lowing condensed statement (taken from the papers of Snyder
and Thompson) is made. In Reticnlitermes flavipes there are
five castes, three of which are fertile or reproductive, and two
that are sterile or non-reproductive. Males and females occur
in each of the five castes. The reproductive castes are: (i) the
first form, which has three well-defined phases of development:

(a) the nymphs, with long wing pads, creamy white body 1.3-1.4
mm. long, and light brown eyes; (6) the winged adults, with
long wings, dark brown body 6 mm. long, and black eyes; (c] the
older males and females, with enlarged abdomens and the scales
of the shed wings, body 7-14 mm. long: (2) the second form,
which, like the first form, has three well defined phases of devel-
opment: (a) the nymphs, with short wing pads and colorless
body and eyes; (b) the young adults, with short scaly wing ves-
tiges, straw-colored or grayish body 6-7 mm. long; (c) the older
adults, with wing vestiges, enlarged abdomen, body length 7-12
mm: (3) the third form, which also has three developmental
phases: (a) the nymphs, wingless, with white head and body,
and eyes that are invisible in the living or unstained specimen;

(b) the young adults, wingless, head and body pure white, opaque
and not transparent, about 6 mm. long; (c} the older adults, with
enlarged abdomen, wingless, heiid and body white, 7-9 mm.
long. The other castes, the sterile castes, are: (4) the worker,
wingless with grayish abdomen, only two developmental phases,
i.e., nymphs and adults, always feeds on wood and there is no
degeneration of the jaw muscles, always harbors protozoa once
an infestation has been acquired, salivary glands small and very
little fatty tissue is present in the body; (5) soldier, wingless with
elongated head cove ml with thick yellowish chitin, mandibles
dark brown long slender and curved, abdomen shorter than in
other castes and more flattened, nymphs and adults only, no
post adult growth.

Thompson (1917) showed that the newly halt lu-<l nymphs of

INTESTINAL PROTOZOA OF TERMITES. 183

Reticulitermes flavipes, i.i mm. long, although externally all
alike, could be differentiated by their internal structures into two
distinct type-, namely, (a) reproductive nymphs, from which
develop the three fertile adult castes, and (b) the worker-soldier
nymphs, from which develop the two sterile adult castes. By
the time the- reproductive nymphs had attained a body length
of i.} 1.4 mm. tlu-y could be differentiated by their intern. il
character- into nymph- • »f the first form and nymphs of the sea >nd
form, which de\ eloped, finally, into the two re-pecti\e repro-
ducii\f adul' 5 'Her and worker nymph- could be dif-

ferentiated, internally. \\hen they had attained a body length
ot .V75 mm. Nymphs of the third form could not be differentiated
until a body length < .f 4 mm. was attained.

-•\rr.il intermediate forms have been discovered by various
in\i tors, intermediates between the first and second form-,

three fertile -oldiei- I l< ith, 1902)', and others; but the-e need
not be . on-ideied 1;.

Tin- ti\ •' cur in most termites. Known exception- are

as fol]o\\-: i In- third form reproductive caste occurs in k-w, if
. of iht I '( Tiniti'l i •; in fact none have been reported trom the
Nearcti< gem this family, and it is doubtful it" thi- ca-tr

urs in an\ i^enus of the Termitid;e. Apparently thi- caste
ha- ln-t-n lo-i in this, the most highly specialized, famiK of ter-
mite-. The third lOmi is less common than the first .m<l -ecoiid
in Reticuli Jl wipes, but in Termopsis it is more c..mmon

than the tir-t .md second form. Reticulitermes is about midway
bet\\eeu the most primitive and the most highl\- >p» •• iali/ed
termite-. / \ •> n:n/'\is is a primitive genus. A true \\oiker caste
i- not pie-cut in Termopsis and Neotermes, but a lar^e-hc.nled
\\oikir-like rejiroductive form is present. Two genera. Kalo-
ternifs and ('ryptotermes, have no worker caste. The ^enu-
. 1 it/>f>.'' has no soldier caste.

Grassi calls the first forms "true" or "perfect insects," or

1 \\ 'Social Life among the Insects. Sci. D M nthly. 16, 1923, p. i

.md workers are normally strrilr. hut soim-timus tlu-y become

intiii- .iu.1 pp'l'ulily on such occasions n-i>r..(li:«- thfir n\vn > Snj-der

(personal communication) soys Heath now admit-: IK- \va> \\rniig. Also experi-
ments \\hiih 1 have conducted recently reveal the fact that the soldiers of at least
:.il, .i;id pmhaMy all. \>'I«>phagus termites cannot live to themselves

vvlu-n K'VI 'ii thi-ir iiuimal diet <if wood.

184 L. R. CLEVELAND.

"the royal forms." The second and third forms he calls "sub-
stitute" and "complemental " forms, \vhich forms he thinks are
always ready to take the place of the royal forms in case of need.
He recognizes nymphs of the first and second forms, but these,
he thinks, if taken young enough and given the proper food,
may be changed into anything else, i.e., any of the castes. But,
Thompson (1919 and 1922) and Snyder (1920) have presented
valuable experimental evidence in support of the view that each
of the three reproductive castes breeds true, that is, produces
reproductive adults only like itself.1

b. Food — Direct.

Wood is a direct food of most termites. Some of them feed on
dry heartwood, while others feed on moist and more decayed
wood, usually sapwood. Kalotermes schwarzi, for instance, is
often found in furniture stored in an attic, since it does not re-
quire access to the soil in nature as does Reticulitermes flavipes.
However, it is quite probable that R. flavipes requires access to
the soil only in order to obtain moisture, since it may be kept in
captivity indefinitely without access to soil provided sufficient
moisture is given it. Water, of course, is a direct food, though
some termites require very little of it. Amitermes tiibiformans
Buckley, for example, feeds principally on dry cow chips and
soil, but occasionally eats the roots of grass and other vegetation.
The manure is often carried down into chambers below the
ground. Many of the Termitida? feed on plants, sometimes at-
tacking flowers, shrubs and field crops. Fungus gardens are
cultivated by many speices of this family. These termites carry
large quantities of vegetation into their huge nests and take
advantage of the digestive action of the fungus on the hemicellu-
lose, the fungus and hemicellulose being eaten by them as food.
It is interesting to note that leaf-cutting ants (CEcodomd] do the
same thing. They carry thousands of leaves to their nest, where
they use them to make a compost heap, on which fungi, to which
the ants are very partial, growr. Termites in captivity may be
induced to eat almost anything, but our knowledge reg.inlm^

1 Each of the adult sexual forms can reproduce itself and the forms below it in
rank. Thus, the first form adults reproduce all castes: the second form individ-
uals like themselves, third form, soldiers and workers; the third i»im indivi<l
like themselves, soldiers and workers.

INTESTINAL PROTOZOA OF TERMITES. 185

their habits in nature, especially the Termitidae, is so limited
that it is impossible to adequately discuss their food.

c. Food — Indirect.

The liquid diet is either a salivary secretion or regurgitated
material from the mid-gut. It is very probably the former. It
is a colorless and distinctly alkaline liquid which collects on the
labrum as a small drop, which may be used either as food for
others or for the termite secreting it. Many termites also employ
this liquid in building. In Reticulitermes flavipes the young
larvae of all the castes and the second and third forms during the
third developmental stage feed solely on this liquid diet which
they receive from other members of the colony. Later, the larvae
of all the castes eat the excreta as it is passed from the anus of
the older members of the colony, and it is at this time and in this
way that they first become infected with protozoa. The larvae
also begin to eat wood at or about the same time that they begin
to eat excrement — which in this termite is composed of a few
wood particles, but primarily of protozoa and liquid. The
soldiers, workers and first form reproductive males and females
always feed on wood, though they — especially the first forms
during the third developmental stage — partake of a liquid diet
as well. The exuvue are usually eaten and, sometimes, the dead
bodies of other members of the colony are eaten.

2. TIIK PROTOZOA.

Lespes (1856) made the first recorded observation of termite
protozoa, though he did not describe them. Leidy (1877 and
1881) was the first in America and the second in the world to
observe these organisms. In 1881 he gave descriptions of the
protozoa which he had observed in Reticulitermes flavipes,
creating three genera for them, Trichonympha, Pyrsonympha and
Dinenympha. This termite really harbors several other genera,
must of which were included in Leidy's drawings as stages in
tin- life cycle of the three genera mentioned above. More than a
><<>re of other investigators have studied these protozoa, and to
d.ite 41 genera have been described. For a complete account of
the work done by all the students of termite protozoa and for a
classification of all the protozoa which have been described from

L. R. CLEVELAND.

termites the reader is referred to the first article of this series,
"Correlation between the Food and Morphology of Termites
and the Presence of Intestinal Protozoa."

All of the known termite protozoa, except Gregarina termitis
and Nyctotherus termitis, have been placed in two orders of flagel-
lates, the Polymastigina and the Hypermastigina, the latter being
the most highly specialized order of flagellate protozoa.

3. BIOCHEMICAL CONSIDERATIONS.

Dore (1920) gives the following analysis of yellow pine (Pinus
ponderosa) oven-dry (100° C.) basis. Results in percentages.

Benzene extract 2.22 Lignin 29.47

Alcohol extract . 1.49 Xylan . . 3.49

Cellulose 57-72 Mannan 6.37

Galactan 0.78

The two following examples are taken from Ritter and Fleck
(1922), who have analyzed a large number of American woods:

Analysis of a mean of four samples. Results expressed in
percentages of oven-dry (105° C.) material.

Western Yellow Pine Tanbark Oak

(Pinus ponderosa). (Qtiercus densi flora).

Moisture 6.42 .... ... 3.66

Ash ... 0.46 .. ... 0.83

Solubility in cold water 4-°9 4.10

Solubility in hot water 5.05 5.60

Solubility in ether 8.52 0.80

Solubility in i % NaOH 20.30 23.96

Acetic acid i .09 5.23

Methoxy. . .. 4.49.. . 5.74

Pentosan 7.35 19-59

Mfthyl pentosan 1.62 none

Lignin 26.65 24.85

Cellulose 57.41 58.03

In the cellulose:

Pentosan 6.82 22.82

Methyl pentosan i .98 none

/l//>//a-cellulose 62.10 56.77

ZJe/a-cellulose 10.56 19.92

Cawwa-cellulose 3°-i3 23.93

That the principal food of most termites is cellulose is a fact
which appears to have escaped the attention of physiologists and
biochemists. Because of this, and also owing to the fact that our

INTESTINAL PROTOZOA OF TERMITES. 1 87

r -books of chemistry and physiology give little, if any, of the
information which has been gained in recent researches on the
chemistry' and digestion of cellulose, it seems expedient to briefly
mention — with no pretention at a discussion — some of the salient
features brought out by a few of the more important of these
researches.

According to Hibbert (1921) "cellulose is nothing more than a
polmerized dextrose glucoside of dextrose," and Irvin and Hirst
(1922) in their most recent paper on the constitution of poly-
saccharides have shown this to be true. They state that "the
average yield from the polysaccharide (e.g., cellulose) to the
hexose (e.g., glucose) is thus 95.5 per cent, of the theoretical
amount. Considering the standard of purity in which the mixed
methylglucosides were isolated, there can be no further doubt
that cotton cellulose is composed entirely of glucose residues."

1'ringsheim (1912) observed cellobiose — a disaccharide which
bears the same relation to cellulose that maltose does to starch
and which spilts into two molecules of glucose (Fischer and Zem-
plen, 1904 and 1910), when acted upon by cellobiase — as an in-
termediate product of bacterial cellulose digestion. The exact
mode of linking together of the two glucose residues in cellobiose
was determined by Haworth and Hirst (1921). Groenewege
(1920) showed that the bacterial digestion of cellulose is a hydro-
lytic procoss.

Celluluse is known to be produced by various moulds, soil
fungi, pathogenic fungi, actinomycetes, soil and intestinal bacteria.
Cellase is to be distinguished from cellulase in that it acts on the
hemicelluloses and is not an endo-enzyme (Pringsheim, 1912).

Some rather divergent views have been expressed regarding
the products of cellulose digestion ; different investigators working
with different cellulose-decomposing organisms have gotten dif-
ferent results. But according to Cross and Doree (1922) most of
these differences may be accounted for. The typical result of
i he digestion of wood cellulose is about as follows: Cellobiose,
ij 1 1 cose, acetic acid, lactic acid, butyric acid, alcohol, CO2, H2,
CH4, CO. l'nder excessive aeration alcohol and acetic acid are
i-i'inpletely oxidized to CO2 and H2O. The gas liberated is never
I >ii re CO2; there is either H2 or CH4, or a mixture of the two.
Alcohol is never obtained without acetic acid. \Yuksman (19--

188 L. R. CLEVELAND.

is of the opinion that some soil bacteria cause CH4 fermentation
of cellulose and that others cause H2 fermentation. McBeth
(1913) thinks cellulose-fermenting organisms are not responsible
for the production of gases. Gas, he thinks, is produced by con-
taminating organisms.

Abstracting Klason's remarkable researches on lignin, Cross
and Doree (1922) state: ".Beto-lignin (CJ9 HIS O9, and that part
of lignin precipitated by arylamine bases is designated as acro-
lein-lignin or alpha-ligmn. The other part, which is not precipi-
tated, appears to be a carboxyl group and is called carboxyl
lignin or fo/a-lignin) must therefore be bound to cellulose and
probably to a//>/?a-lignin (C22H22O7). Lignin cannot well be
assumed as a secondary product derived from cellulose, but
appears as a direct assimilation product of CO2 and H2O or for-
maldehyde. Therefore the formation of lignin is a function of
chlorophyll." Klason thinks lignin is possibly present in wood
as a glucoside and that it may be built up from pentose.

THE RELATION OF THE PROTOZOA TO THEIR HOST.

A. HISTORICAL.

i. True Parasites.

Most students of termites, and termite protozoa, have focused
their attention on the morphology 1 and behavior of either the
termites or the protozoa which they harbor; consequently, the
interrelation of host and parasite has been very little investigated.

Grassi is really the only investigator who has not confined his
attention almost exclusively to a study of either the termites or
their intestinal protozoa. He has studied the termites and the
protozoa, but has never carefully investigated their relationship.
His work on the two groups of organisms, termites and termite
protozoa, has been largely a study of two separate and distinct
problems; namely, the morphology and systematics of the pro-
tozoa, and the physiology and behavior of the termites, with
special emphasis on the underlying factors in caste production.

Grassi and Sandias (1893) think that the relative abundance
of protozoa present in the various adult individuals in a termite

1 The morphology of termite protozoa is a fascinating subject and occupied the
writer's entire attention for a year before the present investigation began. He will
publish in a later paper a description of several new forms.

INTESTINAL PROTOZOA OF TERMITES. 189

colony is inversely proportional to the degree of gonad develop-
ment of their host. According to these investigators, the pro-
tozoa, with a small amount of saliva, function in the develop-
ment of soldiers and workers. Why the same thing (many
protozoa and a small amount of saliva) functions in the produc-
tion of two separate and distinct castes, they do not consider.
They observed that the protozoa were scarce in the winged sexual
forms, and usually totally absent in the neoteinic l individuals.
When they found protozoa abundant in the neoteinic indivi-
duals, the gonad development was imperfect. They remarked
that the developing neoteinic forms were fed largely on a salivary
diet, which they thought killed the protozoa, though it is more
probable that the protozoa died because their host no Ion.' r
furnished them with wood for food. Then, according to Grassi
and Sandias, it is impossible to say whether the maturation of
the gonads is clue solely to the change from a wood to a salivary
diet or to the absence of protozoa. In this paper they reach no
definite conclusion, though they do not believe the presence or
absence of protozoa a sufficient cause to control gonad develop-
ment, for they say: "It is a moot point whether the maturation
of the generative organs is due solely to the saliva or to the ab-
sence of protozoa as well; but the latter, as my- previous re-
marks show, is not by itself a sufficient cause." Then in the
next paragraph," I have frequently asked myself whether the
protozoa have not an important digestive function, since the
comminuted wo(xl passes almost entirely through their bodies.
It is probable, but not proved."

Hrunelli (1905) followed up the idea of Grassi and Sandias, and
states that in the queens of two genera harboring protozoa,
Kalolermes flavicollis and Reticiditcrmes lucifugtis, there is a
correlated destruction of the oocytes — a kind of indirect parasitic

1 The term neoteinia was first introduced by Camerano (Bull. Soc. Ent. Ital,.
. pp. 89-94) to note the persistence during adult life of part or all of the char-
;uuti-iu-s normally peculiar to the immature, growing, or larval stages. Grassi
and Sandias probably refer to the second and third form reproductive castes since
it has been shown by Thompson (1920) that the jaw muscles of these forms de-
rate to the extent that they cannot eat wood and must be fed on a fluid diet;
and, of course, we should not expect them to harbor intestinal protozoa when they
ili> not eat wood.

: Grassi wrote most of the paper. The part written by Sandias is placed between

I9O L. R. CLEVELAND.

castration ("castration parasitaire"). Feutand (1912) says
what Brunelli regards as signs of degeneration is in reality nothing
more than an alteration of tissue brought about by the histo-
logical reagents.

Grassi, nineteen years later (1911), collaborating with Foa, is
of the opinion that the protozoa do not aid their host in the
digestion of food, and he points out that many wood feeding
insects do not harbor protozoa. He states l that several workers
(species not mentioned) of Kalotermes ~ were placed in small
boxes with wood and incubated at about (circa) 35° C. It is
interesting to note that none of the protozoa harbored by Kalo-
termes schwarzi Banks, from Miami Beach, Florida, are killed
when incubated for three days at 35° C. How long Grassi's
species of Kalotermes was incubated is not known, for he does not
mention this in his paper. But he does state that the incubation
killed off nearly all the protozoa, all of the large forms being
killed off every time. These termites, he states, lived in full
activity for several months.

Recently Jucci (1920 and 1921), in a preliminary paper re-
ported before the Academy of Lincei, has announced the dis-
covery of a particular diet which brings about caste production.
But the carefully controlled breeding experiments and the numer-
ous field observations of Thompson and Snyder (1919, 1920, etc.)
make it highly probable that at least three of the castes, namely,
the first, the second and the third form reproductive males and
females, only give rise to fertile reproductive males and females
like themselves and to infertile males and females of the soldier
and worker castes.

2. Commensals.

If these entozoic protozoa benefit from their association with
the termites, and by so doing do not harm or benefit their host,
they are commensals. They may feed from the waste food-

1 Per mio conto in questi ultimi anni ho cercato di decidere la questione in altro
modo: se si tengono ad una tcmperatura di circa 35° C. scatolette contencnti legno
picno di Calotermes, i Protozoi muoiono, talvolta tutti, piu spesso restanu in vita
soltanto le forme piccolo; si hanno cosi delle colonie di Calotermes senza Joenia e
Mesojoenia, e talvolta con molti individui anche del tutto privi di Protozoi. lo le
tengo in vita prospera da parecchi mesi; perci6 ritengo che i Calotermes possano
digerirc il legno anche senza gli speciali Protozoi (Joenia e Mesojoenia).

2 Also spelled Calotermes

INTESTINAL PROTOZOA OF TERMITES.

products and bacteria in the termite's intestine, or they may be
food-robbers and feed from the same table as their host ; in either
case, they are commensals, provided the host's table always
contains sufficient food for both host and parasite. Kofoid and
Swezy (1919) are the principal exponents of the commensalistic
relationship of termite protozoa. They remark: "One of the
most curious and unique faunal associations to be found among
the parasitic l protozoa is the group parasitic ' or commensal in
the intestinal tract of the social termites. . . . This distension
is caused by the vast numbers of parasitic and commensal pro-
tozoans which fill the lumen of the intestine. When this is
opened a thick milky fluid exudes. Under the lens this is found
to be composed of great quantities of these small forms, thickly
massed together, along with fragments of wood upon which the
host, as well as some of the commensals, feeds."

T>. Symbionts.

hums (1919) in studying the termite Archotermopsis li'rongh-
toni Desneux made sections of twelve worker-like females and
found numerous protozoa and nearly full sized ova ready for
fertilization, and in no instance did he observe even an indication
of oocyte degeneration. He also found well developed ovaries
in live soldiers, whose hind-intestines were filled with protozoa,
hums suggests that the presence of protozoa may be correlated
with the wood feeding habit of the termites. In the young
larva-, and in the queens and kings, which receive the fluid diet
from the other members of the colony, he found, as a rule, no
protozoa. In the older larva?, soldiers, workers, and sometimes
in the winged sexual forms,2 protozoa are abundant and a wood
diet is the rule.

1 This term, loosely used, means any organism which lives in or on another
organism.

2 It is interesting here to note the statement of Snyder (1920), regarding one of
hi- cross-breeding experiments. He says, "A large series of young second form
l. in. il.- adults of Relifuliternifs flavipes, which possibly may have been fertilized by

;id form males, were taken from a fairly small colony and placed with mature
lii-t form dealated males, which had not copulated, in small shallow cells in de-

I wood sunken in moist sand in glass jars and tin boxes.

After a period of ten days to two weeks all these second form females had died,
but the first form males were still living and were active; they evidently were pre-
l-.iii-d to forage for themselves, whereas the second form adults needed the care
and nourishment usually afforded them by the workers. Possibly the jaw muscles

IQ2 L. R. CLEVELAND.

The wood, according to Imms, after being crushed by the
mandibles and later by the gizzard, passes rapidly to the hind-gut
without undergoing any chemical change. The pyloric valve
prevents the readmission of wood particles into the mid-gut. To
quote Imms: "On reaching the hind-intestine the wood is in a
condition of minute fragments and particles, and the greater
bulk of it gradually becomes taken up and absorbed by the numer-
ous intestinal protozoa. Within the protoplasm of these organ-
isms the woody material undergoes chemical changes, and, when
ejected from the bodies of the protozoa, much of it is in a condi-
tion capable of being assimilated as food by the host termite.
A significant fact is that lignous particles are not subjected to the
action of the secretions of the mid-gut."

Imms further points out that, in his opinion, the experiments
of Grassi and Foa (1911), where these authors claim to have kept
Kalotermes (species not given, nor is the number of termites)
alive and active for several months with all but a few of the in-
testinal protozoa killed off by incubation at about 35° C., has
no value, for during incubation the termites very probably sub-
sisted on the large amount of fecal matter l which had accumu-
lated in the jars where they had been kept previous to incubation.

Imms, then, through an ingenious method of reasoning, reaches
the conclusion that the protozoa are symbionts, a conception
already reached by Buscalioni and Comes (1910), who claim, by
means of various microchemical tests, to have shown that the
wood is digested inside the bodies of the protozoa ; and since the
protozoa digest the wood, these authors conclude, that they are
symbionts. 2 They state that Trichonympha agilis, harbored by
Reticulitermes lucifugus, when treated with iodine dissolved in
iodide of potassium , gives a characteristic glycogenic reaction in
a region near the nucleus, and that this reacting region is sharply

in these young second form adults had begun to degenerate through disuse in masti-
cating wood. It is believed that second form males — like the females — would have
succumbed, without the workers." First form adults, so far as my experience
goes, harbor protozoa, but the second forms do not. This probably explains the
inability of the second form adults to live without workers, that is in the event the
jaw muscles permitted them to eat wood. Of course the protozoa are lost, possibly,
because the host docs not eat wood. I have experiments started which I hope will
determine this point.

1 The fecal matter contains sugars and possibly other foods.

5 A discussion of this may be found on pages 193-4.

INTESTINAL PROTOZOA OF TERMITES. 1 93

defined from the rest of the body. Cutler (1921) could not locate
a definite glycogen reacting area in Pseudotrichonympha pristina,
harbored by Archotermopsis wroughtoni, but, on the contrary,
found that the glycogenic reaction was diffused through the en-
tire organism. Grassi and Foa (1911) state that they in part
attempted to verify the work of Buscalioni and Comes but that
their results did not justify the drawing of any conclusions.

Dobell (1921) says: "'When the association benefits both part-
ies, the condition is one of symbiosis — a not very frequent state
in nature. An example is afforded by some of the flagellates
living in termites. In return for the food and lodging which the
termite gives to the flagellate, the latter helps the former digest
its own food." Regarding this statement Dobell (I922)1 says:
"It is merely a general statement of my own personal view. It
is not based on any particular experiments which I have made,
but is an inference from what I myself have observed and others
have recorded. For many reasons I consider it probable that the
flagellates of termites aid their host in digesting the wood on which
they feed. The observations of Buscalioni and Comes (1910)
and of Imms (1919) especially seem to me to justify such a con-
clusion.

"However I am quite ready to admit that the matter has not
been definitely settled beyond all question. In fact I recently
discussed the problem with I >r. Koid/umi and we planned
several experiments to determine the question. . . ."

/•{. OKK.INAI. OBSERVATIONS.

At best some of the experiments of Buscalioni and Comes (1910)
were conducted in a rather crude manner and, in addition to this,
unwarranted conclusions were drawn due, often, to misinterpre-
tation of the results obtained. The only conclusion which may be
drawn from the results of their experiments is, that the protozoa
digest the wood which they take into their bodies. But the fact
that the protozoa digest the wood in no way hinders the termite
from doing the same thing. Many animals ingest more food than
they can use. The protozoa may feed from the wood particles
in the termite's intestine, as one would naturally expect them to
do, since nutrition with most of them appears to be holozoic.

1 Personal communication.

194 L- R- CLEVELAND.

In order to establish the fact that the protozoa are symbionts,
rather than commensals (feeding from the same table as their
host), it must be shown that they actually aid their host in some
way by doing something for the host which the host itself cannot
do. The experiments of Buscalioni and Comes do not in any way
fulfill this condition. They have only shown that the protozoa
digest wood particles. The ability or inability of the termites to
do the same thing was not studied. That the termites aid the
protozoa, by furnishing them food and lodging, is obviously
demonstrated by the fact that the protozoa, so far as known,
cannot live outside of termites. But wre cannot call the protozoa
symbionts, until their ability to digest wood has been shown to
be a necessary aid to the termite.

Realizing the difficulty of determining the relation of the
protozoa to the termites so long as the two were associated to-
gether, several experiments were begun, with the hope of effect-
ing a method for removing the protozoa from their host, and at
the same time leave the host uninjured. It is needless to say that
hundreds of experiments — some of which are given in Table I.—
met with failure, for the host was either killed or severely in-
jured in an effort to remove the protozoa from it.

When wood, previously soaked in a 5 per cent, aqueous solu-
tion of either sodium, potassium or calcium chloride, was fed to
a colony of Reticulitermes flavipes (elsewhere in this paper — unless
it is otherwise evident — when the words "termite," "termites"
and "host" are used they refer particularly to this termite)
harboring protozoa, within four to five days an occasional ter-
mite, free, or almost free, of protozoa was found, when many
individuals from the colony wrere examined. The number free of
protozoa increased as time progressed but, before all the termites
in the colony had lost their protozoa, it was quite evident that the
termites themselves were not normal. They had in some \vay
been affected ; whether by a direct action of the chemicals or by a
gradual loss of the protozoa, it is impossible to say from the data
available.

About this time a much more promising method for removing
the protozoa was evolved; namely, the incubation method, and
the experiments employing the use of salts, etc. (see Table I.),
to remove the protozoa, were discontinued. It might be well to

INTESTINAL PROTOZOA OF TERMITES.

mention here that, because of the success of the incubation
method, many of the other attempts to free termites of their
intestinal protozoa were, perhaps, not given a fair trial. If
different dilutions of some of the chemicals had been used, the
results, possibly, would have been more promising. Had X-ray
and ultra-violet light been used more extensively, excellent re-
sults might have been obtained.

It was found that all — not one was left alive — of the protozoa
could be removed from the intestine of ReticuUtermes flavipes by
incubating them for 24 hours at 36° C. This was a much more
rapid method than the feeding of salts, as described above, where
several days are required to remove the protozoa from all the
individuals in a colony and, besides, the termites after incubation
seemed to be perfectly normal in every way. All other attempts
to remove the protozoa were now abandoned in favor of the in-
cubation method, which may be briefly described as follows:
Sometimes slumps and large pieces of wood, containing thousands
of termites, were brought to the laboratory without molesting
or disturbing the colony. The stumps and large pieces of wood
were placed in glass jars and incubated for 24 hours at 36° C.
At the end of the incubation period the jars were taken from the
incubators and, after many termites from each jar had been
carefully examined to be certain that the protozoa had all been
killed, were left at .room temperature, which was usually about
2o°C. Sometimes the stumps and logs were split before they
were brought to the laboratory and the termites were taken from
them and placed, with a good quantity of wood, in glass jars with
metal tops. These jars were placed in the incubator as soon as
they reached the laboratory. In other experiments, where ter-
mites were counted, small shell vials with cork stoppers were
used, and from 10-100 termites were placed in each vial, and
incubated in the same way that those in the large jars were.
Several vials of termites without wood, and several with pure
rellulose ' (Whatman's filter paper) only, were incubated. Other
experiments in incubation were carried out with results as shown
in Table I. One very significant fact brought out by these ex-

1 This paper was made by \V. & R. Balston, Ltd.. and contains .0004 grams of
;i-li per circle of 55 mm. diameter. When "pure cellulose" occurs in this paper it
iners to this grade of the genuine Whatman filter paper.

196 L. R. CLEVELAND.

periments is that the thermal death point of the protozoa is
3-4° C. lower than that of the insect host.1

After incubation the termites were closely observed to note if
they at any time became abnormal; to prevent the growth of
moulds in the jars, for moulds are very destructive to termites;
to be certain that the jars contained neither too much nor too
little moisture — a thing that can only be learned from long ex-
perience in keeping large numbers of termites in captivity and an
intimate knowledge of their habits in nature. Every known pre-
caution was used to keep the termites during incubation in ex-
actly the same environment, except temperature, as the unincu-
bated controls. After incubation the incubated and the unin-
cubated termites were kept under identical conditions, as re-
gards food, temperature, moisture, light, etc.

The termites become abnormal in from five to fifteen days
after they have been removed from the incubator. They are less
active and their abdomens, if carefully examined, may be seen to
be smaller and slightly flattened. The length of time required
for abnormalities to appear depends no the kind of wood fed
after incubation, the more decayed the wood the longer it is
before any abnormalities appear. In two to five days after this
first symptom is noticed the abdomen becomes still more flattened.
Soon it is very much smaller and almost flat, and two or three
days later the animal can scarcely be made to move at all, for
only a modicum of vitality remains. Death occurs, on the aver-
rage, from 10-20 days after incubation. It occurs, as shown in
Table II., in some instances in less than ten days and in a few
cases after twenty days, the longest being 27 days. In the less-
decayed wood, as mentioned above, death occurred early, and
in the most decayed, it occurred late. But why did it occur at all?
Why did the incubated termites die?

The death of the incubated termites — sterile, uninfected,
uninfested and defaunated, so far as protozoa are concerned-
may be due to the change in temperature, i.e., the incubation
per se; it may be due to the killing off of all the intestinal pro-
tozoa; or it may be due to the killing of intestinal bacteria and

1 Termopsis sp. from Ashland, Oregon, is killed at a lower temperature. It is
killed in less than 24 hours at a temperature of 35.5 degrees C. Work on the T. D. P.
of termites from various parts of the world is now in progress and the results will be
published in a later paper.

INTESTINAL PROTOZOA OF TERMITES.

197

fungi, which, possibly, in some way aid their host in the digestion
of wood and pure cellulose.

TABLE I.

METHODS AND RESULTS OF VARIOUS ATTEMPTS TO FREE TERMITES OF THEIR

INTESTINAL PROTOZOA.

u «; w

X * '- ~ Z •-

W I'Ctllt C

E L ~

C. - f ^- ~ x

!^- i

Wood,' 5% N'aCl . .

Wood.' 5%CaCli..
• 1.' 5% KCI....
1,

30

10

20
20

500

IOO

500
500

Kills protozoa in 10-15 days, but host
is almost dead by this time
Same result as with NaCl
Same result as with Nad
Host killed in Dil. of 1/15,000

Wood <'-<)«

5

IOO

Killed host

<juinine sulphate. . .
In iiia bark . . .

2
2

IOO

too

Killed host
Killed host

• ian violet

15

Dil. i /i 00,000, host and protozoa un-

; fill ll-itl

IO

IOO

affected
Dil. i '100,000, host and protozoa un-

Anhydrous CaClj*. .
Cone. HjSCV

6

5

500
2OO

affected
Many protozoa killed, host injured
Many protozoa killed, host injured

1 1' 'MM I '* IH'St

5

2.OOO

Host ancl protozoa unaffected in 2 mos.

1 1 urnus

IO

IO.OOO

Host and protozoa unaffected in 3 mos.

Soil

5

4.000

Host and protozoa unaffected in 2 mos.

Soap

3

IOO

Host killed

Leather

3

Host and protozoa unaffected in 3

Low temperature. . .

•. .ti inn

5
20

2OO

I .000

weeks
2-3° C.. no effect in 3 weeks
Manv protozoa killed, host injured

4

IOO

I min. no effect on protozoa or host

1- nines of CSH»OH.
..•s of CHCU .. .
I V. Light. 10 in. . .
y, -• -kin dos. . .
In. ' J5 '.5 (\ 24 hrs.
Inc4 36° C. 24 hrs.. .
Inc4 42° C. 10 min.
Inc4 43° C. jo min.
liu ' 44° C. 10 min.
In. ' 45° C. 10 min.
Inc4 .Vi* C. 10 min.
In. ' .jr: C. 10 min.
In. « 4.x-"1 C. 10 min.

4

2
10
IO

15

1

IO
IO
10
IO
IO
10
IO

2OO
5°

500
500
5.OOO
I OO.OOO
5OO
500
500
500
500
5OO
5OO

2 min. no effect on protozoa or host
1/2 min. no effect on protozoa or host
15 min. no effect on protozoa or host
N'o effect on protozoa or host
Few protozoa killed, host uninjured
All protozoa killed, host uninjured
Few protozoa killed, host uninjured
All protozoa killed, host uninjured
Host not killed
Host not killed
Host not killed
Host nearly all kill.-<l
Host all kil'led, 48° C. is T. D. P. of host

1 WIMII! moistened with 5 per cent. Aq. solution of NaCl, K lj and with

in.iny dilutions of CuSO4 and HgClj

- fhese chemicals were used in dehydration experiments.

3 More than 2,000 experiments were carried out.

4 Incubated.

L. R. CLEVELAND.

Kalotermes schwarzi, K. jouteli and Prorhinotermes simplex
when incubated for 24 hours at 36° C. are defaunated, but enough
material was not available to carry out extensive experiments
on these forms as was done with Reticulitermes flavipes.

TABLE II.

SHOWING THE LENGTH OF LIFE OF INCUBATED AND NON-INCUBATED TERMITES
WHEN FED THEIR NORMAL DlET OF WOOD. 30 EXPERIMENTS
SELECTED AT RANDOM FROM A SERIES OF 500.

J

"5 c <-i

Remarks.

g

. •- c

^

o _ o

E

*~7 "^ P
f~ < c>

c

b*

^ *"*

.= •- .

~

Treatment.

cd S"

«£ £

C •-; rt

w

.i £ W

°i ""' Q

^r

x ™ —

" "3

d

yi

107

Incubated

4,000

14

207

L'nincubated

2,000

60

Jar became too dry due to neglect

11015

Incubated

500

i?

7015

L'nincubated

i, 200

In.1

Alive now, 9 months later

12015

Incubated

400

15

8015

L'nincubated

I.OOO

In.1

Alive now, 9 months later

13017

Incubated

300

20

14017

Incubated

500

20

15017

Incubated

400

21

1017

L'nincubated

700

In.1

Used in other expts. after 6

months

2017

L'nincubated

2,000

In.1

Alive now, 8 months later

2027

Incubated

300

18

2627

Incubated

2OO

21

3627

L'nincubated

7OO

In.1

Alive now, 5 months later

3027

L'nincubated

500

In.1

Alive now, 5 months later

id27

Incubated

40O

7

Mould caused early death

2C27

Incubated

6OO

15

1028

Incubated

2OO

10

2028

Incubated

IOO

14

3028

LTnincubated

20O

45

Used in other experiments

1031

LTnincubated

I.SOO

In.1

Alive now, 4 months later

2631

Incubated

50O

12

3631

Incubated

7OO

18

2031

Incubated

I.OOO

8

Wood decayed very little

4031

Incubated

2.OOO

10

Wood decayed very little

1^32

L'nincubated

2.OOO

In.1

Used in other expts. after 3

months

2C32

Incubated

I,5OO

27

Wood very much decayed

ld32

L'nincubated

3,000

In.1

Used in other expts. after 3

months

2d32

Incubated

2,000

20

1/33

L'nincubated

I.OOO

In.1

Alive now, 4 months later

To be certain there was no error in the experiments which
had been done up to this time, more than live hundred jars of
termites were collected; some were incubated and some were not
incubated, as shown in Table II. No doubt 100,000 termites

1 Indefinitely.

INTESTINAL PROTOZOA OF TERMITES. IQ9

were used in these experiments and, in every case, death occurred,
as in the preliminary experiments, within 10-20 days on the aver-
age after incubation. In no case did the unincubated and faun-
ated controls die. In fact, they are still alive now, more than six
months since many of the experiments were completed.

How may the cause of the incubated and defaunated (with
the protozoan fauna removed) termites' death be determined?
Is death a direct resit of the incubation? That is, if the termites
did not harbor intestinal protozoa which are killed off by the
incubation, would death result in 10-20 days? In other words,
is the death of the protozoa harbored by the termite in any way
connected with or responsible for the death of their host; or is it
an entirely independent phenomenon having nothing at all to
do with the termite's death. In an effort to answer these ques-
tions many experiments were carried out, but only those experi-
ments whose results throw some light on the questions involved
will be mentioned and discussed.

Insects, like the cockroach, rather closely related to the ter-
mites morphologically, and many wood-boring beetles, rather
closely related to termites in habits, were incubated at the same
temperature for the same length of time, and for a longer time,
as were the termites. These were carefully observed for two
months after incubation and no abnormalities were ever observed
in any of them. It seems, then, that this temperature is not
detrimental to these insects. But the fact that a cockroach or a
wood-boring beetle is not killed by the incubation method em-
ployed in killing the protozoa of termites, does not show that the
termites themselves were not killed directly by incubation,
because these insects do not harbor a vast multitude of intestinal
protozoa that are suddenly killed off, as do the termites; and
besides, they are, after all, quite different in some respects from
termites.

Nasuliternies morio Latreille, from Porto Rico, one of the Ter-
mitida-, and, of course, a termite which does not harbor intestinal
protozoa, was procured for experimental purposes, the idea being
tli.it if this termite could withstand the same incubation tempera-
ture that Reticiilitermes flavipes had been subjected to, and re-
crive no ill effects therefrom, as evidenced by its ability to live
indefinitely after incubation, the probability that Reticiilitermes

20O L. R. CLEVELAND.

flavipes was not killed directly by incubation was very much
augmented. Then, some cause other than the effect of the in-
creased temperature on the cells of the termite might be expected
to be responsible for the termite's death. But Nasutitermes
morio could not be kept in closed jars for a longer period than
5-10 days before death occured, regardless of whether they had
been incubated or not. Amitermes sp. from Uvalde, Texas,
another genus of the Termitidae, was obtained, but it also could
not be kept in closed jars. Other genera of the Termitidae,
termites not harboring protozoa in nature so far as known, which
might possibly be kept in closed jars — though nothing at all is
known regarding this — were not obtainable. It then became
necessary to answer in another way the question, does the heating
kill the termite directly?

An effort was made to keep the incubated and defaunated
termites alive by feeding them substances other than wood, their
normal diet. Dextrose, peptone and starch, separately and
altogether, were fed them; and it was evident, from the results
(see Table III.) obtained by the feeding of these substances,
particularly dextrose, that the lives of the incubated and de-
faunated termites had been prolonged.

But before these experiments had progressed very far a much
better method of prolonging the lives of the defaunated termites
was discovered, and no more experiments employing dextrose,
starch and peptone were carried out. They were placed in glass
jars and were fed a diet of fresh humus, which seemed to be a
very satisfactory food, as the results tabulated in Table III.
show. When they were given humus before any of the charac-
teristic symptoms of wood-fed defaunated termites had appeared,
no such symptoms ever appeared. If these symptoms had ap-
peared before humus was fed, the termites after being fed humus
for three or four days became much more active, and within six
to seven days they were normal again. Even when all the indi-
viduals in a jar of defaunated termites were almost dead, due to
being fed on wood, many of the almost lifeless ones could be
brought back to normal by feeding them humus. The humus fed
defaunated termites appeared to be perfectly normal in every
way long after all of their wood fed defaunated controls were
dead. One hundred jars of defaunated termites were used in the

INTESTINAL PROTOZOA OF TERMITES.

2OI

TABLE III.

SHOWING THE RESULTS OF ATTEMPTS TO PROLONG THE LIVES OF INCUBATED

AND DEFALCATED TERMITES BY FEEDING THEM VARIOUS SUBSTANCES.

27 EXPERIMENTS SELECTED AT RANDOM FROM A SERIES OF 100.

6

4J

'" -n

Sub-tunce.

t 1

1.5

Remarks.

-

£. -

V

-

1 i.

2OO

I •

2OO

34

•

1 i.

2OO

.40

2OO

4

Di

IOO

30

I '• ;i.«

IOO

45

I'. ;..»

IOO

43

Dextrose

ISO

30

111 > ; i , ,se

2OO

34

I.OOO

'-5

Experiment discontinued'

•

HUD

I.OOO

45

Allowed to become too do-

•HUB

5OO

125

Experiment discontii.

•

I.OOO

60

Experiment discontinued

IIUS

I.OOO

Experiment discontinued

mis

I.OOO

Experiment discontinued

-

•HUB

1 ,000

Experiment discontinued

Humus

I.OOO

Changed to a wood i!

I.OOO

Changed to a wood d

Hurnus

I.OOO

2O

' ....: . • it' a v.

I.OOO

50

Experiment not done with care

•

2OO

45

Experiment not done with care

ct.1

50

40

Experiment discontinued

ct.1

So

45

Experiment discomi:

SO

34

Experiment discontinued

i i. r.»

50

60

Experiment discontinued

1 . 1 > C.»

So

60

Experiment dif J

I D. C.»

50

60

Experiment discontinued

luiniu- feeding experiments, and from ten to more than a th<iu>.uid
u-nnit. e placed in each jar. It does not seem feasible or

necessary to state when each of the one hundn-d j.irs \\viv col-
and incubated and just what happened to each of them,
tin- results of each experiment are -iniilar, as Table III.

-c and peptone.

s Intestinal contents of wood-boring insect la
1 Xo evident signs of death at this time.
4 Death resulted in 10-20 days later.
* Fungus digested cellulose.

(To be continued)

Vol.XLVI May, 1924 Xo. 5

BIOLOGICAL BULLETIN

THE PHYSIOLOGICAL AND SYMBIOTIC RELATION-

5HIPS BETWEEN Till: INTESTINAL PROTOZOA

M| IIKMITKS AND THKIR HOST,

\\ I I II SPECIAL REFERENCE TO

Rkl /< L'LITERMES FLA VIPES

KOLLAR.*

L. R CLEVELAND.
(Continued.)

CONTENTS.

.1 to ihcir Host:
:m (continued) :
•••c| termites die when transferred to thi-ir normal

•• !

' I - u K.ir « to ccllulottc

. • -i u I and cellulose

use lignin a» food?.

,'. fiM«l of protozoa-harboring termites

'lulose. when fed defaunated termites, prevci-

- cannot digest wood or cellulose

- die because they cannot digest their food

•Mcnts

ted with protozoa can digest wood

!>.ir!>or protozoa to be able to digest wood

•tozoa aid their host in the digestion of \v

I y

.; the wood for their host

t ing and non-wood ingesting protozoa -' i 1

•In- l>acteria and fungi of termites -M'>

'•ntiul to the life of their host -'7

on of termites and protozoa -' 1 7

-•uinm.iry

* From tin I > ; .irtment of Medical Zoology. - .ni'l I'ublic

li. John< Hopkin-; I'niversity. Baltimore. Maryland.

a series of papers on th: L Tlie rir.-t paprr »i the

•n l«-tween the Food and M'>rplK. !•>.!;>• r,t Ti-Tniti •- an-1 I
i lntr~tiii.il I'r.«tn/oa," deals with tin- moro >jfni-r.i tin- sill

and will upiM-.ir in tin- July Uyjj) number ol tin- 1 m '.mrnul of li

14 203

204 L- R- CLEVELAND.

After the incubated and defaunated termites had been fed
humus and kept alive for a longer period of time than they would
have lived had they been fed wood, as shown by the death of all
the wood fed defaunated controls; from time to time transfers
back to a wood diet were made, with the result, in each case, of
death in ten to twenty days, the usual length of life of wood-fed
defaunated termites. The last of the transfers back to wood was
made after the defaunated termites had been kept alive, active,
and apparently normal in every respect, on a humus diet for
three months. Death resulted, as was expected from the pre-
vious experiments, and two weeks after all the defaunated ter-
mites which had been transferred from the humus to the wood
diet were dead, the humus-fed defaunated termites were alive and
active, at which time the experiments were discontinued. That
progressive development took place is evidenced by the fact that
molting occurred in the humus-fed defaunated termites three
months after all the protozoa had been removed by incubation.
Deeply pigmented first form reproductive adults were formed in
several of the jars during December and January.

From these experiments it is obvious that the humus which was
fed to the defaunated termites is potent to prolong their life
four months — and probably indefinitely — beyond the death
point of such termites when fed a diet of wood. The unincubated
controls, which had been fed wood and kept in the laboratory
in the same way as the incubated and defaunated termites being
fed wood and the incubated and defaunated termites being fed
humus, were alive and active at the time the experiments were
discontinued. Now, since the incubated and defaunated ter-
mites live indefinitely when fed humus, but die in two to three
weeks when fed wood, and the unincubated and faunated termites
live indefinitely when fed wood, it is evident that the death of the
incubated and defaunated t;ermites is not due directly to the
incubation temperature, but to an acquired inability to make use
of wood as food. In other words, the termites after incubation
and the removal of the protozoa, are no longer able to maintain
themselves on their normal diet of wood, and death results. The
incubation has in some way incapacitated them as wood users.
They feed on the wood, as may easily be determined by watching
them or by examining their intestinal contents, but evidently are

INTESTINAL PROTOZOA OF TERMITES. 2O5

unable, for some reason, to utilize it — perhaps digest it, since
humus, which is decomposed or digested wood, will keep them
alive indefinitely.

When the protozoa are treated with a hydrocholric acid and
phloro-lucinol nilution a distinct pink color appears inside the
bodies »I the wood investing protozoa. Imms (1919) used this
reaction to prove th.it the particles which he saw inside the bodie-
of ihe protozoa were wood particles— "ligneous particles."
< )-l lima ' I'd') pi .i' i •<! phloroglucinol and hydrochloric acid on the
termite inte-iiu.il contents and interpreted the macroscopic pink
color rea< lion, which appeared immediately after the solution
was pi." < d "ii the intestinal contents, as a positive test for li-nin
and, from tin- experiment, he drew the conclusion that li^nin

•ermite's (Coptotermes formosanus} alimentary
c.m.il mi." i<d upon. Buscalioni and Comes (1910) interpreted
the phloro-Jucinol reaction as indicative of the presence of wood
ITS in the bodie- of the protozoa, which sugars, the-e investi-
gators < l.iimed, the protozoa had by enzymatic action elaborated
from the in^eMed lignocellulose particles. According to Sherrard

[922 no \\ood sugars can be obtained from wood except by
extr.n i in^ the cellulose. Mannan, for instance, very probably
exists in the form of manno-cellulose. In even' case the quantity
ol' i ( •llnlosf removed corresponds to the quantity of Mi-ar pro-
diiir.l. In other words, the amount of sugar obtained is pn 'por-
tion.(1 to the cellulose extracted. Hitter and Fleck (1922 . in
making .m anlyasis of some American woods, obtained a yield
(a mean of four samples) of 6.82 per cent, of pento-an content
I P. m western yellow pine (Finns ponderosa) extracted cellulose
and a \icld of 22.82 per cent, from tanbark oak (Quercus ilfiisi-
Jlorn\ extracted cellulose.

Aldoju litres, such as /-xylose and /-arabinose, when wanned
\\ith phloroglucinol and hydrochloric acid, give a cherrv'-red
color to the solution. But this color reaction does not take place
in the cold. Crocker (1921) states: "No case^is known where
materials conceivably present in wood, other than aldehydes,
c.\\\ react with the phloroglucinol reagent, in the cold, to produce
a red color." Crocker reaches this conclusion after having tested
a large number of chemically pure compounds. Klason l thinks
1 Svensk Pappers-Tid; 23 (1920). p. 70.

206 L. R. CLEVELAND.

the principal color-forming aldehyde of wood is coniferyl alde-
hyde. Cross and Doree (1922) conclude: "We have thereform
a direct confirmation of the view that the color reactions of lig-
nocellulose are those of derivatives and associated furfural
products."

When tested with aniline acetate most of the wood ingesting
protozoa are colored yellow, which coloration is probably brought
about by the presence of methylfurfuraldehyde (Cunningham
and Doree 1914). When tested with the same chemical, at the
same time, a small percentage of the wood ingesting protozoa
and many of the wood particles free in the termite's (Termopsis]
intestine are colored pink, which coloration is probably brought
about by the presence of a trace of furfuraldehyde, for Cross and
Doree (1922) state that "the aniline acetate test differentiates
the aldehydes." Wood, before being ingested by the termites
gives the same pink color reaction to phloroglucinol and hydro-
chloric acid that appears inside the bodies of the protozoa. Also,
aniline acetate reacts with uningested wood in the same way that
it does with the wood particles inside the bodies of the wood in-
gesting protozoa and the wood particles free in the intestine.

Now, what does the phloroglucinol reaction show? \Vhat is
its value? It may show that the reacting substance (the substance
producing the coloration) is a derivative of lignocellulose or a
substance associated with lignocellulose. If the reacting sub-
stance is a derivative of lignocellulose, the lignocellulose bond or
linkage has perhaps been broken, and it is highly probable that
the protozoa are responsible for the breaking of it. Not many of
the wood particles free in the intestine exhibit this reaction, and
it is possible, of course, that the particles which do react have been
inside the bodies of the protozoa. If the reacting substance is
an aldopentose, as Buscalioni and Comes (1910) thought it was,
it may be that the linkage or bond between aldopentoses and
cellulose (Sherrard 1922) has been broken and, if so, it is highly
probable that the protozoa have broken it. But, since the wood,
before being ingested by the termites and reingested by the
protozoa, gives the same color reaction that it does after inges-
tion, it is perhaps best to suspend judgment until more is known
regarding the specificity of the phloroglucinol reaction.

However, the clear-cut, definite, and unmistakable glycogenic

IXTI.-IIXAL PROTOZOA OF TERMITES. 2OJ

reaction which appears inside the bodies of the wood-ingesting
protozoa when treated with iodine in a solution of potassium
iodide, is, perhaps, sufficient evidence to show that the protozoa
rt a di.^e-tive action on the wood particles which they take
into their bodie-, f»r after the host has been fed cellulose (What-
man's filter p.. i per. Vide infra] for three months ^lyco-en is
still pre-ent in the bodies of the protozoa and it cannot be demon-
strated n»w. or at .my time, in any of the cells of the host. The
cellnlo-e i - probably split into cellobiose, the cellobio>e into
gln< o-e, 1'ri'in whifh the glycogen is formed. The fact that no
glycogen i- pre-ent in the intestinal protozoa which do not in^e-t
the v.o.,,1 p.irti< le- op the cellulose particles, nor in the rells of
the ho-t , iin lie. ite~ -irongly that the glycogen is formed inside the
bodic- of the proto/oa that do ingest the wood particle- or cellu-
|Mitiil«- .nid that these protozoa are responsible for it-
fonnation. The intestine of the host is slightly alkaline (pi 1 7
which \\oiild make acid hydrolysis impossible. The cellul'
tin ii, must be acted upon by enzymes.

1 1 has n a possible to demonstrate the presence of fat-

in the bodies of the protozoa, though the presence of fats in the
inieMin.il and other cells of the host may be easily demonstrated.
I . i« unites which receive a fluid diet from other member- of
the colony accumulate an abundance of fat.

It i- highly possible that the protozoa are dependent on their
ho-t for proteins, since they live much longer on a cellul'
inorganic culture medium to which proteins have been added.

In -ind\ini; many of the termites of Japan, Oshima (. I<H
came to the conclusion that a preference was e-- 1 in favor

of those vo>ods containing a high percentage of cellulo-e. II. is
of the opinion that cellulose is the only constituent of \\ood which
termite- n-e as food. This opinion is based on the lollowin^
perinient which he carried out. Camphor \\<»»\ was analyzed
an. I fed to ( 'nptotermes formosamis; the nest which was constructed
after the wood had been eaten was anali/ed. .ind it was noted
that tin chief difference between the camphor wood and the nest
\\.ts in the cellulose-lignin ratio. The wood contained 4* per
cent, of cellulose and 20 per cent, of lignin (C4oH44Oi5 Beckmann,
I'»_M) while the nest contained 15 per cent, cellulose and 57 per
cent, lignin. The method by which these percentages were ob-

208 L. R. CLEVELAND.

tained is not given, and for this reason we have no way of judging
their accuracy. Regarding this experiment Oshima remarks:
"It is quite obvious that the amount of cellulose is the main
difference between the constituents of camphor wood and those
of the nest. As there occurs no decrease of noncellulose, it is
clear that cellulose has been taken as the food when camphor
wood passes through the alimentary canal; and noncellulose, that
is lignin, which is produced as a decomposed material of ligno-
cellulose by special function of the alimentary canal, is discharged
as the building material of the nest."

Obviously, Oshima's data do not warrant the conclusions which
he draws from them, for it is impossible, without doing a quanti-
tative experiment, to establish the fact that lignin is not used as
food and that cellulose is, since both substances appear in the
wood and in the nest. Oshima completely disregards — purposely
or ignorantly it is impossible to say — all substances which are
present in the wood except cellulose and lignin. The fact that
lignin is present in the feces (and the test used is not a test for
lignin at all) does not prove that it is not used by the termite as
food, for cellulose is also present in the fecal material and has to
pass through the alimentary canal many times before it is used
up. Perhaps it is all never converted into cellobiose and glucose.
Then, it has not been shown that termites do not use lignin as
food; nor has it been shown that they do.

Oshima (1919) fed cotton wool to Coptotermes formosanus and
claimed that these termites lived more actively on this substance
than when fed softwoods, the reason being, so he thought, that
the softwoods have a lower percentage of cellulose. But the
softwoods do not have a lower percentage of cellulose — at least
those that have been anlayzed do not have. However, hard-
woods average about 100 per cent, higher in pentosan content
than softwoods. But Oshima's experiment has no value what-
ever, since the number of termites used and the length of life,
when given a cellulose diet (Oshima's diet of cotton 1), is not

1 Purified cotton yields an ash ranging from o.io to 0.50 per cent., while raw
cotton contains about i per cent, of mineral matter. Ra\v cotton contains about
90 per cent, cellulose and absorbent cotton 99 per cent. There is about i per cent .
of pentosan in purified cotton. Most any filter paper, then, is a better rrllulu-''
diet than cotton in any form. Some filters contain an extremely small pen vnui.m-
of ash. Of course it .cannot be said that the cellulose of filter paper is "normal"

INTESTINAL PROTOZOA OF TKKMITKS. Joe)

state* I. Main ob-ervers have noted that termites attack books,

pastel)*', inl, cloth, and wood pulp, but their ability t<> maintain

them-elve- • MI i he-e -ubstances has not been investigated. Think-

it p"--ible th.it termites in large numbers, when fed cotton,1

do n<>t maintain themselves solely on the substance- pre-ent in

the cotton, but continue to pass the wood, which they have

, iou-ly eaten bcf-.re being changed to a diet of cotton, through

their bod . a portion of it during each pa^~ gi md

thu- .yivin", the fal-e impression of living on either the cotton or

the filter paper, when in reality they were living on the wood

•.Inch miijit -till be present in their feces, I carried out

the follov, ment: One hunderd termites were placed in

ioo vials, IK- termite to each vial, and nothing but pure

cellulo-, \\ 'man's filter paper) was fed them. Many of tl

termite- led .MI active existence and continued to harbor pmto/oa

until t riments were discontinued at the end of four months.

|u-t how long termites can maintain themselves on a cellulo-e

i- not known, but is being investigated by the writer at the

ent time.

Another experiment, which was first done by accident, -et -in-
to thro\\ considerable light on the direct cause of the de.it h of the
incubated and defaunated termites. Many of the cletaun.
lennit- bein;; fed cellulose and a Strong cellulose-decompos-

accidently got into some of the vials, and it was noticed
that the termites in the vials containing the cellulose-digesting
luni;u- did not die within two to three weeks. Some of them
lived in tin- \ials with the fungus for more than three month-,
but in the vials containing some of the same defaunated termite-
and cellulo-e which was not digested by a fungus, death occurred
\\ithin t\\<> to three weeks. When this accidental and preliminary
experiment was repeated (see Table III.) the same result- v.
obtained. It seems quite evident, then, that incubated and
del'aunaied termites cannot digest cellulose.

crllu n is really the only form of normal »<•!] :

aixl ' ii?22) have made a careful study «\ tin- irllulo.-r uf o.tton ami that i.i

iu hide that the two are very similar, it net .il. The two, ac-

'nx tn tliese investigators, are as much alike as tin- n-llulcsc- ulitaiiu-d imui th<-
.-aiiu- \\IM,,I by acid cooking and by alkalini- codkinx. I 'mil iimn- is kno\\-n rt-gard-

;hr nature of cellulose in various sul>-tanccs \VI may o.iHimir tin- usage of
calling o'tton "normal" cellulose.

210 L. R. CLI-.X I I \\D.

Now, since the incubated and defaunated termites remain
alive and active indefinitely (See Table III.) when fed humus—
which is wood (chiefly cellulose) digested or decomposed by
bacteria, fungi, actinomycetes — and digested cellulose, but die
when fed wood or when fed cellulose, it seems very probable that
death is due to an inability to digest wood. Futhermore, it has
been shown that no sugars can be gotten from wood except by
first extracting the cellulose, which probably means that the
sugars exist in combination with the cellulose. Then, we may
reasonably conclude that the inability on the part of the defaun-
ated termites to digest wood is due, perhaps, to the disappear-
ance of cellulase, brought about, in some way, by the incubation
and removal of the protozoa.

If cellulase is no longer present in termites after incubation—
and it does not seem to be — what is responsible for its disappear-
ance? In order to answer this question the two factors involved
in bringing about the disapparance of cellulase, incubation and
the removal of the protozoa, must be separated. Reinfection of
the incubated and defaunated termites with protozoa seemed to
be the principal key with which this puzzling question might be
unlocked, but unfortunately for several months it was impossible
to replace the protozoa in the termites, once they had been thor-
oughly removed. Finally, a rapid and easy method of reinfection
was developed. Since the reinfection experiments are the most
important ones in the paper, perhaps, one of the seven which
were carried out, should be given in detail just as done and record-
ed in my note book. The other six are very similar to the one
recorded here.

Ten large jars of termites were collected and were numbered
l-io/>45. Numbers i~7/>45 were incubated at 36° C. from 3:30
P.M., 10/21/22 to 3:30 P.M., 10/22/22. Numbers 8, 9 and 10/545
were not incubated and were left at room temperature (about
20° C.). Jars i~7/>45 were carefully examined on 10/23/22 and
the termites in them were found to harbor no protozoa. In
making this examination twelve individuals (workers), taken at
random from each jar, were carefully dissected and their entire
intestinal content painstakingly examined under the microscope
for protozoa. If no protozoa were found, the colony was labeled
"defaunated termites." Jar 6/745 was used as follows on 9/24/22,

INTESTINAL PROTOZOA OF TERMITES. 211

the day after incubation: By means of a sieve the termites
were -eparated from all pieces of wood and clumps of excrement
larger than them-elves. Then they were placed in beakers, and
a piece of moi-t filu-r paper was placed in each beaker so that one
corner ju-t touched tin- bottom. The termites would leave the
small pi 1 and excrement and crawl up the filu-r paper,

and when a lar-e number of them were on the filter paper they
wen- -haken from it and transferred to several small beakers.
By tin- method the termites were entirely freed from all parti'
of wood .Mul ex< rement. Now they were taken, one at a time,
from tin -mall beaker- and placed in watch gl. Here the

rijit antenna w close up to the head, and one by one,

ah IT the riijit antenna had been removed, they were placed in
shell vials until ten individuals were in each vial. In this ex-
periment 14 \ial- with 10 defaunated termites each were u-ed.
The-e were numbered [ 1 4/749. Some unincubated and faunat ed
termite- were taken from jar 9p45, collected at the same time
that <>/• js was and taken from the same colony of termite- in
nature. Tin < ed from wood and excrement in the same

way that the defaunated ones were. Then 10 unincubated and
faunated termite- were placed with 10 incubated and defaunated
termite- in \ial- i n/>49; in vials 12, 13, and !4/>4() fifty unin-
cubated termite- were placed. Workers only were used in all
the experiment-. Whatman's filter paper, moistened with di —
tilled v put in each vial; the vials were tightly corked

and left at rooin temperature in the dark. l\e-uh>: 1<> 2 \ 22 at
5 I'.M. three termites from vial 8^)49 were examined. Two con-
tained no protozoa and one contained a few. i<> 27 22 all of the
indi\idual- those previously defaunated and tlio-(- from which
no pn>to/oa had been removed by incubation in ii/->4<> were
examined. Two were dead, 12 harbored proto/«-a, and <> harbored
no protozoa. Three of the 12 that harbored protozoa had the
risjit antenna cut off. This was a clear-cut case of refaun.ition
(replacing of the protozoa in the inte-tine after they had been
removed by incubation). The number of protozoa present in
these refaunaud or reinfected individuals was about half normal.
On 12 2ii 22 six vials were examined with results as follow-:

"

212 L. R. CLEVELAND.

Number 3/749 contained 16 termites, 6 with no right antenna.

" 4/>49 9 " , 7

" 5/>49 16 " , 7

" 6/>49 ii " ,5

n ii ii i i ii ii ii

7/>49 16 , 4

In number 14/^49 fifty unincubated and faunated termites had
been placed with the ten incubated ones and only two individuals
with the right antenna cut off were taken from this vial, though
others were present and might have been taken. Nearly all of
the fifty unincubated individuals were alive. Now the termites
with the right antenna removed which were taken from vials
3, 4, 5, 6, 7, and i4/>49 to be counted were placed in separate vials
i-6/>88 with a label on each vial designating the source. What-
man's filter paper was given them for food. They were observed
regularly. All are alive now, 2/7/23,1 five weeks from the time they
were separated from the unincubated. Two of the individuals
in 3/>49 were examined and were heavily infected with protozoa.
After association for several weeks with termites harboring
protozoa the defaunated termites regain their protozoa and, at
the same time, their ability to live on a diet of pure cellulose or
wood. Now, the important question is, how did the termites
regain their ability to make use of pure cellulose or wood as food?
More specifically, why do they now possess cellulase — and,
perhaps, cellobiase? When the protozoa are removed from a
termite's intestine, the ability to make use of pure cellulose or
wood is lost, but when they are replaced, this ability reappears.
Do the cells of the termite possess cellulase and cellobiase? It
seems that they do not. Then where may these enzymes be
found? The fact that there presence can only be demonstrated
when the protozoa are present in the intestine of the host seems
to indicate that the protozoa possibly possess them. It the
protozoa do possess these enzymes, and the host does not, the
way in which they aid their host is plain.

It seems evident that the protozoa aid their host in some w.iy
to live on cellulose and wood, for it has been shown (Table III.)
that defaunated termites can live indefinitely on the digestion
products of pure cellulose or wood (humus). It now 1 KV< >mes

1 Many are alive now 4/10/23.

INTESTINAL PROTOZOA OF TERMITES. 213

highly probable that the protozoa aid their host in the digestion
of pure cellulose and wood, since the host is able to live indefi-
nitely on either of these substances only when it harbors the
proto/oa. Hut how do the protozoa aid their host in the digestion
of wood;-' Mechanic. illy, or by furnishing enzymes, as suggested
above, which the ho-t it -elf does not possess.

I- i- [io--il.li- that tlu- protozoa mechanically aid their host
-ince the intestine i- completely gorged with them. The intestinal
<e!l- o| the termite in rete enzymes such as cellulase and

<ellobia-e. only when stimulated by actual contact with vast
number- o| inte-tinal pn>to/o.i. 1'avlovsky and Zarin (l<>j_

•inly of the ferment- in the alimentary canal of the bee. Apis
nii-llifcru, found c.ttalase in the stomach and large intestine in
winter onl\ Phree hours after the first flight in spring, r.
\va- pre-eiii in a very small amount, and two days later, not a
i race \\ a- lilt. The-e authors claim that the discharge of catala-e
in i he rectum depends upon the accumulation of feces in it
during hibernal ion. The l>ee has no need for this enzyme except
during hibernal ion, when it seems to regulate the different <
di/iiu; pr- •• and <1< - the surplus peroxides as they ac-

cimiulate. The production of any en/yme may be explained
either .1- a "hunger phenomenon" or due to favorable conditions
• •I nutrition, the latter being, perhaps, the better explanation.
You:.. [918 .in hi- \\ork on inulase, showed that stimulation to
iucre.i-eil production of this enzyme was dependent upon a
direct chemical -timuliis - ub-t a nces present in the medium. It
ha- been imp. --ible to extract cellulase or cellobiase from either
those termites harboring protozoa or those not harboring proto,
delaunated termiti-- . lYin^-heim (1912) claims that cellulase
i- an endc .-en/\ me and i- <1 from the cell when stimulated

b\ direct contact with cellulose. This perhaps explains my ex-

l ractimi failu:

The other way the proto/...i may aid their host is to dige.-t the
\\ood for it. Sin, i- it ha- been shown that defaunated termites
cannot dii^e-t \\ood. the ex[>eriments of Buscalioni and Comes
MHO , \\here t he-e in\ e-t i ^ators demonstrated that the wood
particles iu-e-ted by the protozoa are digested, are now of -reat
value, though thcv \\ere worthless so long as it was not kno-\\ n
that the host could not digest wood without the protozoa. There

214 L. R. CLEVELAND.

are many reasons for thinking that the protozoa digest the wood
for their host. In Reticulitermes flavipes practically all the wood
particles which reach the hind-intestine are immediately ingested
by the protozoa. Hundreds of individuals of this species have
been examined to note if wood particles were ever present in a
very great quantity in the intestine, and in only a few instances
were they ever found in abundance. Perhaps, directly after the
termite has taken a meal, wood particles may be quite plentiful,
in the intestine, but it is certainly not long before nearly all of
these particles of wrood are taken into the bodies of the protozoa.
In Termopsis many wood particles are free in the intestine, though
the wood ingesting protozoa harbored by this genus always have
their bodies filled with wood particles to the same extent as do
the wood ingesting protozoa harbored by Reticulitermes. Quite
a bit of woody material is present in the expressed pellets of
excrement of Termopsis, whereas Reticulitermes, on the contrary
has a liquid excrement.

Of course the mere fact that the protozoa take the particles
of wood into their bodies does not mean that they in any way
alter these substances, for many protozoa ingest substances
which they cannot use as food. Not all of the termite protozoa
ingest solid particles of wood for food, or at all, though most of
them do. Among the protozoa harbored by Termopsis sp., of
Ashland, Oregon, Streblomastix strix does not ingest the wood
particles, and this organism never gives a glycogen reaction.
Streblomastix strix, then, may get its nourishment from the diges-
tion products of the other protozoa which Termopsis harbors,
such as Trichonympha campanula, which ingests great quantities
of wood particles and always gives a very clear cut and definite
glycogen reaction, provided its host has not been wood starved;
or it may derive its nourishment directly from the products which
its host has assimilated, in which case it is a true parasite. If
this protozoon gets its noruishment from the digestion products
of other protozoa present in the intestine of its host, just in the
same way that the termite gets its nourishment, Streblomastix
strix, then, may be a commensal from the viewpoint of the ter-
mite, but from the viewpoint of its relation to the wood digesting
protozoa, it is a true parasite, since it in no way aids the wood
digesting protozoa, whereas the termite does aid them since it

INTESTINAL PROTOZOA OF TKKMITi 5. 215

furnishes the wood which they digest. Of course Streblomastix
stri\- may use both the- products assimilated by the termite and
the products present in the termite's intestine as a result of the
diyv-tion of wood by other protozoa, which products have not
i a --i mi laird by the termite. Those individuals of Tricho-
nympha campanula containing the greatest quantity of wood
•i i heir bodii 'he most distinct glycogen read ion.

The n MI lion i- mm h more pronounced in the posterior half of
i he liody where iiv : he ingested wood remains. Another

lorn i, 'in monas because this organism

• in undulating meml-raiir. a\o-tyle and three flagella and is
not the -aim- or-aiii-m as / .itits tcrmitidis which Kofoid

and v\r/\ [919 d< 1 from Termopsis angusticollis of

I iforni i-sts wood particles but does not give

the i) except in a very few instances. As a matter

"t lac i, .ilx.iit one indhidual in a hundred examined will give a

n reaction. This organism probably does not make
u-r oi ilit \\ii..d j)articles, but, perhaps, ingests them while feed-
ing on barter!. i, since it has a cytostome, an organelle which
Tritium y»: /•>;<!, l>;nen yntpha, Streblomastix, Pyrsonytnpha, and
nio-t i 't tin oilu-r genera of termite protozoa, do not possess.

\m : . tin- pi.>io/oa harl>ored by l\ termes flavipes,

Trichonympha agilis and Pyrsonytnpha vertens ingest wood parti-
ami ^i\c- a vK'cogen reaction. Several of the species of
Dinenympha d" not ingest WOCK! particles or give a glycogen re-
action. \\la-n thr-e termites (Termopsis and Rcticulitcrmes) are
\\.i.id -taiAed. the protozoa mpha campanula, T.

\'m /)//</, etc.) that ijve tin i reaction always

die ott in-t. S. nnetime> ; all>- all of these forms are dead

U-ii'iv ,tii\ diminution in the oilier form- o( rurs. Pyrsonytnpha
is piol>aM\ of '<reat value t" its ho-t (i\\in^ to its peculiar habit
of attaching it-elf to the chitin<ms la\ er investing the inner
surface of the inte-tiiie. Individual- of this genus, instead of
pa--iiu out of the inte-tine with tin- lV<v- when they die, remain
attached to it ! ter drath, and iluis the food which they

ha\e prepared for their ho-t i- kept in the intestine until the
ho>t ha> an oj)|)ort unit y to u-e it. The attachment is, perhap.-,
al>o ([uite useful to the proto/nnn. -ince this organism is not
provided with mimnou- ila^ella as is Trichonympha and by

2l6 L. R. CLEVELAND.

means of the attachment apparatus it is thus enabled to maintain
its position in the intestine in the complex struggle for existence
among the many species of protozoa which completely fill the
lumen of the termite's intestine. No doubt Pyrsonympha, Tri-
chonympha and others, die by the hundreds daily, and thus give
over to their host many substances which they have obtained
from the wood particles. That they give sugars,1 such as xylose,
may be shown by testing with phloroglucinol and hydrochloric
acid. But it is not the intention of the writer to show in this
paper the relation of each species of protozoa to its host. Not all
of the protozoa harbored by the two hosts, Termopsis and
Reticulitermes, have been mentioned, and the protozoa of other
termite genera have not been mentioned at all. In a later paper
this question will be considered in more detail.

Unwilling to conclude definitely that the protozoa were entirely
responsible for the digestion of cellulose, the other microorganisms
harbored by Reticulitermes flavipes were studied. Bacteria were
sometimes numerous, and these were studied in many ways. In-
cubation at 36° C. for 24 hours does not seem to affect their
numbers. All the known methods, aerobic and anaerobic, for
isolating cellulose decomposing bacteria were given more than
fifty trials, but, since the results of all these experiments were
negative, no tabulation has been made. An inorganic medium
composed of

K2HPO4 i .00 gram

MgSO4 0.50 "

KC1 0.50 "

FeSO4 o.oi "

NaNO3 2.00 "

H2O 1000.00 cc.

to which cellulose was added in two forms: a small piece of
Whatman's filter paper was placed in the test-tubes containing
the inorganic medium and a 0.5 per cent, cellulose suspension
was added to the inorganic medium. This was sterilized in the
usual way and the inoculations made. Then the cellulose sus-
pension and inorganic medium plus agar sufficient to make a

1 Provided the conclusions of Buscalioni and Comes (1910) regarding the phloro-
glucinol reaction are accepted. For a discussion of this reaction see page 205-6.

INI!-II\AL PROTOZOA OF TERMIM-. 217

-olid medium \\vn- u-ed. When cellulose was the only carbohy-
drate pre-ent. in i growth ever occurred; but when other carbohy-
dr. itt--. -in h rn-e, were present a rapid growth took place.

nn -re inf'irmation regarding the methods used in studying
cellule- "inpo-in^ microorganisms the reader is referred to

\V k-niaii I'/i'i .UK! 1922'), Kellerman (1913), McBeth [913
and I'd'' . Brown i'M5 and 1917), Schmitz (1919), and oil'..
Alter duplexing all of these methods it was concluded that no
cellule :ni; bacteria are harbored by Rcticiilitcr-

Jin: . im e more than one hundred individual attempt- to

i-"l iposing bacteria met with failure, e\en

r the culturi <• more than two months old. It was an

matter t" such bacteria from the soil and from eo\\-

( hip-.

An .it tempt wa- also made to isolate cellulose decomp"-in-
mould- and actinor. - from termite:?, and this attempt w

failure. '1 lii- e\|>eriment was repeated ten times with the same
re-ult ea< li time.

None i'! the known cellulose-digesting Ixicteria, moulds and
in >m\ • i t« - dies at 36° C\ As a matter of fact this is the optimum
temporal urc for many of the-e organisms. In some install
tin- t ellulo-r is (lccomiM)sed much more rapidly at even a higher
temperature.

Tin . ativr results indicate that Rfticitlitermes flavipes

harbor- no bacteria, moulds or actinomycetes capable of decom-
|>c-in;< cellulose.

Die most logical conclu-ion, then, from the results of all
tin experiments carried out. i- that the protozoa actually aid
tlu-ir liM-t l.\ ili-e-iin^ tlie \\....d--or more specifically by break-
ing up the cellulo-e niolei uK- and treeing many substances which
bound toit in -Mine \\a\ for it, because the host itself cannot
dii^e-t cellulo-e. The proto/n.i can digest cellulose.

The relation-hip, then, between Reticnlilermes flavipes and
-onie of it- inte.-tinal pn>to/oa, particularly Trichonympha and
/' ;. / ^niymplui, i- one • >i -\mbio-is. The termites are able to
maintain thcni-elvc- on their normal diet of wood only when
they harbor pn»to/..a to due-t the wood for them. The protozoa
are harbi.ivd only when their host eats wood or the excreta of

1 TI. . .mtain- .j,; n. [i

2l8 L. R. CLEVELAND.

other individuals containing wood and protozoa. The larvae must
become infested \vith protozoa, by eating the feces of older and
infested members of the colony, before they can maintain them-
selves on a strict wood diet. The host procures the wood for the
parasite, and the parasite digests it for itself and for its host.
Each is a servant to the other; the protozoa are dependent on
the termites for food and lodging, and the termites are dependent
on the protozoa for the protozoal cellulose digestion products.

GENERAL DISCUSSION.

The number of protozoa present in a sigle host is truly enorm-
ous. Here we have an animal whose intestine is completely
filled with flagellate protozoa, the majority of which live in a
symbiotic relation to their host. Perhaps many — if not all-
intestinal flagellates — human as well as animal — are harmless to
their host. At least their presence does not mean anything—
harmful or otherwise. Only careful and painstaking investiga-
tion will reveal their relation to their host.

The examples of symbiosis in nature hitherto described, such
as algae and fungi (forming lichens), hydra and Zoochlorellee,
sea-anemone and hermit crab, are well known. These relation-
ships are much closer than such relationships as ants and aphids,
insects and flowers, etc. The symbiosis exhibited by termites
and their intestinal protozoa is as real or true as any yet described.
How such a relationship was developed is an extremely interest-
ing but very difficult question to consider — or even theorize *

1 Since none of the protozoa of termites are known to occur elsewhere it is in-
teresting to speculate on their origin. Where did these insects get them? The
ability of soil protozoa to digest cellulose has not been studied and little is known
regarding the protozoa of plants. If there are protozoa parasitic on plants, or
protozoa in the soil, capable of digesting cellulose, it is possible that some of them
became inhabitants of termites. It may be that the symbiosis was established in
this way. These insects, before harboring protozoa, fed on soil and wood and, per-
haps, were able to utilize some of the substances in the wood and many of those in
the soil. The intestine of termites, feeding on practically the same food that the
protozoa free in nature feed on, is certainly a much more constant — and hence
better — environment for the protozoa than soil, or even plants. But, if the termites
once possessed the ability to digest cellulose, why did they lose it? If the acquired
parasites had been digesting cellulose longer than the termites and could do it
better, i.e., more completely, this may be responsible for the loss of tin- ability to
digest cellulose. The insects probably lost their ability to secrete i-n/yiiu-s, such as
ccllulase and cellobiase, because it became unnecessary for them to do it. How-
ever, it is also just as reasonable to suppose that the termites nevi-r po d tin-
ability to digest cellulose, since so few animals do.

I \TEsMXAL PROTOZOA OF TERMITES. 2IQ

on — for it has undoubtedly been established a very Ion? time.
The ability of animals to live indefinitely with a sterile intes-
tine is still a live, debatable, question. Osborn and Mendel
i'( 14-'' have -hown that microorganisms are of value to higher
animal- elaborators of protein from non-protein stbstances
1. Armsby IMII showed that non-protein Mibstances
an- .1 source "t protein to herbivorous animals, probably due to
tin- formation of due-table bacterial protein in the digestive
tract. Then- i- < . rtainly a very small amount of protein in wood,
and it may be th.it termite protozoa aid their hoM in >. -me Mich
fa-hion ,i- int. -~iin.il b,i. t.-ria aid their host. Kiankin ' MI; re-
opened t!ii~ question by a review of all the literature. The
tesl quantity of microorganisms — in the ca-c of mammal-
and other vertebrat - located in the nondi^e-tive portion-;
of the alimci tial. Al-o the increasing number of animals

which may I- .lly indicates that an a-eptic exist-

ence m.iv I -I- po~-ible in the majority of cases. Bacteria carry
de. oin 1 1. 1~ it ion to a lower level, yielding unassimilable -nl>M.i;
Mich as methane, carbon dioxide, indol, skatol, phenol.

IVrhap- the chief beneficial role of the intestinal flora i-, the
-\nthetic |K. \\er- of the microorganisms.

Cut the relation of many insects to microorgani-m- i- a differ-
ent qncMion, and at lea-t from one viewpoint, if from no ..ther,
a more inti-re-t ini; one and also one which may be Mudied mon-
-ati~fact<.iil\ . HaumlH-r^er (1919) has shown "that Hrosophila
living in fermenting fruit are dependent for their fo"<l >ii].pl;
the synthetic and ab-orptive powers of yeast cell-." In a similar
m. inner hi- -tiidv of " the relation of Mtisca don:' to manure,

..f Desmometopa to decaying meat, and of Scia-cn and Tyroglyphus
to deca\iiu \\oo.l shows clearly that these arthn>])oils .ilso feed

Aiu-tlirr i>.i-«iljiluy. .iii.l jxiluips a more plausible om-. is that th..- tern.

with protozoa, probably iVd on humus. Some v

vnu -cd — along with the liumus. The protozoa \vi-rc

uik.-u in and f..iin.l thu .-nvironment suitable. Rapid multiplication and few <!•
in t: tnent. soon filled the termites with them. The •

,\\K lial'it nt" the termites gradually increased. At present every conceivable
ii,. in noii-v. iinR termites to strict wood '.ell illustrated in

•

Hut what hapjx n. .1 tn the ancestors of prc-tcrmite protozoa — the organisms
which lived in the s..il . >r in wood? They have probably evolved along another line,
in an »i .dividiials who became inhabitants ot" termites.

15

22O L. R. CLEVELAND.

on microorganisms." However, microorganisms are not asso-
ciated with these insects as symbionts — certainly not in the same
sense or degree that termite protozoa are associated with their
host — but as food.

The nitrogen content of birch-wood is very high as compared
with that of other woods. It is 0.108 per cent. In many woods it
runs as low as 0.08 per cent. From the work of Baumberger
(1919) and others, it is evident that many insects, due to the low
protein content of the substratum on which they live and feed,
are dependent on microorganisms for the proteins which they
contain. Since such a small amount of protein is present in the
food of termites it is quite likely that they, too, are dependent on
microorganisms for a part of their proteins. Fungi may often be
seen in the intestine of many termites.

The relation of ambrosia beetles to fungi has been studied by
Hedgcock (1906) and others, and the fungus-growing habits of
the Termitidse have been studied by many investigators, chief
of which are Doflein (1905 and 1906), Fetch (1906) and Escherich
(1909), but the exact relation of these insects to the fungi is not
known. The fungi may simply furnish food (protein) for the
insects, and, on the other hand, they may elaborate substances
from the wood and grasses which the beetles and termites can-
not elaborate, which substances are very important in the in-
sects' metabolism.

Internal symbionts (yeasts) have been described from the
beetle Anobium panicetim, by Karawaiew (1899) and Escherich
(1900), and Portier (1905) claims that a micrococcus and a fungus
live symbiotically with the caterpillar Nepliliica, but in no in-
stance have intestinal flagellates been shown to live symbioti-
cally with their host. The flagellates living in termites are the
first example of such a relationship.

As a rule the second and third reproductive forms do not
harbor intestinal protozoa. Why? For one of two reasons per-
haps: (i) the liquid diet upon which these individuals are fed
kills the protozoa; (2) the failure of the host to eat wood causes
the death of the protozoa. There is more evidence in favor of
the latter view. In the first form reproductive adults protozoa
are usually present — though never in abundance as in workers
and soldiers — and this form eats some, though never as much

INTI>IINAL PROTOZOA OF TERMITI •-. 221

wood as the workers and soldiers. When workers and soldiers
are wood starved the wood ingesting protozoa which they harbor
die first. When workers are fed cow chips the protozoa disap-
pear in three to four weeks. But here, as in the case of saliva, it
is impo— ible to determine whether the change in the intestinal
content brought about by the change of food, results in the loss
of protn/M.i. or whether the failure of the host to provide the
•A ith wo. nl ir-nlt- in their death. Since there is a ^reat
dr.ll d|" rellulo-e in the c,,\v i hips it is more likely that the change
in the intr-tiii.il content i- responsible for the death of the pro-
tozoa. I IK b.K teria of the cow chips do not kill the proto/
f. n -triilr COW (hip- arc ju-t as effective in removing them This
• |iie-tion i- now brin;^ aii.n krd from many angles ami it i- hoped
tliat a more deliuitr aii-\\er may result.

Their i- ron-ider.tble difference of opinion re^ardin^ tin- extent
to which cellulo-e i- mili/rd in the animal organi-m. Very little
i- really knov.ii about the mechanism which vertebrates -and
most, it not all, invertebrates too — employ in making n-e of
cellulo-e. It is intcrc-tin^ in this connection to compare the viev. -
e\ ; \v<> jirominent physiological cheini-' Hawk

i that m.my of the herbivora use as much as 25 per

cent, of r cd cellulose, that less than 5 per cent, i-

u.-cd by doga and the amount used by man is "too -mall for it
to pi. iv a role of importance in the diet of a normal individual."
1 1, saj B: "In neither man nor the lower animal- lias there been
( lrmoii-t rated .111 v formation of sugar or glycogen from crllnl-
Y,,n Furth IMI that 30-70 per cent, of the cellulose

eaten by herbivon >u> doiu«--tic animals is dige-trd, and that man
digests about 50 per cent, of the cellulose which he consumes, and
in , .f habitual coii-tipation he may dig« much as 80

prr cent. The tillered mammalian intestinal content i- inactive
to cellnlo-r i rrin^-heim, I«)I9). Pringsheim (1919) also claim-
that no cellobio-r--plitiing enzNTnes are pre-ent in any of the
organs of cattle. Biery (1914), Bier\- and ( ".i.ija (1912 and
Hillard (1014) report the presence of cellulose-splitting enzymes
in hepatic secretion- of certain mollusks and cnisiace.ni-. Har-
rington (1921) claims to have demonstrated the pre-rncc of
cellulo-e--pliitin^ ni/ymes in Teredo. Dore and Miller (1923)
made a comparative analysis of the wood eaten and the borings

222 L. R. CLEVKI.ANT).

passed by Teredo navalis, and, as a result of the analyses, these
investigators conclude that "the wood lost about 80 per cent, of
its cellulose, and from 15-56 per cent of its hemicelluloses, in-
cluding from II to 40 per cent furfural yielding constituents such
as pentosans, etc.," during its passage through the animals'
digestive tracts. They state further that "The simplest explana-
tion of the disappearance of this carbohydrate material is that
the cellulose and hemicelluloses of wood are partly digested by
the teredo and probably hydrolyzed to simple carbohydrates
which the animals can use." In all these investigations it should
be noted that many substances other than cellulose were present.
In no instance was cellulose alone fed the animals. It would be
very interesting, indeed, to study the microorganisms of these
crustaceans and mollusks. No mention is made by these investi-
gators regarding the possibility of microorganismal cellulose
digestion.

Recently a cellulose-digesting anaerobic bacterium has been
isolated from 60 per cent of the human stools examined (Khou-
vine-Delaunay, 1922), and it is quite possible that future re-
searches will reveal that intestinal microorganisms perhaps play
a more important part in cellulose digestion than has previously
been thought. Intestinal bacteria and fungi quite often aid their
vertebrate and invertebrate hosts in the digestion of cellulose and,
since it has been shown in the present investigation that intestinal
protozoa can digest cellulose, it is now possible that the Infu-
soria, such as Diplodiniiim, Entodinium, Bittschlia, Isotricha,
Dasytriclia and Ophryoscolex, harbored by ruminants, notably the
ox, goat, sheep, camel and reindeer, may aid their hosts in the
digestion of cellulose and hemicellulose. The Equidre also harbor
similar Infusoria.

GENERAL SUMMARY.

1. There are four families of termites and all the species and
genera of three of them, Kalotermitidae, Rhinotermitida? and
Mastotermitida?, that have been examined have been found to
harbor enormous numbers of intestinal protozoa. No termite of
the other family, Termitida?, has been found to harbor intestinal
protozoa.

2. The principal food of protozoa harboring termites is wood
and the principal compound in the wood which the termites use

INTESTINAL PROTOZOA OF TERMITES. 223

is cellulose. This was demonstrated by keeping termites alive
and active indefinitely on a cellulose diet.

3. The protozoa harbored are all killed off by incubation at
36° C. for 24 hours, while the termites apparently are not injured
at all by the incubation.

4. The incubated and defaunated (with the protozoan fauna
removed) termites die within 10-20 days, on the average, after
incubation, if fed tlu-ir normal diet of wood.

5. When the incubated and defaunated termites are fed di-
gested wood (e.g.,humus) or fungus digested cellulose, they live
indefinitely.

6. The ilr.ith of the incubated and defaunated termites is not
dm- to the. incubation t>cr se, but to an inability to digest wood.

7. When the incubated and defaunated termite- are rein-
IV. ted with pro1 'heir ability to utilize wood, their normal
diet, reap; Mid they live indefinitely.

8. The reni-Aal, then, of the protozoa seem- to In- tv-poii>ible
for the loss of the ability to utilize wood as fc.nl. l'o determine
this i|ue-tion the ability of the bacteria, fungi and protozoa,
harbored bv termites, to digest wood or pure cellule-^- w.i- care-
fully >t in lied. It was loinul that the bacteria and fungi could not
digest cellulo-e, but that some of the protozoa could.

.). Now, -in- e the termites die in 10-20 days, if fed a \\ 1 diet ,

after the pi"io/oa have been removed from them bec.iu-e they
cannot dige-t their food (wood), as shown by the fact that they
do not die, but live indefinitely, when fed digested WO.M) <ir when
reinfei ted \\ith protozoa and fed wood, and >ince the proto/<>a
do digest tin- \\ccd particles which they take into their bodie-, it
i> highly probable, if not certain, that the termites are dependent
on the plot../, i.i t.. digest their food for them.

to. Net all of the protozoa harbored by Retictditennes flavipes
dLe-t \\ood particles. Some of them either live at the expense of
the wccil dke-ting protozoa or their host, or both.

II. The proio/oa receive from the termite- food and lodging,
for which they give in return protozoal wood digestion products.

u. The relationship between some of the protozoa, particu-
larly Trichonympha and Pyrsonympha, and their host, Rt'licnli-
termes flavipes, is one of s

224 L- R- CLEVELAND.

LITERATURE CITED.

Armsby, H. P.

'n The Nutritive Value of the Non-protein of Feeding Stuffs. Bur. An. Ind.,

Bull. 139-
Banks, N., and Snyder, T. E.

'20 A Revision of the Nearctic Termites with Notes on Biology and Geographic

Distribution. U. S. Nat. Museum Bulletin 108, pp. 1-228.
Baumberger, J. P.

'19 A Nutritional Study of Insects with Special Reference to Microorganisms

and their Substrata. Journ. Exper. Zool., 28, 1-81.
Bierry, H.

'14 Ferments digestiefs chez Helix pomalia. Compt. Rend. Soc. Biol. de Paris,

76, 710-12.
Bierry, H., and Giaja, J.

'12 Untersuchungen iiber die Mannane, Galaktane und Cellulosen angreifende

Enzyme. Biochem. Zeitschr., 40, 377-89.
Billard, G.

'14 Note sur les ferments hydrolysant les hydrates de carbone chez I' Helix

pomalia. Compt. Rend. Soc. Biol. de Paris, 76, 566-67.
Brown, W.

'15 Studies on the Physiology of Parasitism. I. The Action of Botrytis cinerea.

Ann. Bot., 29, 313-48.
'17 Idem. Ibidem, 31, 489-98.
Brunelli, G.

'05 Sulla Struttura dell' Ovario dei Termitidi. Atti R. Accad. Lincei, Rend.,

14 (2 sem.), 122-27.
Buscalioni, L., and Comes, S.

'10 La Digestione della Membrane Vegatali per opera dei Flagellati contenuti
nell' Intestino Termitidi, e il Problema della Simbiosi. Atti Accad.
Gioenia Sci. Nat., Catania, ser. 5, 3, Mem. XVII, pp. 1-16.
Crocker, E. C.

'21 An Experimental Study of the Significance of "Lignin " Color Reactions.

J. Ind. Eng. Chem., 13, 625-28.
Cross, C. F., and Doree, C.

'22 Researches on Cellulose IV. London, Longmans, 253 pp.
Cunningham, M., and Doree, C.

'14 The Production of w-hydroxy-s-methylfurfuraldehyde from Carbohydrates
and its influence on the estimation of Pentosans and Methylpentosans.
Biochem. Journ., 8, 438-47.
Cutler, D. W.

'21 Observations on the Protozoa Parasitic in Archolcrmopsis wroughtoni Desn.
Part III. Pseudotrichonympha pristina. Quart. Journ. Micro. Sci., 65,
247-64.
Dobell, C., and O'Connor, F. W.

'21 The Intestinal Protozoa of Man. London, Bale, 211 pp.
Doflein, F.

'05 Die Pilzkulturen der Termiten. Verh. deutsch. Zool. Ges., 15, 140-49.
'06 Ostasienfahst. Ergebnisse und Beobachtungen eines Naturforsches in
China, Japan und Ceylon. Berlin, Teubner, S. 459-73.

INTESTINAL PROTOZOA OF TERMITES.

Dore, W. H.

'19 The Proximate Analysis of Wood. J. Ind. Eng. Chem., n, 556-63.

*2oa The Determination of Cellulose in Wood. Ibidem, 12, 264-69.

'2ob The Proximate Analysis of Coniferous Woods. Ibidem, 12, 476-79.
Dore, W. H., and Miller, R. C.

'23 1 : i km 01 Wood by Teredo navalis. Univ. Calif. Publ. in Zool..

22. 383-400.

Escherick, K.

'oo rcyi.-lnui- rkommen von Sprosspilzen in dem Darmepitlu-1

t-iii' - K B nt.. 20, 349-57.

'09 ni<- pil// . :i Termiten. Ibidem, 29, 1-27.

Feytaud, J.

'12 i ontl . i Termite lucifugt. Archiv d'Anat. Micros.. 13,

Fischer, E., and Zemplen, G.

'04 V gegen einige Enzyme. AMI. i In in ,

365. i

'10 '-.•

Furth, O. Von.

'16 I ical and Pathological C! M.-t.il>"liMii.

.us., Philadelphia. Lippini»ii. \\ pp.

Grassi, B., and Sandias, A.

'93 ••>) dclla societa dei Termiiidi. Atti • • ni.t

r. 4) 6, Mem. XII. e 7. Mctn.
:iia. 1893. 1 • .h-h tran :i iju.nt. Jmirn.

^j 315 and 40. 1-75.)

Grassi, B., and Foa, A.

'ii 1 oi dei Termitidi. Nota Prelim in i:> . K. n I. K. A. ..pi.

20. 7*5-4'-

GroeneweRe, J.

'21 i die Zerwtxung der Zelluloae durd I--M.

Jardin Botan. de Buitzenzorg. i 3).

Harrington, C. R.

'.•i \ . siology of the Ship-worm ( 7

.i.. 15. :;'• 41-

Hawk, P. H.

'21 Practical il Chemistry. Blakiston. 0751

Haworth, W. A., and Hirst, E. L.

'21 1 lif » lonBtitUtioD of Disaccharides. Part V. i\-lM>iose (cellose). J. ( 'ln-m.

. 119. 192-201.
Heath, Harold.

'02 The Habit ifornia Termites. BIOI . Hn i... 4.

Hedgcock, G. G.

'06 StU'li'.-' up"ii S. mie Chromogenic Fungi which Discolor Wood. 171)1 Ann.

Mo. li.n.in. Card., pp. 59-114.
Hibbert. H.

'21 Tin- c'.'ii-titution of Ci-llulo-r. J. Ind. Eni;. Chem.. 13, 3
Imms, A. D.

"19 On tin- Structure and Bioln-y <>i Archotfrmopsix, together with I )i--<rription3
(.1 nc\\ - "i Inti-itinal Protozoa, and (jeneral Observations on the

[soptera. Roy. Soc. London, 2098, 75-180.

226 L. R. CLEVELAND.

Irvine, J. C., and Hirst, E. L.

'22 The Constitution of Polysaccharides. Part V. The Yield of Glucose from

Cotton Cellulose. J. Chem. Soc., 121, 1584-91.
Jucci, C.
'20 Sulla differenziazione della caste nella Societa dei Termitidi. Rend. R.

Accad. Naz., Lincei, 29,
'21 Idem. Ibidem, 30,
Karawaiew, W.
"99 Uher Anatomic und Metamorphose des Darmkanals der Larve von Ano-

bium paniceum. Biol. Cent. 9,
Kellerman, K. F., and McBeth, I. G.

'12 The Fermantation of Cellulose. Centralb. fur Bakt. und Parasitk., 2

Abt., 34- 485-99-
Kellerman, K. F., McBeth, I. G., Scales, F. M., and Smith, N. R.

'13 The Identification and Classification of Cellulose-dissolving Bacteria.

Centralb. fur Bakt. und Parasitk., 2 Abt., 39, 502-23.
Khouvine-Delaunay,
'22 Un anaerobe de 1'intestin humain digerant la cellulose. Compt. Rend. Soc.

Biol., 87, 922-23.
Kianigin, I.
'17 The Effect on Higher Animals of the Sterilization of the Inhabited Medium.

Journ. Physiol., 50, 26.
Kofoid, C. A., and Swezy, O.
"19 Studies on the Parasites of the Termites I-IV. Univ. Calif. Publ. in Zool.,

20, 1-116.
Leidy, J.

'77 On Intestinal Parasites of Termes flavipes. Proc. Acad. Nat. Sci., Phila-
delphia, 1877, 146-49.
'81 The Parasites of the Termites. Journ. Acad. Nat. Sci., Philadelphia, 8

(New Series), 425-47.
Lespes, C.
'56 Recherches sur 1'organization et les moeurs du Termite lucifuge. Ann.

Sci. Nat., Zool., (Ser. 4) 5, 227-82.
McBeth, I. G.

'16 The Decomposition of Cellulose in Soils. Soil Science, i, 437-87.
McBeth, I. G., and Scales, F. M.

'13 The Destruction of Cellulose by Bacteria and filamentous Fungi. U. S. Bu.

Plant Ind., Bull., 266, pp. 1-52.
Osborne, T. B., and Mendel, L. B.

'i4b The Contribution of Bacteria to the Feces after Feeding Diets Free from

indigestible Components. J. Biol. Chem., 18, 177-82.
Oshima, M.

"19 Formosan Termites and Methods of Preventing their Damage. Philip.

Journ. Sci., 15, 319-83-
Pavlovsky, E. N., and Zarin, E. J.

'22 On the Structure of the Alimentary Canal and its Ferments in the Bee (Apis

mellifera). Quart. Journ. Micro. Sci., 66, 509-56.
Fetch, T.

'06 The Fungi of Certain Termite Nests. Ann. Roy. Botan. < lanl.. IVtailmiva,

3. 185-270.
'13 Termite Fungi. Ibidem, 5, 303-41.

INTESTINAL PROTOZOA OF TERMITES. 22~

Portier, P.

'n Recherches physiologiques sur les Champgnons Entomophytes, Paris,

pp. 1-44.
Pringsheim, H.

'12 Uber den fermentativen Abbau der Cellulose. Zeitschr. fur Physiol. Chem..

78,
Pringsheim, H., and Adelheid, M.

'19 I-Yni!<-iH\' .in Zelluloseabbauprodukten. Zeitschr. fur Physiol. Chem.

105.
Ritter, G. J., and Fleck, L. C.

'22 ( 'ln-iiii-.tr- V. Results of an Analysis of Some American Woods.

J. In-!, l Chem., 14. 1050-57.
Schmitz, H.

'19 stii'li- •- "ii tli' P of Fungi. VI. The Relation of Bact Cel-

lul' : Induced by Fungi, with Spiii.il 1 .• to the

Decay of Wood M • Botan. Gard. Ann., 6, 93-136.
Sherrard, E. C.

'22 M.ui! ;cc Cellulose. Science, 56, 431.

Snyder, T. E.

'20 I: ive Adults of Termites. l'i i 'in. So.

\\ .i-h . 22.

Thompson, C. B.

'17 'I In- i ' '•« of the Common Termite / K.«>1..

J'lUtn. Mm |'li.. 30.
'19 Hi.- I ).-\i-l«>|iiin-nt "i the Castes ot Nine Genera aii'l "Ihiitn-n >;••

i.i... 36. 379-98.

'22 I i ;r. Journ. Morph.. 36. 495

Thompson, C. B., and Snyder, T. E.
'20 The Third Form Wingless Reproductive Type of Termites: Rtctiiulii, -•

an . Journ. Morph., 34, 591-'

Waksman, S. A.

'19 St : .ibolism of Actinomycetes. Journ i :.ih . 4. 180-

4

'22 1 irganisms. Absts. Bact.. 6. 265-99; 6, 331-60.

Young, V. H.

•18 cting Inulase Formation in PI. \V»iM.

21 ' 14-33.

THE CELLULAR ELEMENTS IN THE PERI VISCERAL
FLUID OF ECHINODERMS.

JAMES ERNEST KINDRED,
BIOLOGICAL LABORATORY, WESTERN RESERVE UNIVERSITY.

A consideration of the cellular elements in an animal group,
particularly those cellular elements which are present in cavities
derived from the primitive ccelom, should include references to
the origin of these cells and any other factors which may be con-
cerned with their modifications, such as the habits, the powers of
regeneration and the topographical anatomy of the organ sys-
tems of the animals examined. It is generally conceded from
the evidences of palaeontology and comparative embryology that
the Echinoderms of to-day are a fairly ancient group which have
been able to adapt themselves to diverse environmental changes.
The same pentaradial symmetry is present in all of them, al-
though masked in some, but the principal variations which are
characteristic of the classes are concerned primarily with the
character of the body wall and secondarily with the distribution
of the breathing organs. In the Echinoderms we can distinguish
three types of organization of the body wall and breathing organs:
the first type is characterized by a fairly flexible body and diffuse
breathing organs (e.g., Asteroidea and Ophiuroi'dea), the second
by a rigid test and limited breathing organs (e.g., Echinoidea),
and the third by a well-developed muscular body wall and limited
breathing organs (e.g., Holothuroidea). In all of the classes
except the Holothuroidea, the movements of the body are very
sluggish, consequently the oxygen requirements for muscular
activity are very low and a system for rapid transfer of oxygen is
not needed. But in the majority of the Holothuroidea, the de-
velopment of muscle necessitates a great available supply of free
oxygen and a mechanism for carrying this oxygen must be present.
Therefore one phase of this investigation will attempt to corre-
late the appearance of different types of evils in the perivisceral
fluid with the character of the body wall, the distribution of the

breathing organs and the movements of the body as a whole.

228

PERIVISCERAL FLUID OF ECHIXODERMS. 22()

We know that the Echinoderms have great powers of regenera-
tion and may conclude, therefore, that all of the cells in the body
are very labile, but the question arises as to which cells in the
perivisceral fluid are the most generalized. Cells which could be
regarded as the most generalized would be those which are con-
stant in the perivisceral fluid of all of the Echinoderms and which
are observed to have diversified functions; those cells which
under nn < ssity of local needs remove foreign material, wornout
IIH nt- (if cellular origin and which could give rise to modified
Another : •! this investigation, then, is to determine

if there .IP • 1U in the perivisceral fluid, and if so, what

i- the n. it lire of < ti\ ities.

I. M \llkl\L AM)

The material u-e«l in this investigation v. '.lerted in the

vicinity of the I 5 1 '• • ilogical Station, Friday Il.irb"r,

Washington. K< [>n -i-ntatives of the four clas : luhino-

derms found in this (-reused. The following are the £

which have i ..nniiied :

< i ^SS I. .\~ii-i-oi.;.

I . /

nii.

3. 2 icata.

4. :tS.

5. /.' ;.v.

6. Henri: . iscnla.

~. P : aides.

( i tSS 1 1. Ophiuroiilea.

I. C)/>///<'/'/;C.'/N- ''/.

( "i ^SS III. 1 -• hinoidea.

i. > rotus drobachiensis.

j. Strongylo( enlrotusfrandscanus.

tv Echhnirnilnn'ns ccccntricns.

CLASS 1\'. 1 lolothuroidea.

1. CncHnniria japonica.

2. Cucnnniria chronjhelmii.

3. StichopHS californicus.

23O JAMES ERNEST KIXDRED.

Perivisceral fluid from individuals of each of the above species
was drawn from the perivisceral cavity and studied by the hang-
ing drop method. In studying the phagocytic activity of the
cells, a concentrated suspension of finely granulate carmine or
india ink in seawater was used. This suspension was injected
into the perivisceral cavity through a minute opening in the body
wall by means of a delicate hypodermic needle. The amount of
suspension injected varied with the size of the animal, but it was
found that the usual dose sufficient to affect the phagocytes was
8 cc. The injected animals were kept in a live box for a day, so
that there would be time for thorough ingestion of the particles.

The clotting activities of the cellular elements were studied in
drops of perivisceral fluid which had been allowed to stand for
varied lengths of time.

A saturated solution of seawater and methylene blue, another
of seawater and neutral red were made and allowed to stand for
several days before using. The supernatant solution which was
free from particulate matter was decanted off and the solution
used drop for drop with the perivisceral fluid. These stains were
used for intravitam staining in certain phases of the investigation
since they were found to be specific for the vacuoles of the leuco-
cytes.

II. OBSERVATIONS.

i . The Leucocytes.

In his paper on the leucocytes of the invertebrates, Goodrich
('19) called attention to the fact that the leucocytes of Asteracan-
thion glacialis are characterized by the presence of extensive mem-
branous processes of the ectoplasm. He says, "The freely pro-
jecting pseudopodia usually described are either figured from
optical sections of the folded membranes or from cells which have
produced them under abnormal conditions. These pseudopodia
may be present on cells in the fluid withdrawn from the body and
which has been allowed to stand, and are probably derived from
preexisting membranes." Goodrich calls all of the cells leuco-
cytes, making no distinctions.

In Arbacia (Kindred, '21) I observed the formation of a syncy-
tium in the perivisceral fluid by the anastomosis of filiform proc-
esses which had been derived from the membranous flaps of the

PERIVISCERAL FLUID OF ECHIXODERMS. J.U

leucocytes, thus supporting Goodrich's assumption. Theel
('21), however, in a review of his descriptions of the types of
amoebocytes in the perivi-ceral fluid of Asterias nibens and Pare-
chitius miliaris, states that he distinguished two types of amcebo-

es in the coelomic cavity of these forms, the "white or hyaline
plasma-amceboc} -M<\ the "bladder amoebocy These

description- were n< e<l to by Goodrich nor myself althov

they a; if to both of our papers. A< cording

Nieel the i\vn type- "l aincebocytes are very dissimilar, the

" ill. i'l' NT ami i -I" 'not being flattened nor spread out on the

stirfa< e 1*1 the ;Jass when a drop of the fluid is placed on a glass

slide, but remaining thick and compact with the bladder lying in

ml superimpo-ed l.:\ers. On the other hand, the "hyaline

ma-am«i •' " in the fresh fluid are characteri/ed by the

;<>n of : cell body with longer or shorter

ido|l(idi.i. I \presses doubt as to whether or not one

type mav be < ! fmin the other. Following the terminol.

• 1 by ('.oodrirh. I ,nn calling the cells with the membranous
llap^ leucocytes, and regarding Theel's "hyaline pla-ma-amoebo-

;se of the leucocytes from my ol^er\M-

tion-..n the formation of syncytia. Theel ('21 in e\pre--in^r an
opinion as to tin- possibility of morphological changes in such
celN BS H it may be presumed that the character of

tin- MiiTMiindin.; meiliiim may play an important part in that,
and abo\e all mo\i-ment .md relative stillness, the former ]>re-
ventin^ ami tlu- latter forwarding the pro i tran-mntation.

It" for in-ianer .Hi .im.i-bocyte leaves the c«i-luinic cavity in onl« r
to immigrate into tin- ti—ues of the bod\- wall, it mu-t n& ess irily
tin«l<-r^. • < .-Main rh.m^es of form. When a cell passes over from a
pa— i\ e drift to an aetive motion, its primitive globular eonhyura-
tion mn-t be exch inged for another ami accommodated to creep-

ilU about. "

Therefore, thi • nee of two fc.rms of cells in the freshly

drawn drop doe- imt »i neee--ity mean that because of this
occurrence u e are -I- ilin^ with two distinct types instead of an
active and pa— i\c form of one type of cell. It is reasonable to
suppose that these cell- have a cycle of life and that as they be-
come older this change to a more passive condition leads to a
change in form. Since the activities of these cells are comparable

232

JAMES ERNEST KINDRED.

to the activities of the leucocytes in other animals, they are so
termed.

FIGS. 1-14. Active leucocytes of the Echinoderms, camera lucida, X 650. i.
Le plaster las hexaclis. 2. Solaster simpsonii. 3. Dermaster imbricata. 4. Pycno-
podia helianthoides. 5. Pisaster ochraceus. 6. Henricia leviuscula. 7. Evaslerias
troschelii. 8. Strongylocentrolus drobachiensis. 9. Ophiopholis acttleata. 10. S.
franciscanns. n. Echinarachnius eccentricus. 12. Cucumaria japonica. 13. C.
chronjhelmii. 14. Stichopus calif ornicus.

Text-figures 1-14 are camera lucida drawings of the active
phases of the leucocytes (" bladder amcebocytes" of Theel) of the
whole series of Echinoderms studied. Examination of these
figures shows the general morphological similarity of these cells
to each other, the only distinct difference being that of size, which
varies from 7-14 microns in endoplasmic diameter. In all of the
cells it will be noted that the ectoplasm is clearly marked off from
the endoplasm and is produced into a varying number of rapidly
changing delicate flaps. These flaps are constantly being with-
drawn and extended and may be regarded as modified pseudo-
podia. By means of these flaps the leucocyte progresses slowly
through the fluid. That the surface of the cell is covered with a
sticky fluid is evidenced by the manner in which particulate
matter adheres to the flaps. When the flaps are withdrawn the
particles which adhere to them are ingested.

PERIVISCI K AI FLUID OF ECHIXODERMS. 233

The endoplasm of nearly all of the active leucocytes is granular
and opaque, the exceptions to this opacity are found in the leuco-
cytes of Ophiopholis acnleata (Fig. 9) and Cucnmaria japoi:
(Fig. 12 i. In these leucocytes the nucleus- with its content of
large chromatin granules is easily discernible. The endoplasm
tonally contain- a varying number of hyaline vacuoles which
-tain blue with the methylene blue-seawater solution and red with
the neutral red-seawat ition.

The active leiiro, vie i- a phagocyte and is always found loaded
with <annine or India ink particles when those subMamv- are
introdured into the peri\ i-< eral cavity. In the Holothuroidea
the-e pha^' re "b-erved to deposit the india ink parti*

in the skin (Srlmlt/. . \werinxew ('n) considers that the

r of th- in M.ine of the Echinoidea is due to this pliago-

cytic activity of "aiiio •••>" and that the color difference in

\aiietir- df .s'. drobachi is dependent upon tlu- color of the

fond, the pigments of which are taken up by the "am<eboi-\ •
and i arried to tin- -kin. Since the leucocytes are the only type-
of < rlls in the peri\i-vral tluid which are pi tic, they

\\ithoilt doubt the j>h :c "anio?lx)C\ activi

are <!<•-' iii'< <l b\ these and "ther authors. A further function of
the pli the removal of germ cell- remaining in the

j;onads after the bulk of the ^ i metes have been >hed (Caullery
and Siedlii-ki. \\\ . ( 'erno\ odeanu and Henri (\>(>\ <»b>erved
that bai teria inj( into the body cavity of sea un liin- W

taken up 1>\ "ain-r: h long pseu<lo|)odia, '\-lls which

are donbile-- K-ur.u\ti ^.nce mention ha- been made above
of the relation < ,f the lei; • s to the transfer of f. M id i)roduct>, it

i- pertinent to en«|uiri- arther obserx at ion- ..f their |»arti( i-

palion in tlii- pha-e of \it.il at tivity.

("uenot ('<)ib . one of the tirst to call attention to this relation,
a--nmed that the sub-lance- passed from the intestinal cells into
the inte-tinal lacun.e are taken up by the "amoeboc> ind

stored up in them t<> be « arried to other parts of the body. The
" amoebocytes " which are so com erned become metamorphosed
into "amoebocytea with spherules." Frenzel ('92) was of the
opinion that the " anwebocytes " pu>hed between the epithelial
cells of the intestine and into its lumen where they disintegrated
anil their remnants acted as a digestive ferment. Enriques ('02)

234 JAMES ERNEST KINDRED.

thought that digestive substances were carried from the rete
mirabile peritoneum in the Holothuroidea to the stomach epithe-
lium by " amcebocytes. " Thus there are several conceptions
concerning the relations of the leucocytes to the digestive activi-
ties which need further investigation.

As evidenced by their tendency to remove foreign particles
from the body cavity the leucocytes may be regarded as excretory
agents and further observations as to their excretory activities
should be considered. Durham ('88) and Chapeaux ('93)
observed that the phagocytic cells leave the body through the
papulae in the Asteroidea. I have observed such a migration in
Leptasterias hexactis, in which, after injection with carmine, the
papulae are reddish and a smear from the outer surface reveals a
number of leucocytes laden with carmine particles. "Amcebo-
cytes" (particular type not stated) have been observed to leave
the body cavity of the Holothuroidea by diapedesis through the
walls of the branchial tree into its lumen and thence to the out-
side (Herouard, '95; Schultz, '95). Therefore, there is evidence
that the exit of the phagocytes in the Echinoderms is through
the body wall.

The exact relation of the leucocytes in the removal of waste
substances from the tissues has not been proven, but Delage and
Herouard ('03) thought that substances absorbed from the tissues
by "amcebocytes" are reprecipitated in them in the form of
granules and may possibly give rise to the various "amcebocytes
with spherules." List ('97) pointed out earlier that substances
absorbed by the "amcebocytes" may be the cause of the develop-
ment of a crystalloid in the nucleus of cells of this type, which by
growth causes a degeneration and finally the destruction of the
cell. Thus the accumulations of crystalloids observed scattered
throughout the body of various Echinoderms may be regarded
as the remnants of degenerate excretory "amcebocytes."

Another activity of the active leucocytes is the formation of
plasmodial masses which are very numerous in any drop of peri-
visceral fluid. That these plasmodia are formed by the fusion of
active leycocytes has Uvn <>l>~cr\ <•<! frequently and plasmodial
formation is one of the activities which distinguishes the active
from the passive phase of the leucocyte and from other cells in the
perivisceral fluid.

PERIVISCERAL FLUID OF ECHIXODERM-. 235

In addition to the active leucocytes there are always present,
even in a drop of freshly drawn pern i-ceral fluid, numbers of
flattened cells with filiform processes which float pa>- ively
through the fluid f FL-. i =; 22). The endoplasm of these cells is
granulated, opaque and \\icuolated. is the endoplasm of the
active leucocytes and it reacts the same to methylene blue-sea-
water or neutral red--eawater solutions, i.e., the vacuoles stain
blue and red respectively. Since I have ol»er\ed active leuco-

: 1 till- Ki-llilli.lllTIIl-:. i.ini'-I.l luri'ki, X '

15. / •-.•iuscula. i :f>">iii-ii. 18.

•'•r simp-
:ta.

cytea • hani;inL; into the-e in .\rbac i>i. I think that they are pa>-ive
phar-e-. of the leiu oc\ i, -. Tlieel re.^anU them as di-tiiut in
tlu-in-eU'e- and 'ill- them "hyaline pla>ma-am<ebocytes. "
Their pre-eiuv in the fre-hly drawn fluid would indicate that they
are normally prc-etit and that their occurrence is not due to
abnormal conditions .1- •'.-•odridi suggested. Although these
cells may occur Miigly. they are most frequently met with as

236 JAMES ERXEST KIXDRED.

elements of a synrytium and it is this syncytial formation which
is their most valuable vital activity, in that it is concerned with
the repair of injury and the replacement of lost parts in the
organism. This is due to the fact that these syncytia close
wounds and form the basis for the growth of the other tissues
and further that these cells themselves may take part in the for-
mation of other tissues.

Theel ('21) states that there is a fibrin-like substance which
occurs in the coagulation of the ccelomic fluid in the Holothuroi-
dea in addition to the meshwork of the fibers formed by the ar-
borizing leucocytes. That this fibrin-like substance is extra-
cellular is evidenced by the fact that a fibrous meshwork occurs
when there are far too few cells to form such a meshwork so
rapidly, by leucocytic syncytia alone. I observed this super-
fluity of fibers in the drops of perivisceral fluid of all of the prepa-
rations of fresh material and was at a loss to account for it. The
origin of this fibrous substance is not known, although it has
been suggested by Schafer ('83) that it might have been secreted
by the leucocytes. That it does not coagulate in the fluid in
the perivisceral cavity is due to the various ciliated cells which
keep the fluid in motion (Cuenot, '01). This movement of the
perivisceral fluid does not prevent the formation of plasmodia or
syncytia, but it does seem to inhibit the fibrin-like coagulum.

From the observations of Theel and others it is well known that
in the development of the test of the Echinoidea the leucocytes
form syncytia within which is secreted (intracellularly) the
spicules which fuse and form the stereom, a definite beam and
rafter skeletal structure. Leucocytes containing spicules have
been observed in the perivisceral fluid of various Echinoderms.
I have observed them in the perivisceral fluid of Henricia levins-
cula (Fig. 16). Theel from his observations on the occurrence
of these cells containing spicules is of the opinion that we may
presume that the skeleton of the Echinoderms is due to the
activity of migratory " plasma-amoebocytes " and their syncytial
fusion. This assumption may be true, but there has been no
evidence presented which shows that these cells (my leucocytes)
are concerned in the replacement of resected areas of the stereom .
Hence in order to determine the scleroblastic activity of the leu-
cocytes a series of resections of the test of S. drobachicnsis were

PERIVI5CERAL FLUID OF ECHIXODERMS.

237

made and the subsequent regeneration observed. But before
discussing the observations the relationships of the stereom and
the adjacent region should be described. Thus in a section

pr.

r> -

st

>rg.mbr.

nc.sy.

••

l-i R :i. Slron Jruhia. •. vertical

i. X 75-

!•'[«.. j j. l>i-t.n! . : :• ;ifuluiii . membrane of bod; • .ill >lio\vn in

Semi-diagrami

!MK'. J.s. Detail . •'. n in l-"m. 2.5. S-mi-liugrammatic.

< 650. am idth -I'li'-iuliV; cp., i-pidermis;

la.. l;u 11: : nn; pr.. \>T< .A area; pt.. peritoneum;

rg.mb.. n-yi-ii- yncytium; trab.,

trabecula.

through the bi'.h \\.iil "t .S. drobachiensis which had been decalci-
fied ami -laiiu-d the following relations of the stereom and the
surrounding ti— tie- may be observed. The epidermis (Fig. 23)
appeal- nn the outer surface of the section and beneath this there

238 JAMES ERNEST KINDRED.

is a thin layer of syncytial reticular connective tissue (pr.} which
is directly continuous on its inner surface with the definitive
stereom (s/.). This syncytium is loosely organized and has
large intersyncytial spaces within which are found wandering
" amcebocytes with spherules. " The cytoplasm of the syncytium
is fibrous in appearance and the remnants of spicules could be
observed in it. The traces of spicules were more marked in the
region bordering on the stereom where the syncytium had the
trabecular organization characteristic of the stereom. This
region I have designated as the prestereomal area. The stereom
is made up of I-shaped trabeculse which are apparently joined to
each other end to end, so that the whole stereom is a framework of
beams and rafters with regular lacunar spaces. The substance
of the stereom when observed in the unstained condition is clear
crystalline in character and when stained is intensely basophilic
in reaction. A prestereomal area is also found between the
stereom and the peritoneum. From the organization and rela-
tion of the prestereomal areas to the stereom it is evident that the
growth of the stereom takes place peripherally by the gradual
deposition of skeletal material within the prestereomal trabeculae.
Nuclei with prominent nucleoli were observed in the prestereomal
trabeculae, but none were observed in the trabeculse of the ster-
eom. This condition would indicate that trophic activity of the
cells is lost in giving rise to the stereom and that the whole cyto-
plasm of the syncytium becomes converted over into skeletal
material while the nucleus degenerates. Now if a cut were made
directly through the body wall and a piece of it removed, the cut
surface would present three regions, a middle stereom region and
two peripheral prestereomal areas. It is obvious that the pre-
stereomal areas would be capable of replacing certain parts of the
test, but the question arises as to whether or not these cells are
aided in this regeneration by the leucocytes.

Although the body wall of several specimens of S. drobachiensis
were resected in an attempt to answer this question, the results
are far from convincing and the description. of the regeneration of
the test which follows is to be studied more in detail at a later
date.

Eight specimens of S. drobachiensis were injected with carmine
in seawater through a minute perforation in the peristomial mem-

PERIVISCERAL FLUID OF ECHINODERMS. -"39

brane and were allowed to remain in a live box for twenty-four
hours, so that the leucocyte- would have a chance to take up the
c.trmine particle- Al the end of this time, a piece of the body
wall, I cm. square, was removed from the aboral surface of each
-penmen. The -imens were then put into a live

and o! tions were made over a period of four week- upon

the < hanges which were taking place in tin ' ed area. When

! it was noted that a membrane v. dually

the opening in the body wall. Thi- membrane \\a- icd-

dish in color as conti -.\ith the green i-h color of the -urround-

i.ody wall, and : I llv from the margin of the open-

ing BO that tin- dimini-hed slowly in diameter. In -even

oft ening in the body wall was 1 in

t\\o \\eek-. At tir-t the membrane closing the opening was very

delicate, but : lally became firmer ami in >evcral -pi'cimni-

Jiriicil. 1'lii- ph. i- tou^henin:^ of the mrmbrane

iin-.l dining the third \vrck. It was then noticed tint >krlr-

tal mairrial b»-^aii • '11 the margin of the membrane

which-/ with the original cut surface. Purin^the

fourth \\t-ck tin- df|)«,~iti.)ii nf -keletal matt-rial \\cnt furtlu-r and
ral individuaU -ho\\cd a portion of reK«-n»-ratrd test. At
thi- time tin- ti--iu- ;>enin^ was remo\ed lK,iu -« -\rral

-pc. iiut'ii-. spread OUt on a nd studied in tin- frr-h condition.

Tin- wholi- ma-- liad a crim-mi color and \\ ' i tan^lrd hbrou-
COnsistency in \\hich \\CP irent leu« • •ntainii

miiu- granule- \\hich had fornu-d more or It--- « if a -yncytimn. in
\\hich the cell boinida: ndistinct. Se\eral piece- i.f the

membrane and the adjacent body wall were pi ion-

iu- in order to determine the relation of the cell- to the repl
inent of the -kelet.in. Thc-e pieces wi ; •• « 1« •• ilcified and a series

of -ei tiou> made.

1 i^uic J^ i- a semi-diagrammatic drawing of a vertical -ection
through tin- regenerated membrane and a portion of the adjoin-
in- body wall, the -tructute of the latter having been disi'iissed
above In tin- membrane thei- • gions of ditlereti-

tiation. The region m«'-t di>tal to the te-t and forming the
center of the membrane is thick and reddi-h in color, thi- color
beiiu due to the |>re-eiK'e of minute particles of carmine in the
cells which make up the -\iic\tial reticulum of the membrane

24O JAMKS ERXEST KINDRED.

(Fig. 24). The color diminishes laterally and also in the region
proximal to the stereom, where the syncytium is trabeculated in
the same manner as it is in the prestereomal peripheral areas
(Fig. 25). Since the leucocytes were the only cells in the peri-
visceral fluid which were observed to be phagocytic and form
syncytia it is probable that they aid the prestereomal cells in the
formation of the membrane and gradually develop skeletal
material for the formation of the stereom. The cytoplasm of the
reticulum in the membrane in addition to the carmine particles
contains fibers which are probably the remnants of decalcified
spicules. The nuclei of the reticulum are round, have a distinct
nucleolus and are the same type as are present in free leucocytes
observed in the lacunae of the stereom and in the prestereomal
trabeculae, so that it is probable that the leucocytes and the con-
nective tissue cells of the prestereomal area are of the same series,
except that one has become specialized for the production of the
stereom under ordinary conditions of growth, whereas the leu-
cocytes only take over this function when the body wall is in-
jured. The only evidence for a line of demarcation between the
two is in the presence of the carmine particles in those cells which
make up the reticulum of the membrane. It is therefore evident
that the membrane is formed by both the multiplication of the
prestereomal connective tissue cells aided by the anastomosis and
syncytial formation of the leucocytes which make up the bulk of
the membrane. Within the spaces-of the reticulum of the mem-
brane are found large numbers of "amcebocytes with spher-
ules" which probably carry nutrition to the cells of the syncy-
tium, enabling them to carry out their scleroblastic function.
These "amcebocytes with spherules" are very few in the region of
trabecular formation adjacent to the stereom and are entirely
absent from the lacunae of the stereom, thus they seem to be
massed in that region where repair is going on rapidly and tin-
cells of which are being differentiated .

Thus in brief there are three regions present in the regenerating
area. First, the syncytial region which forms the bulk of the
membrane which has closed the opening in the body wall and is
composed of a syncytium of leucocytes with small lacuna •. con-
taining large numbers of "anxrliorvte- with spherules." Sec-
ondly a prestereomal area, definitely trabeculated, with large

PERIVISCERAL FLUID OF ECHIXODERMS. 241

lacunae and continuous with the peripheral prestereomal areas
which enclose the stereom of the adjacent region of the body wall.
The "amoebocytcs with spherules" are very few in this region.
Thirdly, the definitive regenerated stereom which is crystalline
in character, devoid of nuclei and in the lacuna- of which there
are no "amcebocytes with spherules."

It i- obvimis that the cellular elements which form the mem-
brane have either been derived directly from cells in the region
of the pre-tereomal area- or from cell- of the peri\ i-ceral fluid
which ha\e migrated to tin- ...iva and formed -yncytia. Since
the leu 'lit- only cells of the peri\ i-<vral thiid which

have l.ft-ii ob-rrved to form -yncytia and since they are the
COenCXT initial wandering mt-eiichym.il cell>, then is

on t" suppose thai they h.i\ t- -on ie part in this process. That
com formation doc- not nece— ai ilv -tart from the cut >ur-

idenccd by t hi- appear. ince of indc-

pemlt-i illation in the membrane I

A more detailed -ludy of ilu- n l.iiioii^hi| >- of the K-u- -to the

Deration of tin- te-t i- to b«- more closely followed in a more
eMi-ntlt'l ' riim-i

These observation • • neral \\ith Tin > :iciu>ion^

that tl. obl.i-tic cell- "oii^ht to \ • 'K'd .inion^ the true

pla^ma-amo i 3, though t!n-\ are di-Minilar ,i- to their

functional manife>tation-. an<l t oii-e<|iient Iv, too, as to their
i hentic. il ( on-titution. !• i unlikely that their (]iialit\ ot

taking ii]> a >ii])ertluit\ of inor^.iiii. -.ilt- from the Mirroimdini;
medium ha- inlhu-nced the cell-, their power of anneboid move-
ment haxill- been Ml|)|ire--ed or limited." Me conclude- that
the direction- of the p-eiidopodi.t predetermine the characteri-tic
piotru-ioii- of the cr\-tal- in I'^olns p>h<i utapiis. The cr\-ial
wlu-n formed Occupies the whole of the cell except for a -mall -pace
occupied by the nucleii-. In the formation of the >tereom in
6'. ilrobachicnti* a -imilar predetermination of the or^ani/ation
of the skeletal element- is apparent in the pre-tereonial area-.

I'rom the above observation^ it may be concluded that the
leucocNte- in both the active and passive phases are those celU
which are of conMaiit occurrence in the Echinoderm-. The\' are
phagocytic, thrombocytic and possibly scleroblastic. These
cells are omnipre-ent in all regions of the perivisceral cavity.

242 JAM1.S KRNKST KIXDRED.

From their activities they must be recognized as the most gen-
eralized cells in the perivisceral fluid since no others have been
observed whose functions are so diversified. The question of
their origin is a vexing one, although it was once pointed out
th.it the dorsal organ may, in the Asteroidea, be an organ of
leucoblastic function, a function which Cuenot ('oi) denies this
organ, claiming that it is more likely an organ for the elimination
of wornout cells of the perivisceral fluid and that the "amcebo-
rytes" (general term) are probably peritoneal in origin and may
also arise from each other. In all probability the number of
leucocytes increases rapidly when an animal is injured, particu-
larly in the region of the injury, where they form the framework
for the regeneration of other tissues. They may be regarded as
one of the important agents in the replacement of lost parts in the
Echinoderms.

FIG. 26. "Amcebocyte with red spherules," Strongylocentrotns drobachiensis.

cam. luc., X 1300.

FIG. 27. "Amcebocyte with red spherules," S. franciscanus, cam. luc., X 1300.

FIG. 28. "Amcebocyte with red spherules," Echinarachnius eccentricus, cam.
luc., X 1300.

FIG. 29. " Amoebocytc with yellow spherules, " E. eccentricus, cam. luc., X 1300.

2. The "Amcebocytes with Spherules.'"

There are two groups of "amcebocytes with spherules" present
in the perivisceral fluid of the Echinoderms. One group of these
is characterized by the presence of pigmented spherules in the
cytoplasm and the other by colorless spherules. In both types
the spherules fill the cytoplasm to such an extent tli.it the nucleus
appears merely as a light space in the center of the cell. All of
the "amcebocytes with spherules" are further characterized l>\
the presence of very blunt pseudopodia and when fluid con-
taining them is allowed to stand they tend to assume a spherical
shape.

Amcebocytes containing red spherules were observed in the

PERIVISCERAL FLUID OF ECHIXODERMS. 243

Echinoidea alone, those containing yellow >pherules in the Echi-
noidea and some of the Holothuroidea, and those containing color-
less spherule- in the Ophiuroidea, Echinoidea and Holothuroidea.
In the Echinoidea, the "amoeb' with red spherules" were

numerous and largest (12 micnm- in diameter) in S.francis-
(Fig. 2~>\ in S. drobochiensis i Fig. 26) they were

numerou- and -mailer (9 microns in dian In }:..hinarach-

jiin ley wt • less numerous and -mailer

[0 mi< rons in diamei.-r . The pnd< miinant type of pigmented

ani"i te in / -'.ricns wa- the type with yellow -pherulc-

,!]d the-e wnv ji M licularly mas-ed on the peritoneum

(if the inte-tine. S d yell' mentcd amcel « •« \ If- \\ ere

ol,-.T\.-d in ,< but ii" . iatioii with .\

•t-m \\a- not-

Tl:<- K"| puincnt in the pi^meiiird amoebocytCS of tlie lu
noide.i \\a- de-i^nati-il c. hii n- by M.Miinn (.'t^.SL The

\ ii-w «.f tin- author and < .ninth- I '.^7 < that the amcebocytes «>n-

t. lining (•• hinnchi-iii' •! with u transportation

ha- lle\»T been fill! pled. ( Uellot ('gib Wa- the lir-t t«»

(,j-p..-e thifl a — nmjitioi) .uid Stated that there was n.. change in
the depth "•! the ( • 'lor when the cell- Were allowed to -land ill the

air, and that the contained pigment instead of being an oxygen-

Carryinj • d -up food material which the (ell- had

taken from the inte-tine. Further, \Vinter-tein -Imwed that a
-olutioii of ecliin.ichrome doc- !)• up more «>\\ -en th.m the

-anie amount of -i-a\\ater. Tlii- a--ertion of \\inter-tein'- i~

•liticaut. because it open- up tl . -tion a- to the ,,\\

ie(iuirenient- ot the I'diinodi-rin-. Of course ii i- ol >\ i..u- that
all of the free OXJ gen u-ed by the l;.« hiiioderm- i- taken from the
Beawatei and further it ha- been -hown that except for a ililler-
etice in albuminoid content the periviMvral tluid of the l:.chino-
derms i- the .-a me den-it\ a- the -ea water a'tieiiot . '«)!). There-
fore. \\e can a — nine that there i- a dilfu-ioii of the si-.iwater
through the breathing organ- of the Fcliinoderm- and that the
oxygen c..ntent i- the >ame in the peri\i-ceral tluid .1- it is in the
out.-ide seawater. \Yith tht 'tioii of certain of the Holo-

thuaiide.i. all of the Fcliinoderm- ha\e a very slight develop-
ment of rapidly contractile muscle elements and hence no need
fora large amount of oxygen for muscular activity, from which it

244 JA.MKS KRNEST KIXDRED.

follows that in the Asteroidea, Ophiuroidea, Echinoidea and
Holothuroidca with tests (e.g., Psolns), which live in regions of
high free oxygen content, due to tidal currents, there is
enough oxygen in suspension in the perivisceral fluid for ordinary
metabolic activity. Therefore in these forms no oxygen-carrying
cells are developed and the cellular modifications which occur in
the cells of the perivisceral fluid may be regarded as independent
of the relations of the breathing organs. Further we would
expect to find efficient oxygen-carrying cells in the Holothuroidea
which have a relatively high development of muscle, for the
needs of which the free oxygen content of the perivisceral fluid
is not sufficient. The hemocytes are the cells which fulfill this
requirement in the Holothuroidea and will be discussed later.

If the "amcebocytes with red spherules" are not to be regarded
as oxygen-carrying cells, are they then related to the transfer and
storage of food as suggested by Cuenot? As a partial answer to
this query are the results and conclusions of Awerinzew ('n)
who carried on investigations on the habitat and food relations
of two varieties of S. drobachiensis . Awerinzew observed that
the two colored varieties of this species lived in different types of
environment, the green-yellow forms living on a mud and stone
bottom and the red forms amongst red algae. He assumed that
the pigment in the food of the red forms was carried from the
intestine to the skin and deposited there. The pigment in the
prrivisceral cells would then be due to the food eaten. He
checked his results by injecting carmine particles in solution into
the alimentary tract and found that these particles were carried
to the skin, but the distinctions between the types of cells en-
gaged in this activity were not made clear, so that it is possible
that the cells were leucocytes carrying on their normal activity as
phagocytes and there is no case for the pigmented cell- as food
carriers .

The "amcebocytes with red spheruk-s" are far more numerous
and larger in the perivisceral fluid of S. fraud scan us th.m thry are
in S. drobachiensis. This fact leads to the suggestion that the
color of the body wall is due to a difference in numbers of the red
cells in the two species.

Since it has not been proven that tin- " amoebocy tes with red
spherules" are developed from other cells by the ingestion of

PERTVISCERAL FLUID OF ECHIXODERMS. 245

pigment from food, it may be that they are the descendants of
the pigmented cells (chromatophores) of the larval Echinoidea.
These cells are present in the segmentation cavity of the larval
Echinoids, but I have not found any reference as to their occur-
rence in the larva? of other classes. If this is true, then we are
dealing with an amoel" which is specific in the Echinoidea

and may yet be found lo be derived from the colored -ub-t.mce-
characu-ri-tic of the Kchinoid ovum.

;. ".V .). 30.

-

,. ('

" Amu 1 \\ itii colnrle les" (1 re more

widely di-tribuied in the Kchim -derm- lli.in tin- pum« -uted ones
.ind are loimd in the < )phiuroidea. lu'lnnoidi-.i .uid llolotlm-
roidc.i. "riu-frll^of tin- i\ [»r ,irr much mmv abundant and .iriivi-
in the I lolotliiiroidi-.i than in thr other classes. Th«.-y an- al-o
much lr— >tabk-aud i li~inti-^raic -01 >n after withdrawal from tin-

bod\-. Tlu-x arc drr|ily stained b>' nu-tlu'lene blur--ea\vatri- or
iH-uiral red--ra\\aH-r -olution- ction \vhich i- not i haractcr-

istic of the "amoeb with colorle-- -plu-ruli--" in the other

classes.

It is to be noted that the -i/.- of the >pherule~ i- constant for
any i;i\ en am»d'oc\ ie, but that there is a X'ariatioii in the H/e- of
the spherules of amoebocyte- characteristic <pf dilti-rent animal-.
Theel has also remarked upon this and ha- pointed out tli.tt the
"amocbocytes with colorless spherules" predominate in the
I lolothuroidea which he studied. He says: "In the holotlmrid-
the white corpuscle- an- thoroughly predominant, though, tl

246

JAM1.S I KNKsT KlMiRKM.

also. -pcries are to be met with which are characterized by both
white and red corpuscles. Thus, for inst., Lalnt/oplax buskii
makes an exception by lodging cells with hyaline granules to-
gether with such ones which contain red pigmented granules,
both kinds being nearly equal in number."

The origin and functions of the "amoebocytes with colorless
spherules" is problematical. According to Cuenot ('Qib) they
may arise from leucocytes by the addition of spherules of protein
and are hence to be regarded as food carriers.

42

41

43

FIG. 38.
FIG. 39.
FIG. 40.
FIG. 41.
FIG. 42.
FIG. 43.
FIG. 44.

Hemocyte, C. japonica. Side view, cam. Inc., X 1300.
Hemocyte, C. japonica. Surface view, cam. luc., X 1300.
Hemocyte, C. chronjhelmii. Surface view, cam. luc., X 1300.
Vibratile corpuscle, Stichopus californicus, cam. luc., X 1300.
\ 'ibratile corpuscle, Ophiopholis aculcala, cam. luc., X 1300.
Yibratile corpuscle, Strongylocentrotits drobachiensis, cam. luc., X 1300.
Yibratile corpuscle, S.franciscanus, cam. luc., X 1300.

3.' The Hemocytcs.

Theel ('21) describes a series of "red blood corpuscles" from
the perivisceral fluid of the Holothuroidea, excepting the forms
with tests, as Psoltis. In his discussion of these cells, he calls
attention to the fact that the content of these corpuscles was
first noted by Semper ('68) who suggested that this material was
of the nature of hemoglobin. Howell ('85) advanced the s.mie
suggestion with regard to the same type of cell in certain species
of Thy one and Cucumaria. Cells of this type were also observed
in the Holothuroidea by Cuenot ('91 1>>. Knoll ('93) and Kollman
('08). These cells were described as spherical or elongate ovoid
cells with a definite limiting nn ml UMIIC which was elastic, but did

PERIVISCERAL FLUID OF ECHINODERMS. 247

not form pseudopodia. The cytoplasm was of a homogeneous
color of the same shade as hemoglobin, and in it was embedded a
slightly ovoid nucleu-.

In my observation- on the- genera Ciicuniaria and SlicJiopns I
have found these cells to be limited to the Cucumaria. For
re. i-< Hi.- Liiven below I h,r Us as hemocytes.

The hemocytes of Cucumaria are flattened, biconvex di-c-, ovoid
in shape. The cell nieinbraiie i- pla-tic. but the cell- exhibit no
aniu-boid movement and ried pa--i\ely ill the perivisceral

tluid by contraction- of the aid movement- of the body.

The . \ top], (-in i- a hoino-eneon-, ma~- o| . , -\ ellow color in

which i- located a -mall oval nucleus, eccen t rically placed (I

\o). In ' tu the granular content of the nucle-

us i- very clearly apparent I ig. 401. A mass of t: :ls pre-

sents a crimson appearance < on i parable tot! 1 blood

of vertebi The hen I C. jaf>o>: red bo.ly wall)

are much larger and ui<>re mmien.us than tho-e of ( '. chronjhclmii
<\\ hit c body wall . « < >n-e< |iiently the deep Color of the former ma\-
:ue ill part i" thi- dilferencr in numl

^.lnll(•l llr\di '22 , although giving no reference to the earlier

suggestions that l' • !U coulaiu ' 'oliin. can'icd on a

Series »f experimeni- on tin- content of heiuorx te- in Thyone
briarcu^. \\< - Union obtained from tlu-e celU to

the >pectr and oblained a band character!-! ic of ox\ hemo-

globin; ii|ion icduciioii tin- solution gave the >in-K b.iiid charac-
ter i>tic of hejiio^lc.bin and when shaken \\ ith air, the double band
characteristic of o\\ licino-li ,bi;i a]iprared: hemin-likt cr\ -taU
\\cic obtained from the content of the hemoc\ tCS, 'I'he-e eX|»eri-
nu-nt- together \\ith -i-\t-ral other clu'inical te>ts have lei Van
der He\de to i "iiclude that the Mib-t.mce i> hemoglobin. As 3
re-lilt of these ob-er\ation- I ha\e di>i^nati-d the-e cell-. t>ju-
of \\hich al-o occur in Cncnmariti, the lu-moc\te-. a term which
is briefer and more conci-e than the appellation red blood cell-.
YYe may a^k if there i- any rea-oii why Oxygen-carrying < ells
should be present in certain of the Holothurnidca au<l not in
other members of thi- cla-s or other Kc-hinoilerm classes. 1
have surest ed earlier in thi- paper that oxygen-carrying cell-
may be regarded as unnece->ar\ in the . \-teroidea, Ophiuroi<lea
and Kchinoidea because of the relati\el\- lo\\ ox\ -en need- of the

248 J A MI'S KKNKST KINDRED.

body which can be supplied by free oxygen of the perivisceral
fluid. But in the Holothuroidea of the- types like Cucunmria
and Thy one, the body is made up of highly contractile muscle
elements which may possibly need more oxygen than can readily
be supplied by the perivisceral fluid. Consequently, it is sug-
gested that the hemocytes, as oxygen carriers, have appeared in
association with the development of the muscular system and
greater muscular activity, ?o that the contractile elements may
be supplied. Further evidence for this suggestion is found in the
absence of these hemocytes in the Holothuroidea without mus-
cular bodies (e.g., P solus] which are comparable to the test-
bodied Echinoidea.

4. The Vibratile Corpuscles.

The vibratile corpuscles are those cells in the perivisceral fluid
which are minute and bear flagella. Cells of this type were
observed in Ophiopholis aculeata (Fig. 42), 5. drobachiensis
(Fig. 43), S. franciscanus (Fig. 44) and Stichopus calif ornic us
(Fig. 41).

In 0. aculeata the vibratile corpuscle is very minute (3 microns
in diameter). A small peripheral layer of granular cytoplasm
encloses a relatively large nucleus. A single long flagellum is
present at one end of the cell. The vibratile corpuscles of
Strongylocentrotus are much larger, each has a relatively smaller
nucleus as compared with the amount of cytoplasm and a shorter
flagellum than the vibratile corpuscle of 0. aculeata.

In S. calif ornicus , the vibratile corpuscles are very minute
(i micron in diameter) and are colored with a yellowish pigment
which obscures the nucleus, if such be present. Since hemocytes
are lacking in this species it may be assumed that these cells are
oxygen-carrying cells and that the pigment is a hemin compound
related to the content of the hemocytes of the other muscular-
bodied Holothuroidea.

Cuenot ('02) suggested that the function of the vibratile cor-
puscles is to keep the fluid content of the perivisceral cavity in
motion, thus aiding the ciliated peritoneum in causing a circula-
tion of the fluid.

PERIVISCERAL FLUID OF ECHINODERMS. 249

III. SUMMARY.

From a comparative study of the cellular elements in the peri-
\Nceral fluid of a repre-ent.:- jroup of Echinoderms found in
the vicinity of the Puget Sound Biological Station the following
conclusions have lu-cn reached:

1. The leucocyte- are < on-tant in the peri\ isceral fluid of the
I ' liinodrnn-.

2. Tin; li-in - in all of the 1 .< him idrnii- >tudied are phago-
cytic and thromboblastic and in -< •! ippear to be sclero-
Ha-iic and a-soeialed \\ith • • nieill of re-ectrd -ke'.etal

IS.

;v llcni' IN with IH-IIK^I..!.!!! . are found only in cer-

of the I liilnihuniidi-a and ted with the de\ rlo])-

iiirnt of a highly mu-t ular 1»><1\ .

4. Modilii alion> ol tin- 1'icalhin^ organs ha\t- a] •] lan-nt ly no

"ii llu- cellular (fntci, be 1 •«•! i\ Nn-ral fluid, so that

no spi-( ih. 'It-M-lopi-d in tho-i- forms

with a rigid noii-nniM. ular 1«><1\ \\all dr.-pitr the limitations of the

breathing

5. < >f tllr ".HlKll with S|ilnTlll< ~, " tlloM- \\itll Cololi'

s|)lu-i ulr- an- pu--»-iit in tla- Ophiu: I .1 hinoidra and Holo-

tlmroidra and arc prrdoininaiit in the l.i-t « 1 "Amn-liocyti->

uiih red -phi-ri. ' in tin- 1 '.< hiiioidra, whrre the

>i/e .Mid numlu-r pie-ent i.- correlated with the depth of color of
the body \\all.

6. YiluMtile corpu>rle- are pn-ent in the Ophiuroidea, I'.rhi-

noidca and in >'//</. <>n;itn* "alone of the Hnlothumidra

studied. In the latter -; • he\" are pi^nieiited and are re-

led as « i-lN \\hich function a- do the hemocyir- in other
mu-cular-liodied 1 Joint humid'

A. KN«)\\I I DGM] NFS.

I \\i-h to take thN op]mrt unit\" to e\|>re» m\- thank> to the
Ohio Aiadeim of Seieiiee for a re-e.irch grant which aided me in
carrying on thi> in\ < -i i^ation.

Awerinzew, S.

'ii (jber die pigmente von S. droba Arch. Z. EXJH.T.. \'III: i \iii.

Bush, M.

'21 Rrxisni Key to the Echinoderms of Friday Harbor. Publ. Pu.yt >•!. Biol.
tii'ii. Ill: 65-78.

250 JAMES ERNEST KINDRED.

Caullery, M., and Siedlecki, M.

'03 Sur la resorption phaijorytaiiv des produit- yiih.iux itiutilio-s i-lit-x 1'Echino-

cardium cordatum Prim. ('. R. Acad. Sri. I'.m-. 137: 496-498.
Cattaneao, G.

'91 (,li amoeboi iti <i<i < '• •falopodi e loro confronto con quelli d'altri inverte-

brati. (".. Atti. Univ. Geneva. 50 pp.
Chapeaux, M.

'93 Sur la nutrition des Echinoderms. Bull. Acad. Belg., 26 : 227-232.
Cernovodeanu, P., and Henri, V.

'06 Phagocytose chez les Oursins. C. R. Soc. Biol. Paris, 60 : 882-884.
Chun, C.

"90 Die Bildung der Skelettheile bei Echinodermen. Zool. An/.., 15 : 470-474.
Cuenot, L.

'pia Etudes Morphologiques sur les Echinoderms. Arch. Biol., n : 303-680.

*9ib Etudes sur le sang et les glands lymphatiques dans la serie animale. Partie
2, Invertebres. Arch. Zool., ser. 2, 9 : 613-641.

'01 Etudes physiologiques sur les Asteries. Arch. Z. Exper., 9 : 233-259.

'02 Organs agglutinants et organs cilio-phagocytaire. Arch. Z. Exper., 10 :

79-97-
Delage, Y., et Herouard, E.

'13 Traitc de Zoologie Concrete. Tome 3, Les Echinodermes.
Durham, H. E.

'88 The Emigration of Amoeboid Corpuscles in the Starfish. Proc. Roy. Soc.

London, 43 : 327-330.
'91 On the Wandering Cells in Echinoderms. etc., More Especially with Regard

to Excretion. Q. Jour. Mic. Sci., 33 : 81-121.
Enriques, P.

'02 Digestione, circulazione, e assorbimento nelle Oloturie. Arch. Zool. Un.

Itali, I : 1-58.
Frenzel, J.

'92 Beitrage zur verglcichende Physiologic und Histologie der Verdauung. i
Mittheilung der Darmcanal der Echinodermen. Arch. Anat. Physiol.
Abth., 81-114.

Geddes, P.

'80 Observations sur le fluide perivisceral des oursins. Arch, de Zool. exp. et

gener., 8 : 483-496.
Goodrich, E. S.

'19 Pseudopodia of the Leucocytes of Invertebrates. Quart. Jour. Mic. Sci.,

64 : 19-27-
Griffiths, A. B.

'92 Sur 1'echinochrome: un pigment respiratoirc. C. R. Acad. Sci. Paris, 115 :

419-420.
Henri, V.

'06 Etude du liquide perivisceral des Oursins. Ek-nu-nts figures. 1'hmomene
de la coagulation et son role biologique. C. R. Soc. Biol. Paris, 60:
880-882.
Herouard, E.

'95 De 1'excretion chez les Holothuries. Bull. Soc. /"<>1. Franco, 20 : 161-165.
Howell, W. H.

'85 The Presence of Hemoglobin in the Erhmodrmi-;. Johns Hopkins I'niv.
Circ.. 5 : .v

PERIVISCERAL FLUID OF ECHIXODERMS.

'86 Xote on the Presence of Hemoglobin in the Echinoderm id. Biol. Lab.

J. II..;. kin-, 3 : 289-291.
Kindred, J. E.

'21 Phagocytosis and Clotting in the Perivisceral Fluid of Arbacia. Biol. Bull.,

41 : 144-152
Kollmann, M.

'08 Ri- Ann. Sc. Xat. / . ~ ie 8

Knoll,

'93 — n Thieren. .02.

List, T.

'97 '"• :ikrystall<>i' lfii in di

\Vai Anz.. 14 : 1X5-191.

McMunn, C. A.

"85 On tli 1 of Some I : . Jour.

M .'- M^.

Michel, A.

'88 Sui l.i ^vmphati C. 1<

A. .1-1. >6.

Saint-Hilaire, C.

'97 ' rmwande d< up.

;8.
Schafcr, E. A.

"83 ( >n tin i ' i . a Urchin. 1 34 :

Schultz, E.

"9; l>ei den Ho' :iii.ilbl.

15
Semper, C.

'68 I>i. II pel den Philippien, II.. i.

Theel, H

'94 N ;'>n in ilif I

i

'96 Kfiii.ii ka on tl iKi'lt-riii i • *ch.

l.i!1
'19 ' "" A- :• i iin-ii.

1. A . Aii hi\ tgivet

Bd, 14.

*JI < >ii Ai: pl ( itlu-r t'n-1. in 1 the F ' uy

"i I i in-;. 11 1- I |..!..tliuti :- An ln\ Mil. 1.5. N : 25 :

;".
Van der Heyde, H. C.

'22 1 1. ni.'Ji.lun in . .: HIDL. Bfi.L., 42 : v5

17

CELL BEHAVIOR IX TISSUE CULTURES.

HUBERT B. GOODRICH,
WESLEYAN UNIVERSITY AND THE MAKINI: BIOLOGICAL LABORATORY.

This paper deals with certain free wandering cells observed in
tissue cultures from the minnow, Fundulns lieteroditus and
Fundulus majalis. The types of cells which became free and
isolated from the spreading tissue growths were, chromatophores,
amoeboid mesenchyme cells and, most abundant of all, certain
cells having curious fan shaped projections. These cells proved
to be identical with those studied in their association in tissues
in cultures by Dr. Dederer ('21). She identified these as mesen-
chyme cells which in cultures were the means of attachment of
the outgrowing sheets of ectodermal cells to the surface of the
coverslip. In the work here presented isolated cells were sought
as the best objects for the study of cell behavior. The mode of
locomotion and the tactile reactions were more especially studied
and for the latter work the Barber microdissection apparatus was
utilized. The tissues were cultivated in the sea water medium
(M. R. Lewis, '16) using, however, in many cases more dilute
solutions. Observations were usually made within two days
after planting. The work was done at the Marine Biological
Laboratory at Woods Hole, Massachusetts during the summers
of 1921 and 1922. I wish to acknowledge my indebtedness to
Mrs. D. B. Young for the drawings from the stained preparations
and to Mr. S. C. Williams for aid in certain of the observations
on the rate of motion of cells.

The Fan Cells — Among cells of this type an abundant form
was that for which I came to use the descriptive term "Canoe
cells." When first observed these seemed to be elongate spindle-
shaped cells such as indicated by many outlines in Fig. 4. I
supposed that there were delicate psuedopedia at either end but
upon plotting the direction of the motion of these cells I was
surprised to find that they were moving steadily at right angles
to the long axis. More careful observations upon living and upon

fixed and stained preparations showed the presence of a delir.it r

252

CELL BEHAVIOR IN TISSUE CULTURES.

253

film along what proved to be the anterior side of the cell. This
film or fan rounded about the ends of the elongate cell, thus
giving the canoe-like form (Fig. i). By use of the microdissection

1

i. Typical " Canoe " cell, Drawn from stained preparation.

I x. _• and 3. !»....• o-ll (2) and double fanned or bipolar

ci-11 (3) showing microdi- l>i"ly of cell and beneath the cover

i. The fans arc firmly at t nd the needle can not be pushe '-n tin-m
the

apparatus the relation of the cell to the cover glass was determined
more accurately. It was possible to slide a needle between the
more visible spindle shaped part of the cell and the cover glass.
The, fan, however, was firmly attached (Fig. 2). The fan wa^
clearly of ectoplasm in the gel state while there was more fluid
protoplasm within the body of the cell.

I believe these fans to be the motor organs of the cell. There

254 HUBERT B. GOODRICH.

seemed always to exist a definite relation between the position of
the fan and the direction of motion. The fans were the only
portion of the cells in contact with the solid support and there
seemed to exist no mechanism for locomotion in a fluid medium
without support. Other types of fan cells, as described below,
illustrate these points even more clearly than the "Canoe" cells.
The rate of motion of the "Canoe" cells was studied by plotting
their course with a camera lucida. Fig. 4 shows the history of such
a cell. In most cases the outline of the fan could not be observed
with the camera in position and an outline of only the body was
drawn. However, in positions I, 12, 21, 28, 30, 32 and 34 the
probable form of the fan is indicated by dotted lines. These
outlines were based on observations with the prism of the camera
removed. At position 8 the cell became attached by a psuedo-
podium-like projection on the right which may have terminated
in a fan. A similar process occurred at positions 24 to 35 during
which period a small fan could clearly be observed at the right.
At position 27 the cell under observation collided with another
cell. The two cells became attached and the newcomer formed
an irregular projection at the upper right of the original cell
(positions 27 to 36).

The average rate of motion of freely moving cells excluding
such cells as proved to be slowing down prior to the death of the
cell, was 6.3 microns per minute. The cell shown in Fig. 4 moved
at a rate of 5.3 microns from positions I to 25. An attempt was
made to study the effects of changes of temperature upon the
movement of these cells. For this purpose the cultures were
studied under the microscope in a warmed box at temperatures
varying from 21° centigrade to 42°. Above 40° the cells withdrew
their fans and became rounded. Observations were made at
constant temperatures and also during an increase of tempera-
ture. It soon became apparent that variations in the conditions
of individual cells would preclude the possibility of constructing
a temperature curve for the rate of locomotion. Cells becoming
free from the main growth of tissue moved for a variable period
and then became rounded and this condition probably pivceeded
the death of the cell. A slowing of the rate of motion was appare-
ent before the contraction took place. Also the varying torm of
the cell and the probable occasional attachment !>v small sub-

32 — -

36 12.14

12>U 28

11.15

11 07

< 2 VJ '

FIG. 4. Tin- hi>toiy i.i" a ' Vunoe" cell for one hour and seven ininutrs at room
temperature. The solid line- outline the cell body in so far as it was visible \vln-n
projected by the camera luciila. The dotted lines show boundary of fans based on
obsei with tin- pii-ni of the camera raised.

256 HUBERT B. GOODRICH.

sidiary fans, as in Fig. 4, which are invisible when making camera
drawings, doubtless influence the rate of motion. Nevertheless
the results seem to indicate that an increase of temperature causes
an increase in the velocity of the movement. The average of all
(15) observations at room temperature showed a rate of locomo-
tion of 6.3 microns per minute. The average of all (10) observa-
tions at higher temperatures (in most cases varying) was 7.3
microns per minute. The records of greatest speed were 11.5
microns and 10 microns per minute and were attained by cells
at the higher temperatures. In the latter case the cell was fol-
lowed for 39 minutes. The details of these observations are
recorded in the tables in the appendix to this paper.

A second form assumed by these cells was that exhibiting
two fans (Figs. 5, 6, 7). These fans were usually at opposite
poles and under their influence the cell became greatly attenuated.
In such cases a micro-dissection needle could be passed between
the coverslip and the body of the cell (Fig. 3), but the fans were
found to be firmly attached. The cell was thus freely suspended
like a hammock between two supports — the fans forming means
of attachment and also of extension. Two typical double fanned
or bipolar cells are shown in Fig. 5 and 7 which were drawn from
stained preparations. In Fig. 6 are shown two cells attached
with one fan pulling in a direction not directly opposed to the
other.

Another extraordinary mode of motion was observed in which
the contractility of the cell functioned as well as the gliding motion
of the fan. The history of such a cell is shown in Fig. 9. The
account begins at 2.44 P.M. with the cell at position I and with
a fan at either end. Suddenly a release of the fan occurs and the
cell contracts and is thrown into position 2. It again elongates,
fans are found at either end — positions 3, 4, 5 and at 3.01 a
second contraction occurs, throwing the cell into position 6. The
process is repeated four times — positions 6-9, 9-12, 12-13, and
13-18. In each case the cell is elongated by the pulling of the op-
posed fans. This unusual mode of motion was observed in a
culture from a 17-day embryo. This marked contractility of a
mesenchyme cell is almost suggestive of the behavior of muscle
cells as described by M. R. Lewis ('20) except that these cells

8

FiCS. 5. 6, 7, 8. Typii-al fan cells drawn from stained pr<-parati<>:
and 7 are bipolar or double fanned cells. The limits of the fans may have extended,
further than indicated. Fig. 6. Fan cells — possibly shortly
I i-^. 8. Two cells apparently fused forming syncytium. \' >ut 1800

times.

258

HUBERT B. GOODRICH.

contract completely to a spherical form and then gradually
expand.

i V

2.44

100 MICRONS

FIG. 9. The history of a bipolar fan cell for one hour and eleven minutes at
room temperature of 23°. Movement is by method of alternate expansion and
contraction. For details see text.

The cells having fans were found only in cultures showing
epithelial growths such as described by M. R. Lewis ('16) and
Dederer ('21). I have planted cultures from embryos of various
stages but have been unable to obtain this type of growth from
those of less than six days of age when developing at the tempera-
ture of the running sea water in the laboratory (19° to 22°). The
most favorable period for planting and obtaining such growths is
shortly before hatching (about 17 days, !9°-22°). Sections of
embryos of these later stages reveal a conspicious layer of cells
beneath the surface epithelium which are almost completely

CELL BEHAVIOR IX TISSUE CULTURES. 259

wanting in the earlier stages. Dr. Dederer's observations have
shown the close relation between epithelial cells and the fan cells
and it is then not improbable that this layer contains the fan
cells. Moreover she has shown these fan cells to be necessary
for the spreading of the epithelial growth l>y attaching it to the
rover glass. Therefore the absence of the epithelium in cultures
from the younger embryo-; may be due to the absence of this
sub-epithelial layer with its included fan cell-.

In a few cultures other impumented cells of presumably mes-
enchymul origin and of exceedingly irregular form were noted.
Tlu-ir mode of motion -reined typically amoeboid.

Tai'tili- I ns. — The tactile reactions of various types nf

"ell- were, studied. For this purpose the Cells were touched with
a delicate 1.1 die moved by a Barber mien t-di--cct i< >n ap-

paratUS. It" the body of a double fanned Cell -uch a- that sho\vn
in 1 i:_ .} i~ sharply stimulated the fans are rule ised iiid the whole
cell Contracts, becoming spheri< J. A fail may be pried loo-e. and
a similar reaction en-iies. The contraction seem- in part due to
an ela-tic ten-ion of the cell. I have not found it possible to
stimulate a < ell b\ toiichr i partially mutil.itiiu -i -mall

portion of the fan I .pla-m does not -eem to conduct a

stimuli!-. In -.-me ie stimuli!- near the boundary of

a fan and the cell -talk of a greatly elongated cell will call-c . oiu-
plete contra« tii'ii. The material of the fan -eem- (<• flow to-
gether, making a ball of protupla-m which i- carried along with
the -t. ilk to the center of the cell.

I ha\e tried a few experiment- to determine the chemutactic
response "f -u<-h ti — ue cell-. A .loud of methylcne blue in-
jected by a mi T"-pipette will cau-e the contraction <,f the-e
i-olated cells. I ha\e not -ucceeded in obtaining an\' more def-
inite respon-e such a- a change in the direction of motion.

Previous writers, Bancroft ('12), Stockard ('15' and \e\vmann
('18) ha\e dc-cribcd two t\ pes of chroinatophores in the Fnndithts
embryo. These are black chromatophores or melauophie- and
the brown (or red) chromatophores. Bancroft ('u ha- de-cribcd
a third type which also appeared in these tissue cultun Th*
were \cllo\v and -mailer than the other types and -iiowed few or
no pseudopodia. As a group, the chromatophores are but -lightly
responsive to tactile stimulation. If a needle is pushed a

26O HUBERT B. GOODRICH.

a pseudopod with sufficient force to slightly indent the ectoplasm,
the pseudopod may slowly withdraw. A brownian movement of
pigment granules is frequently initiated. Of these cells the yellow
chromatophores are most responsive, the red are less responsive
and the melanophores are almost completely inert to tactile
stimulation.

All cells studied seem to show less variety of adaptive response
than the amoeba. Thus the only reaction that I have observed
is contraction either of the whole or a portion of a cell.

I have previously referred (Goodrich, '22) to the motion of the
fan cells as non-amoeboid. The characteristic streaming of
protoplasm which we associate with the amoeba is certainly not
present. The phenomenon seems more akin to the movement of
diatoms. It is, however, not impossible that this gliding motion
may be a factor in the locomotion of many unicellular organisms.
Schaefer ('20) in his discussion of amoeboid movement has called
attention to the importance of a surface film of streaming proto-
plasm external to the ectoplasm and wholly distinct from the
familiar streaming of the endoplasm. This film can be observed
only indirectly as it carries particles that become entangled in it.
Schaefer states (page 106) that "the surface film in amcebas is
powerful enough to enable them to move by it." In this case it
causes a backward motion. It is also probable (see Schaefer,
'20, for discussion) that such a surface film is important in the
movement of diatoms, Oscillitoria and even of Gregarines. No
adequate explanation has been offered for the motion of this
surface film. I have not been able to detect the presence of such
a moving film in the fans of the cells studied in this paper although
cells have been observed in media containing a suspension of
carbon granules. Yet the delicacy of the fan is such as to make
the test seem inadequate and it is not impossible that such a
mechanism may exist. If so we may have some clue to the motion
of many cells in development and in regereration. Moreover this
mode of motion seems allied to the power of adhesion of cells. The
cells here described are attached to the cover glass by means
of the fans. Dr. W. H. Lewis ('22) has raised the question as
to why tissue cells adhere in an organism. The mechanics of
the gliding motion seems to involve this power of adhesion and
thus the two problems may be related.

CELL BEHAVIOR IX TISSUE CULTURES. 26l

SUMMARY.

1. Certain isolated cells from tissue cultures of Fnndulns
embryos have been described.

2. These cells posess fan-shaped films which are adherent to
the cover glass.

3. These films are the motor organs of the cells by means or
which they glide on the under surface of the cover glass.

4. The tactile reactions of these cells and of the chromato-
phores are described.

5. The relation of thi> type of cell movement to amoeboid
movement i> di-< n— rd.

BIBLIOGRAPHY.

Bancroft, F. W.

'12 H'-i'-'lifv >:JuIui I! [OUT. EX] 7.^»\.. \'«\.

i -

Dederer, Pauline H.
'21 I: iltures of Fund:, .':<J with Sp.

k . BIOL. BCLL. \"..l. 41. No. 4-

Goodrich, H. B.

'22 c-ll |. ..• Cultun-s. Abstract! ta. Sod

ni / 13, No. i. p. 100.

Lewis, M. R.

'16 si-.i V. ' ie Cultures. Ai. i". N

'20 M .itures. Carnegie Institution of Waahing-

.ntrilur . No. 35. 1'uMu ati-n No. 272.

Lewis, W. H.

'22 'u.ility • Anat. Rcc., Vol. 33. N

Newmann, H. H.

'18 Ilytni<U ! ' '• . '.. \ '"1. -"'. 1918,

Schaeffer, Asa A.

'20 Ai; ' ' 'ii I ni\ •

Stockard, Charles R.

"15 A Sui'h ••; \\ '.unl.'iiii.i; M. •-. M. i.;. :; • :i<- Living Volk Sac an<l t ln-ir

DevclopMii-ntal 1'io'lui't-; ( himna- . ilai El ahclium and

Blo<"l ( ells. Am. Jour. Anat.. V..: . 3.

262

HUBERT B. GOODRICH.

APPENDIX.

TABLE I.

RECORD OF OBSERVATIONS ON RATE OF MOVEMENT OF CELLS AT ROOM

TEMPERATURE.

Reference

Number.

Average Rate
in Microns
per Minute.

Temperature
in Degrees
Centigrade.

Duration of
Observations
in Minutes

i

3-4

24-S

4i

2

3-3

24-5

42

3

7-6

24

16

4

4-

24.

33

5

4.1

24.

19

6

5-9

24.

ii

7

8.1

22.5

25

•

8.9

22.

15

9

8.2

22.

ii-S

10

7-

22.

8-5

ii

6.1

22.

9-

12

8.

21-5

20

13

7-7

21.5

13

14

8.

21.5

7-5

15

5-3

22.

48.

TABLE II.

RECORD OF OBSERVATIONS OF RATE OF MOTION OF CELLS AT HIGHER

TEMPERATURES.
The temperature was in most cases gradually increased as indicated.

Reference
Number.

Average Rate
in Microns
per Minute.

Temperature
in Degrees
Centigrade.

Duration of
Observations
in Minutes.

16

8-3

29-5-37

25

i?

n-5

27 -33-8

20

18

8-3

33.8-36.5

25

19

3-4

27-35

44

. .

9.1

35

7

21

5-

27-30

12

22

5-8

28-31

3

23

6.4

27-31

13

24

5-i

27-29

10

25

10.

33-35

39

Vol.XLVI June, 1924 No. 6

BIOLOGICAL BULLETIN

STUDIES ON III! -KCONDARY SEXUAL
« HARA< TERS OF CRAYFISHES.

i. MMI: SECONDARY SEXUAL CHARACTERS i\
I'I.MMI- <n ('<;»/ bams propinqitus.

L. TURNER.

ivvioRY. BELOIT COLLEGE.

: In- • ! the writer t"i»un<l five -pecimciis

'inn/Hint* f>r<>[>in<inn\ which had both m.ilc ,m<l Female second-
ary sexual characters. Two of tlie specinu n- were destro

in di— ectini; tin- iiiti-rii.il or-. ins, in examining tin- .innulus for
-pi iin ,m<l in -ep.iratini; the appendages preparatory to drawing
thrin. A r. \ir\\ i.|' the lit rr.it ure revealed that the r<m<Iiti(in is
<-xtrcinel\ rare. Many tin m-. i nds of specimen- ha\r < »mr uml< r
thr do-r -i i ntin\ <.t Faxon, Eia^en, H.iy, Ortni.inn and others
.ind only thirti-c-n cases in the i:i-nus Canibarns h.i\'e Krrii rer. .nU-<l
in which ni.ile .nul I'em.ile -rc< >mlary >r\u.il character> ha\r Keen
fi.unil ti> r\i-t in the -.une indi\ idu.iU. The tiiulin.^ of the live
-pi cinieii- -»-ein- all the nmn- rein. irk. il >le U-can-e all were t.iken
within a -hurt distance of each other in Turtle ( 'reek near lieloit ,
\\'i-con-in on a -iii;Jc field tri|>. Ahout two hundred -pecinieii-
were t.iken on the trip so that the >eeniin.u henna|)hroditic ones
con>titute<l about J.5 per cent, of the en tire catch.

IHli^ent M-arch w,i- made in the same locality during the late
>prin^ with the result lh.it three more similar -pecimen- lieloii^in^

to the -.ime -pet ir- were found.1 The ei^ht specimens are re-
ferred to in the descriptions as numbers i to 8 inclusive.

The search was continued in l<»J3 resulting in the capture of
four females in the >ame locality in Turtle deck. All showed

imens described li.iv \»-<-i\ :\ the Carnegie

Museum at I'ittshurxli.

1 8 263

264 C. L. TURNER.

definite secondary sexual characters of the male. All four were
carrying eggs or embryos attached to the abdomen, making it
certain that they were fully functional females in spite of the
presence of male secondary sexual characters. These four will
be referred to in the descriptions as numbers 9, 10, n and 12.
The brood of one of the specimens was successfully reared. Eight
months after hatching twelve members of the brood were still
thriving and it is hoped that the effort to maintain and breed
them under laboratory conditions will be successful.

Later in the summer of 1923 one hundred females were taken
from South Kinnikinick Creek a tributary of the Rock River fif-
teen miles south of Beloit. Among these were found seven having
the partial development of male secondary sexual characters.
These seven will be referred to as numbers 13, 14, 15, 16, 17, 18
and 19.

Other streams in the vicinity of Beloit have been searched and
a considerable number of abnormalities in the secondary sexual
characters have been found. In no female individual, however,
has there been found any male secondary sexual character.

The fact that this simultaneous occurrence of male and female
secondary sexual characters in the same individual has been found
only thirteen times in the entire genus Cambarus makes the
discovery of nineteen specimens in a single species within a
limited area well worth investigating.

This same condition has been found to exist in southern Wis-
consin to a still greater degree in Cambarus virilis. Eighty-eight
per cent, of all the females taken in Delavan Lake show additional ,
male secondary sexual characters.

It seems likely that the conditions causing the high rate of
peculiarities is a local one and it is proposed to go into the mat 1 1 i
thoroughly and to publish results whenever a body of newdata has
been collected or whenever any conclusions have been reached.

REVIEW OF LITERATURE.

The first described instance of a truly hermaphroditic decapod
crustacean is that of a lobster (Honmris viifyiris) which had
normal external genitalia and internal organs of the m.ile <>n the
left side and those of a female on the right side. The specimen

SECONDARY SEXUAL CHARACTERS OF CRAYFISHES. 265

was described by F. Xicholls in 1730. Another specimen has
been described in the Canadian Naturalist, Vol. XXXIII, Xo. 2.

Rosseau and Desmarest described several cases in Astacns
Jha'iatilis of branched oviducts opening on the third and upon the
fourth walking legs.

Yon Martens described three specimens of Cheraps preisii,
an Australian crayli-h, in which there were openings at the 1
of the third walkin. luit no oviduct- and no ovaries. The

specimen- were normal male- except for this peculiarity.

Yon Ihrii .. I dnnberg, Faxon and Hay working upon variou-

ies of Parastacus, a genii- found in South America, found that

then- were i ' n-tant conditions of the genital inland-, il

ami external ^enitalia. luoc.f these conditions pointed toward

|).irtial hermaphroditism. In -ome species there were

Mmje uenit.il inland-, fit her -permaries or ovarie-. and tv,

. enital iliM i- am! genital openings. Only one of the sets
tube- u.:- fum tioii.il. however. In other spi -ingle genital

inland am! two sets of genital tubes existed but there were no
ital opening- for the non-functional tube-. In -till another
>pe« ie- I lien- wafl n- - < \ i<;' ther of extra genital lube- «>r «.f

extra genital opening-. The normal condition in the tir-t two
Mibdi\ i-inii- o| tin- veiiu- was, therefore, one of partial herma-

phroditism.

1 axon [885 foiiml foul - in the genii- Cani'unii which

\ ve some evidence of hermaphroditism.
One specimen of Can <{>inqu:. >w, 60 mm. loi

had the clau 5 of 3 female, the al»Iominal a|)pemla.;e- of ,i female
ami a well-formed amuilu- but no li-male genital opening- at the
ba-i- of the thiid walking legs. In-tea«l, it had the external

ital opening- of a male at the base of the fifth walking 1.
Hi— i-ction deiunn-trated the pre-ence of a well-de\-elo|n-d o\,u\
MI the -pecimeil \\ 'a- e\ idell 1 1\" a lelllale.

The -ecoml i\i-e wa- al-o one of Cambarns pro pi n<i nits sanborni,
38 mm. loiii^. The specimen \\a- female in every n •-] • ept

that the tir-t abdominal appendages were like tho-eof the m

The third case « -pecimen of Ciimbunts • ^4 mm.

Ion-, in which the external genitalia were tho-e of a female with
the exception of the tir-t abdominal appendage- which wi-re

266 C. L. TURNER.

almost as large as those of a male but were lacking the charac-
teristic, recurved hooks.

The fourth specimen, a female of Cambarus pro pin guns, 72 mm.
long, was normal wi t h the exception of the first abdominal append-
ages. These were modified so as to be almost identical with but
slightly smaller than those of the male.

Hay (1905) in examining the collection of crayfishes in the
U. S. National Museum which had been made by the U. S. Fish
Commission and by himself found four additional specimens
which showed evidence of hermaphroditism.

In the first specimen, Cambarns spinosns, the organs internally
and externally were those of a female with the exception of the
annulus ventralis which was rather small. In addition there
were the following male secondary sexual characters: hooks on
the base of the third walking legs, abdominal appendages
modified as in the male but slightly undersized, genital papilla at
the base of the fifth walking legs, but no genital pores.

The second specimen, also Cambarns spinosus, was identical
with the first with the exception of the first abdominal appendages
which wrere shorter.

The third case, a specimen of Cambarns propinquus, 53 mm.
long, displayed a preponderance of male secondary sexual charac-
ters on the left side and of female characters on the right side.
The right side had the usual genital pore of the female at the base
of the third walking leg, a small abnormal appendage on the first
abdominal segment and the usual second abdominal appendage
of the female. The annulus ventralis was also well developed.
On the left side there was no female genital opening at the base
of the third walking leg but on this third leg there was the copu-
latory hook normally found in the male. The first and second
abdominal appendages of the left side were those of a normal male
and at the base of the fifth walking leg there was the genital pore
of the normal male. Nothing of the internal structure could be
made out.

The fourth case appeared in a specimen of Cambarns affinis,
1 06 mm. long. Externally it had all the characters of a male
except that it had no opening for the sperm duct at the base of the
left fifth walking leg. It did have, li"\vrvrr, the genital openings

SECONDARY SEXUAL CHARACTERS OF CRAYFISHES. 267

of the female at the base of the third walking legs. A dissection
revealed that there were fully developed ovaries and oviducts on
both sides and a small testis with a -perm duct on the right -ide.

Hay concludes that in the genus Cambarns hermaphroditic in-
dividuals are females "which owing to some ambiguity of t In-
formative cells in the embryo have de\ i t«> a greater or le—
degree the characters of the opposite sex. The condition is a \ vry
rare one and i- u-ually -hown in the external organs only". . . .
" Hermaphroditism i- as uncommon in Astacus as in Cum/urns.
Among the I'ara-t.n id.e the condition of apparent hermaphrodit-
i-m seems to U- e-tabli-hed in the genu- /' It may
also 1'e found in ait is rare or altogether wanting in
otlier genera. "

Ortmaii [905 ';>- nhi-ii ti\e additional cases in the genus
• ibnrns, four colle.ie-1 by himself and one 1>\ I-]. M. \Yilliam-on.

1 li- !ir-t case was< >nr • >f < rns obscnru^ which had as female

-.ual charai tcr- the female t>i"- • <\ • law. a detinite
hut -mall amiulii- an<l no ho. .k- on the third walking legs. I •
spe< imen had. ho\\.-\er. the genital opening- of the male al the
6 <>f the lifth walki: tnd the first and -f.-Mul alxloniinal

ap|)eii' < leal 1\ < if the male type.

The -ennui -p«-< inuMi \\ • one of Cum/Mir. v. liich

had not <ml\- t he aiiiuilu> and the genital openin in- ft-nia It-

hut al-o copulatoiA hook- on i he third leg (a male « haracterisl
The alidominal appendage- Wi I e .1! ui-'i'inal Inn full--i/ed ami
male-like. 1'hc -pecimeii \\a- inter|)reted as deing a female
with the lull Inn not -pi-ciiic <le\ elopni.-m of male . n-gan-.

In the third case the -pecinien ;rt<uii< \\a- a func-

tional female. 71 nun. long with .ttached to the alxlomen.

It -1 lowed the full equipment of SCCOndarj -e\ual character- of t he
fcniali- and in addition a pair of hook- at the base of th.- third
walking 1« j

The fourth individual, al-o 'men of Ciunhnrns Imrtoni

(48 mm. long', had female i law-, a well-developrd aiintllus and
the second abdominal appendage- of the female. The tl-lial
hooks at the l>a-e of the third walking keg- were lacking. 'I •
male secondary sexual character^ were shown in the genital open-
ings at the l>a-e of the fifth walking legs and in the full develop-
ment of the male tir-t abdominal appemlao -

268 C. L. TURNER.

The specimen collected by Williamson was one of Cambarus
rnsticus. The first abdominal appendages were not fully devel-
oped but were clearly of the male type while all the other second-
ary sexual characters were of the female type.

DESCRIPTION OF NEW CASES.

Specimens I, 2, j, 4 and 5. (Fig. 5.) These five specimens are
identical in their peculiarities and it will therefore suffice to
describe one of them and to give the lengths of the other four.
The length of the specimen described is 48 mm. The lengths of
the others are 45, 47, 49 and 52 mm. respectively. In the speci-
men 48 mm. long the secondary sexual characters are typically fe-
male but there are the following additional characters : copulatory
hooks are present at the base of the third walking legs; the first
abdominal appendages are modified to resemble somewhat those
of the male of the first form, being shorter, however, (8 mm. in
length) with the tips undivided and curved toward the median
line (Fig. 10); the second abdominal appendages are those of a
normal male (Fig. 10). A dissection of the internal organs has
been made and ovaries and oviducts have been found. The eggs
are full sized and would normally have been laid within two
months. The ovaries are heavily parasitized by the embryos of
a flatworm l to such an extent that many of the eggs are entirely
replaced by the encysted embryos. No trace of a testis or of a
vas deferens is present.

Specimen number 6 is 42 mm. long. It differs from numbers
i, 2, 3, 4, and 5 only in the degree of development of the first
abdominal appendages. The right member of the pair is in all
respects like the homologous appendage of the normal male of the
second form. The left member of the pair is slightly shorter and
the tip is bent toward the median line.

Specimen number 7 is 54 mm. in length and is typically female
in all its secondary sexual characters except in the case of the two
abdominal appendages. The first abdominal appendages are
modified like those of cases I, 2, 3, 4 and 5 and are of the same
size. The second abdominal, appendages differ only slightly
from those of the female, being a little stronger and the iniu-r

1 The writer is indebted to Professor A. S. Pear -< «i iln I nivcrsity nt" \Vi-.-c m-on
for identifying the parasites as Microphallus opacits Ward.

SECONDARY SEXUAL CHARACTERS OF CRAYFISHES. 269

ramus a little thicker. There is no trace of the triangular shoul-
(U-r which occurs regularly on this appendage in the normal male.
This specimen has been dissected and as in the previous cases
there have been found well developed ova. This case is unque-t-
ionably one of a female with the partial development of male
secondary -exual character-. The mak- char.icu-rs are not so
-tnmgly developed a- in the previous cases, tlu- ho. .k- . >n the third
w. ilking kg- being ab-ent and the second abdominal appendages
being modified hut link- from the usual female type.

Specimen mimher .V i- 47 mm. long. It is without doubt a
Ifiii. ik, h.i\ ing ,i iionn.il .mmilu- ventralis and oviducal opening-
at t IK- h. i M-< if the third walking legs. The fir- 1 abdominal appcnd-
|n-tuliar, however, being unlike on tlu- two -ide-.
Tlu- right OIK- i- a normal <limiiuitive female appendage. The
k-ft on. Fig. II pp have grown as a bifurcated female

appendage hilt there I . .\VI1 OUt of it .M I hi- junction of the

h.i-al and di-tal parts an a| ipendage somewhat -imilar to ami a
little -mailer than the normal tirst abdominal appendage • >l tin-
•nd form male. The [ndximal half of the appendage i- ijuite
>lender. The MM. .n<l abdominal appendage-, hoih in >i/e and
• ral loini i. M nihle the normal female a|i|u-iida^i - bin a dis-
tinct triangular -houlder like that found in the male i- piv-ent
upon the endo|)od: 14 .

Xf>fi inicH unmln-r (j \\.,~ hearing well-de\'i-lo|>ed eggs u|ioii tlie
abdomen \\lien taken SO there was no question a- to the -.-\. It
is (>~ mm. in length. The nviducal Opening^ and the aimulus
\enirali- are tho-i- of a normal female but the following male

M rue in re- are also prcM-ni : copulatory hooks on the third walking

; !u>t .ihdominal ,ippen<l.igi-- much like tin -e of the imrmal
tir-t form male: d abdoiiiin.il append. I*;. •- like tho-^' of the

mak- in form but a little -mailer and with the triangular >hoiilder
on the eildopodhe Mot SO \\ell de\i-loped . .

Specimen numher i<>, 55 mm. long, i- umi-ual in that it po-
an extra o\ iducal |>ore at the ha-e of the left fourth walking leg in
addition to a rather complete complement of both male and
female -ccondary >e\lial character- FigS. J, 9 and 141. The
annulu- \etitrali- and o\ iducal pore- at the base of the third
walking legs are norm.il female character-. An extra oviducal
opening is present on the left fourth walking leg but it is

2~O C. L. TURNER.

small and has no oviduct attached to it. The openings of the
sperm ducts at the base of the fifth walking legs are in the same
position as those of the normal male but are smaller and are not
so prominent. The first abdominal appendages are shaped much
like those of the normal, second form male but they are a little
smaller and are more slender in the basal third. The second
abdominal appendages are like those of the normal female with
the addition of a marked triangular shoulder upon the endopodite,
indicating a tendency on the part of the appendage to develope
specific male characteristics.

This specimen was dissected and it was found that the internal
structures were those of a female. The extra oviducal opening
and the openings at the base of the fifth walking legs had no
internal tubes connected with them. There was no trace of a
spermary.

Specimen number n is 62 mm. long and is identical in its
external peculiarities with specimen number 9. Like specimen
number 9 also, it was bearing young embryos upon the swim-
merets when taken and was of course a functional female. With
the eggs so recently discharged it would be expected that the
ovary would be depleted and small. A large genital gland has
been found, however, which is irregular in shape and seems to be
made up of unlike parts. It is possible that the structure may
prove to be an ovo-testis.

Specimen number 12, 53 mm. long, when taken was bearing
embryos which were in an advanced stage. These embryos were
reared in the laboratory and forty-nine of the sixty reached a
stage in which the secondary sexual characters had become distin-
guishable. All the embryos were female and all were normal.
As in specimen number 1 1 a large irregular genital gland has been
found but there are no ducts except the normal oviducts. Exter-
nally this specimen is with the exception of its size almost identical
with specimen 1 1.

Specimens number ij to ig inclusive (Fig. 6) are alike in their
peculiarities, all being normal females as evinced by their internal
and external organs but bearing in addition one male secondary
sexual character, namely a hook on the left third walking leg.
The lengths of the specimens number 13 to 19 are respectively
63 mm., 47 mm., 49 mm., 47 mm., 42 mm., 45 mm., and 5 omm.

SECONDARY SKXTAI. ( HARACTERS OF CRAYFISHES. 2~ I

DISCUSSION.

The occurrence of both male and female secondary sexual
characters in the same individual is so rare and sporadic that it
has seemed futile to speculate as to its cause. However, when
nineteen instances are found in a limited area within two year:- it
would seem that a close study of conditions might yield results.
Two possible explanation- -u^gest them-eh

I. The gonad- may ha\v been damaged by parasites so that
the po~-ible control ()f tin.- development of the -econdary sexual
character- would be lackiiu. 1 'arasites 2 were found, indeed, in
t number- in both male- and in female- and there can be no
doubt thai they did exten-i\e damage to the ti--ue- of tin- cr
fishes. ' Ovaries, ainu-mary canal-, ^reen -land- and

li\er- wen- often heaxilv infe-ted and partially dc-troycd. The
seeming effe< i of p.u.i-iti-m of the gonads upon the development
of the -e< ondary -e\ual characters has been worked out in i^reat
<let. iil by Smith • Pii" . This author fount • number- . .f the

spider i 'tab (Inachm so p.n-a-itized by Cirrepede Sacculina that
the ovaries and -pel in arie- . .f Inachlts were frequently de-tro\e.l.
The timlin^- wiih regard to the effect upon t; mlary -exual

characters and hi- . .mclu-ioiis may be found in the following
-tatemcnt : " \Ylicie.i- the malrs under the influence . .f the para-
site, wen ;>al'le . all the femai' ndar\ -exual
characters and under certain conditions mi^ht e\eii «le\ elope
OVa in their testes, tin- infected female> on the other hand,
although the ovaries in some iinplei«-l\- ili-ap|ie,[re.l, ne\«-r

approached t" the male primal^ or secondary -exual characters
in the slightest degree." ". . . Tliepi. nlinn can-ed

\-omii; female- uiuler i.; mm. in carapace length to assume ;

matureK llie adult t \ pe • -f abiloineti and abdominal appemla.
". . . it \\ould appear that the male wa- a potential herma-
phrodite and the female purely female."

If the irreguluritk-- de-cribed j,, ('nmhtirits prof>in<]nns are due
to para-iti-m it mi-h.t be ex|u-cie<l that the number of irre-ulari-
tie- would IK- -omewhat j >ropor t ional to the number of heavily
para-iti/ed indi\ idual-. Thi-, ho\\e\er. does not pro\ e to b«
the case for para-iti-m i- unite common and irregular secondary
sexual character- are rare Again, if para-iti-m can-es these

2J2 C. L. TURNER.

irregularities some principle other than that found by Smith for
Inachus must be sought in Cambarns propivqnns to explain the
relations between the gonads and the secondary sexual characters
for in Inachus " the male is potentially hermaphroditic and the
female purely female" while in Cambarns in the instances cited the
male shows no tendency to develope secondary sexual characters of
the female although the female in all the cases described has devel-
oped more or less fully the secondary sexual characters of the male.

II. In Parastacus the presence of both male and female
secondary sexual sharacters in the same individual is a fixed
condition which is transmitted from one generation to another
without change. The question arises as to whether or not the
condition may not be transmitted in Cambarus as in Parastacus
although much less firmly fixed in Cambarns than in Parastacus.
The finding of a considerable number of specimens within a
very limited region would tend to strengthen the view that the
condition was somewhat constant in Cambarns and was being
transmitted. The females seem to be unaffected by the presence
of the male secondary sexual characters and they are certainly
sexually functional as shown by the fact that all such specimens
taken during the breeding season were bearing eggs or embryos.
One of these broods was reared and about fifty out of the sixty
embryos were brought to a stage in maturity where their sex
could be determined. It is rather suprising that every one of
those brought to maturity was a female. Twenty-one were
carried without mishap to a stage when the secondary sexual
characters were fully developed and all were normal, showing no
traces of the conditions which existed in the mother.

It is noteworthy that specimens taken within a short distance
of each other at the same time should closely resemble each other
in their deviation from the normal. The specimens taken from
Turtle Creek in 1922 show the peculiarity in the first abdominal
appendages illustrated in figure 10. Those taken from Turtle
Creek in 1923 possess first abdominal appendages like those
illustrated in figures 9 and n while the seven taken from South
Kinnikinick Creek are all alike in having a single male character,
one hook on the left third walking leg. This would indir.itr that
the individuals resembling each other might have been members
of the same brood.

SECONDARY SEXUAL CHARACTERS OF CRAYFISHES. 2J3

To say that the irregularities in the secondary sexual charac-
ters may be transmitted is not to answer the question of their
origin. It would imply, however, that the cause producing
the effect was deeply seated in the germ cells and was not brought
about by a temporary, somatic influence.

LITERATURE LIST.
Faxon, Walter.

'81 Bull. Mus. Comp. ZoOl., Vol. VIII.. No. 13.
Faxon, Walter.

"85 M.-moirs Mus. of Comp. Zool.. Harvard College, Vol. X.. pp. 12-14.
Faxon, Walter.

'98 Pro .U.S. Nat. MOB., XX.. p. 683.
Hay, Win. P.

'05 ^;mih-..M.m M ;-• «-Ilancou- :ons. Vol. III., Part r, pp. jjj-j28.

Lbnnberg.

'98 / p. 334-335 and 345-

Nicholls, F.

'30 1' • he Royal Society 01 '.XXXVI..

.
Ortmann, A. E.

'05 \I' . Vol. II.. No. 10. pp. 37 iSS.

Rosseau and Desmarest.

'48 Amui!' .juc de France, : • VI., pp. 479

aii'l .jM .
Smith, Geoffrey.

'10 <jii.irt. .!• ''MM. ' ;.j. No. 216. pp. 577-603.

Smith, Geoffrey.

"10 nii.ut. Jiiurn. M . 218. pp. 225-.'.} i.

Von Ihring.

'9- i Mosi

Von Martens.

'70 Sii/unn>l'i.-riL-li: Gesellschaft N i lni.

274 C. L. TURNER.

EXPLANATION OF PLATES.
PLATE I.

FIG. i. Diagram showing the three basal segments of the five walking legs and
the first two abdominal appendages with the second pair of abdominal appendages
turned back. Normal male. I., II., III., IV., V., walking legs; A Pi and AP2, first
and second abdominal appendages; H, copulatory hooks on third walking legs;
S, external openings of sperm ducts at base of the fifth walking legs.

FIG. 2. Diagram illustrating the three basal segments of the five walking legs
and the first two abdominal appendages with the second appendage turned back.
Normal female. O, oviducal openings at the base of the third walking legs; AV,
annulus ventralis. Other explanations as in Fig. i.

FIG. 3. Diagram illustrating the three basal segments of the five walking legs,
the first two abdominal appendages and the peculiar features described in specimen
number 10. Structures may be identified by the explanations in Figs, i and 2.

FIG. 4. Diagram of three basal segments of walking legs and first two ab-
dominal appendages with other features peculiar to specimen number 9.

FIG. 5. Diagram of basal segments of walking legs and first two abdominal
appendages with other secondary sexual characters shown in specimens i to 5
inclusive.

FIG. 6. Diagram to illustrate the condition of the secondary sexual characters
in specimens 13 to 19 inclusive.

BIOLOGICAL BULLETIN, VOL. XLVI.

PLATE I.

ffi

\_D CSZ

C. L. TURNER.

276 C. L. TURNER.

PLATE II.

FIG. 7. Left first abdominal appendage of normal male. X 7-5-

FIG. 8. Left first abdominal appendage of normal female. X 7-5-

FIG. 9. Left first abdominal appendage of specimen number 9. X 7-5-

FIG. 10. Left first abdominal appendage of specimens I to 5. X 7-5-

FIG. ii. Left first abdominal appendage of specimen number 8. < 7.5.

FIG. 12. Right second abdominal appendage of normal female. X 7-5-

FIG. 13. Right second abdominal appendage of normal male. X 7-5-

FIG. 14. Right second abdominal appendage of specimens number 8 and

number 10. X 7.5.

BIOLOGICAL BULLETIN, VOL. XLVI.

PLATE II.

.

/

8

C. L. TURNER.

:

THE SURFACE TE.\M<>\ THEORY OF MEMBRANE

ELEVATION.

I.. V. HKILBRUNX.

When a sea-urchin e^sr i- in-eminated it lift- off a membrane
from it- -urface. A similar phenomenon can be produced by
various reagents. Many t •\planations have been advanced to
• •tint f<.r it. Some •>, I showed that all -ub-t.m.v-

which prodiirc membrane ele\ation are tho-e which misjit be
expe. ted to prod i H ea lowered surf ace tension (i . Al-o.in general,
sub 3 \\l)i« h lou't-r -urface tension markrdlv, pio.Uh i- nn in-

!)iMiu-flc\ aiion. And so I advanced the view that >urfa<v ti-n-ion
lo\vi-rii,. in .ill related to membrane ele\ ation, and I

olltnd a |>h\-iral explanation of the prOCCSS ba-eil on -lich a

loxverin- of surl 'en-ion. The idea \va- nc\v. although

'I 'ran In- u i had pre\ imi-ly -Imwn that wlien substances of a chem-
ical series are '"injured, those with lower surface ten-ion are
relati\elv nion- effective in producing membrane clc\ation.
I i .nbe thoii-ht ihe | involved a secretion on return to

\\ ater.

The -U!' face ten-ion theor\ of mem bra IK- elrvation Wa- .:

SOme Work | but not b\ other-. (",arre\- \ in a review

ol the literature -tale- th..t the theorx demaiul- the exi-telK •
a membrane on the unin-eminated I or tin- < iarrey tind-

no evidence, and he cite- Moor, ^ repetition of /iejn '-
and the llertuu-' ~ < .perimeiH with brokeii-iiji 6J an

argument a^ain-t -uch a membrane. The e\ idi'iice for a :
e\i-lem membrane i- abundaiitK -Upplied ill m\" I«)I5 pa[»er, tO
which ( iarrey doe- not refer.

In a paper published recently, Ju-l (8) claim- that hypertonic
solutions of sodium chloride in sea-water, \\hich do n<it lower
surface ten-ion, nevcrthele— ])roduce membrane elex'atioll.1

A- .i in.ittn <.! :.i, t I. . •ii'larilv

prod tin- h. \\rinl -iin.i. .• t<-n-i..ii. ii.r they cuu-r in. • -'.veiling and tlii- \^T«-'.

1. "\\ctrcl >uri.nx- trillion.

-77

2/8 L. V. HEILBRUXN.

In my 1915 paper, I pointed out that the earlier descriptions
of cortical change in the sea-urchin egg had failed to distinguish
between two types of cortical change. The normal change at
fertilization is a membrane elevation. Many reagents produce
this change, others however produce a swelling of the membrane.
The two types of change, although fundamentally quite different,
are not easy to distinguish morphologically, partly because the
surface of the egg is not especially favorable, for microscopic
observation. I therefore proposed various criteria to distinguish
them. Thus elevated membranes collapse in albumen solutions,
swollen membranes do not. Other criteria are easily established.
When eggs with elevated membranes are crushed, they flatten
and obliterate the perivitelline space. Eggs with swollen mem-
branes have little or no perivitelline space, and when they are
crushed, the thick membranes remain around them as before.

Just tried none of these criteria. The membranes he describes
in his paper seemingly lack a perivitelline space, for he notes that
on return to sea-water from the hypertonic solution, the perivitel-
line space is not obliterated. If the membranes were separated
off as Just believes, and if there were a perivitelline space, one
would think that this space would be obliterated by the osmotic
expansion of the egg on return to ordinary sea-water.

As soon as an opportunity was afforded, I repeated Just's
observations. With the concentrations he used (20—24 Per cent.
2]/2 M NaCl in sea-water), no membrane elevation or separation
could be observed. In such solutions the membrane could be
seen slowly to expand and swell. There was no evidence at all of
a sudden movement of the membrane such as occurs when the
membrane is elevated. Tested with albumen solutions the mem-
branes did not collapse. When the eggs were compressed the egg
contents did not expand to the outer limits of the membrane.
The membranes produced by solutions of sodium chloride in sea-
water, or on return from such solutions to sea-water, were
certainly swollen and not elevated. In the light of this evidence
it appears that the argument advanced by Just is not valid. //
still remains true that all substances which produce typical mem-
brane elevation do in every instance cause a lowering of surface
tension.

THEORY OF MEMBRANE ELEVATION. 2J9

It should perhaps be pointed out that in cases of extreme
coagulation of the protoplasm, the contents of the egg may some-
times be made to shrink a\\ ay from the vitelline membrane. Xo
good case of this is known for the sea-urchin egg, although some
shrinkage apparently occurs after prolonged heat coagulation.
In the Cnmingia egg. which has a much stiffer membrane than the
sea-urchin egg, a membra IK- which does not normally become
elevated, some 191- • rimeiits showed that the protopla-m

would -brink away tn>in the \itelline membrane when the e.u.u
(ontent- had been t lion m^hly coagulated after -r\eral hour-*
e\po-ure to h\ pertonic -olut i< >n-. Such slow < o.i-ul.tti\ e -hrink-
• el\ to be confused with true nu-mbrane elevation,
although in the ('until:, g under certain conditions the mem-

bi.tiie may bei oin- d away from the I'ter it i-

released.

Ju-t make- one oi her point against the surface ten-ion theory.
Il«- ~i. ii<- ih.u in "tin -if Arbacia and Echiinirachniit* any

* oiii]»eient observer car see thai membrane separation following
insemination i- no mere -miare tension effect." Thi- i- not a
\.i\ -ciiou- .dxumeiit . The surface tension theor) claim- only
that the lowered surl ;ion is the cause which. underlies the

pro, ess, not tint all the dri.iil- which follow the initiation , ,t t he
• trii-ion phenomena. A; \\.iyhowi.m

one decide a physical problem by mere visual obsen at ion?

Jll-l, ap|»areiltl\-, ha- attempted to do thi.-. The lifting «>M "t"
the membrane i- a physical pmce— and demand- a ph\>ical
explanation. |i ,-h to thi- i- hi- de-rripiion

of membrane elevation a- in\ol\in^ a "wave of negativity" (9).
Hut he doe- not ti-e the \\-i -rd in any known phy-ical -en

ki I I ki NI i

i. Heilbrunn.
'13 BIOL. BULL., XXIV., 15. ib.. XXIX.. i

. Traube.

'09 Biivli.-m. /,-it-. li.. X\ '[.. i

3. Godlewski.

Winterstein'a Handbuch dei \.-n;lrKh.-ii.l.-n I'lr. :

4. Garrey.

'19 Hi.u . Hi i i ... XXXVII . 287.

5. Moore, A. R.

'12 t'niv. ni ( alitomi.i I'ul.lu atiuns in IMiysiolojjy. I\'..

28O L. V. HEILBRUNN.

C. Ziegler.

'98 Arch. f. Entwicklungsmech., VI.. 249.

7. Hertwig, O. and R.

'87 Untersuchungen zur Morphologic und Physiologic der Zelle, Heft 5.

8. Just.

'22 BIOL. BVLL., XLIII., 384.

9. Just.

'19 BIOL. BULL., XXXVI., i.

THE INDEPENDENT DIFFERENTIATION OF

I ATKD CHK'K PRIMORDIA IN CHORIO-
ALI.ANTOIC GRAFTS.

I. Tin. KYI . NASAL REGIOX, OTIC Ri <.iox,

AND Ml -i:\CEPHALON.
I I I' ,11 HOADLEY.

'I HI. H' I! < ATORY. THE UXIVERMIY . •!• ( llh \t.O.

I. hit:

II. Metl

ill. I

i

i. ... -98

/'.It. i • • • •

IV. I '

v. -313

I. IVIRODUCTIOX.

Tin- r-iudy •>!" thf power of independent growth and diileivntia-
tii.ii i.f < -mbi \onir primordia (self-differentiation, R«>u\ has been
pur-iied l'\ various method-. The experiments on tin- de\ elop-
ninn i.f i-olated blastomeres, inaugurated by Ron--.'- -tudv of the
, ha\e o.ntnl)! ly to our kim\vK-<]-<' nf the

l(»»-.ili/.itii>u "t pot«-iui.ilitii-> in rK-.ivage f-1 More recently,

-!i^.iii«'ii \\tli .mil (littcrciiti.iiinn • .f embryonic tissues

IKTII (Mrrird on \<\ oiln-r iiH-ili«.d~. Tin- classical rxprri-

nu-nt- of Harrison (1907 a, [910 on the embryonic nerve cell

.imphilii.i ill li — uc ciilturi-. folhiwi-d l>v IHIIIHTOU> ^tudir- of otln-r
embryonic » rll- \<\ H.ini-oii (1912 , l.rwi-, M. l\. .md \\'. II.
MI i </, /', etc.)i -"id others, funii-h good examples of the specifi-
city and the decree . .f the iiide|>eiidein power of differentiation
of the isolated > ells of \.iriou- einl.rxcnic organs. Harri-on
(1907 &) and others ha\e .ilso made Mudies of the independent
power of growth and differentiation of limb primordia and other
part-- of the amphibian larva by a method ha\ing many ad-
ditional advantages, that of grafting. The part Mudied was

281

282 LEIGH HOADLEY.

grafted to another larva of approximately the same age. In
such experiments, the grafted parts continue to develop while the
graft is nourished and relieved of its excretions by the host. The
differentiation of the organs of the host, which are developing
simultaneously with those of the transplanted part, must, how-
ever, affect the differentiation of the transplant to a greater or
lesser degree.

The ideal conditions for a transplantation in which it is desired
to determine the independent power of growth and differentiation
of the part, would be in a location where there are no nerves from
the host, where there is no differentiation of host tissue to
influence the grafted part, and where there is a sufficient supply
of blood vessels to insure nutrition and the ability to excrete
waste products on the part of the transplant. This must be in a
rapidly growing tissue which is capable of repair, and such as will
react to foreign tissue with rapid incorporation and a host of
capillaries. The chorio-allantoic membrane of the chick furnishes
such conditions as has been pointed out by Murphy (1913),
Danchakoff (1916), Kiyono (1917), Minoura (1921), and Atter-
bury (1923).

It is my intention to attack the problems of independent growth
and differentiation by the isolation and transplantation of the
various primordia of the chick embryo to the chorio-allantoic
membrane. A general survey of the field has already been made
with a variety of tissues and has yielded promising results. For
instance, when a cross section of the body of the thirty-six or
forty-eight hour chick is transplanted, the grafts obtained show
that cartilage, bone, muscle, nervous tissue, and mesonephros
grow 'well. In one case of transplantation of the primitive
streak, nephrogenous tissue is found to be present. In many
cases of such grafts of cross sections of the body, small buds of
the feather germs form and grow well; they receive a rich
vascular supply. The present paper is confined to the organs of
special sense, i.e., the eye, nasal region, otic region, and one
example of brain tissue, the mesencephalon. The mesencephalon
was selected because of its close association with the eye, and
because of its high degree of differentiation in the birds. Other
parts will be considered in later papers.

CHORIO-ALLAXTOIC GRAFTS. 283

The progamme was suggested by Prof. F. R. Lillie of this
laboratory to whom I wish to express my appreciation for many
helpful suggestions and criticisms during the progress of the
\\ork.

II. METHODS.

In the experiments recorded in this contribution, I have used
chicks of twenty-four, thirty-six, and forty-fight hours of incu-
b.ition as tin- -ource n| material. The- methods of grafting
employed are similar to tlmse used by Murphy and Danchakoff
\\itti certain modifications developed at thi- l.ibor.uory. An i gg
of nine io ten day- of incubation is candlfd ,md the juncture of
two of the larger blood \e--els is marked. The -hell i- then -teri-
li/ed .nnl .1 -in II window i- rut through it at tin- point with a
h.H -k-aw bl.idr. 'I is then rfturned to the incubator \\lnle

the embryo which i- to furnish the material for the tr.m-plant.i-
don is put in physiol - It solution on a warm stage under ..

binocular dissection mil \>c. l;or illumi nation, t he light oi a

i-o-u.itt, nitrogen-tilled tungsten lamp is com cut rated on tin-
embryo by a bull's-! MK-nser. All dishes are sterilized in the

autoclave .md .:11 in-trumrnts are boiled in water. I )i~-c. i in-

die- whit h h.i\e been -round into fine kni\'f- .tie n-ed in the
• iper.it ion-. \\'ith tlie-e. tlu- primordium is i-ol.iied .md i- then
pi. ned on the . horio .!l.intoit- nifmbrane as near .1- po— ible to
tin- jniutiire ot the two vessels noted above. The -hell meni-
ie \\liich i- torn -lightly to admit the ti--ne i- repl.u-ed, the
shell i- returned and tin- rut edges are -e.iled with p.ir.iltiiie.
The egg i- then returned to the iii« nb.itor in -neh .1 po-ition th.it
the \\indo\\ i- do\\n, SO that the weight of the egg in-iire- the
contact of the nieinbratu- and the implanted ti--ne. After
twelve hour- lli« • gf is rotated, and at regular interxal- tli.
after rolled as any de\ eloping t-gg. M.ui\ of i he t ran -plant- are
-me to have more or le-- mechanical injury, and the time reunited
lor their incorp. .r..t i< m \\itliin the membrane of the- ho-t ,md
Mib-e(|iK'iu infiltration with blood ve— el- mn-t \ary. The-i- are
the nio-t im])oriant \ariable- \\ith which the method mti-t cope.
After -i\ or -even da\ - the egg is opened and the development of
the grafted ti — lie i- -uidied macro- and micro-. , ,; .ically.

284 LEIGH HOADLEY.

III. EXPERIMENTS.
A. The Eye.

A series of grafts of the optic primordium of the chick have
been made by this method. The material was taken from chick
embryos of twenty-four, thirty-six, forty-eight, and in a few cases
sixty hours of incubation. In some cases the entire primordium
was excised, in others only a part of it was used. There are
various difficulties to be encountered in such an operation. The
mesenchyme, at these early stages, is quite sticky, and consider-
able care must be exercised to remove the greater part of it from
the vesicle. Particular care is needed if it is desired that no
part of the brain be included in the transplants. At the twenty-
four hour stage, the neural folds are open in the optic region and
only a slight swelling indicates the position of the optic foveola.
The anterior end of the head was removed from such embryos;
the region from which the optic vesicles develop was then bisected,
and the primordium of each side was transplanted separately.
The lens ectoderm is not yet definable so that transplantation of
this region with that of the future optic vesicle is a matter of
chance. At the thirty-six hour stage, the technic is much more
simple. Here the embryo has from twelve to sixteen somites.
Some of the embryos of this age were more advanced, but the
majority fall within these limits. The primordium of the eye is
then in the vesicle stage and is easily removed. Great care must
be taken that the tissue is not harmed in the transplantation.

Another set of experiments was made on the eye at the thirty-
six hour stage. The main part of the experiment was the same,
but the lens ectoderm was dissected away. This is very difficult
to do, for there is a great tendency for the wall of the optic vesicle
to tear. In most of the cases which have been operated in this
way, no lens ectoderm was left on the vesicle, but one case which
later developed small portions of lens tissue is in doubt. The
mechanical injury in an operation, must, of necessity, be great,
and this is probably the cause of the greater degree of disorganiza-
tion found in these grafts. Grafts of the isolated lens ectoderm
were also made but no growth was found in any case, possibly
owing to the very small size of the piece of the embryo trans-
planted.

< HORIO-ALLAXTOIC GRAFTS. ~ -

With the experiments made from the eye primordium of the
forty-eight-hour embryo, greater difficulties with the elmination
of the surrounding tissue began to appear. It is practically
impossible to isolate absolutely the optic cup of this stage from
the surrounding mesenchyme without injuring the cup in some
w.iy. The optic cup can In- dissected out with little difficulty,
but in every case tin- re- may be seen a fur-like border of the
mesenchymal cell- attached to its surface. After a few , it tempts
to remo\e tin-, it was con-idered to be useless, for the manipula-
tion injured the rup in c\ery case. On thi- account, when a
tr. m-plantation of ilie optic cup is spoken of. it means that the
optic cup t-r with the immediate me-cnchyme w.i- taken.

Attempt-- were made 1.- in the experiment- with the thirty-

hour embryo-, t.. n nio\e the lens primordium from the optic
cup. It v. 'ii found that, though thi- i- po->ible, tin

; of me.haiiic.il injury involved. Such experiments
were therefore omitted. Since the sixty-hour "eye" i- in much
the same <"iidition, the -.line difficulties wen- experienced \\ith
it, but it \\a- not con-idered necessary to perform main- ex|
inn nt- \\ith the material I nun this age, for even at thirt\--ix
hour- of incubation. : s of the optic primordium po-

complete \»> \\emf independent self-differentiation.

In the grafts obtained from the foveola of the t \\ent\--four hour
embr\o, there i- much \ariability in growth and differentiation.
Thi- -eem- to be due to t' that there i- a < eitain amount of

cm -hi n- ini he n-mi»\ al of tin- ti--ue, thedi- :/atii>n prodi:

bv tlii- nu-( hanical factor incn-a-ini; a- the >i/e of the operated

pan d. In one of the. there is a primary differentia-

tion into tapetal cell- \\hii-h contain pigment granule-, and
retina-forming cells. Man\- of the graft- >ln>\v iu-r\-ou- ti--ue
but it is not -ulticicntly different iati-il to warrant any po-itive
r.tatement a- (<> it- iiatun-. There i- much other ti — ue tran--
planted with the optic primordium of thi- age. In One of [lie
;jralt- \\ell defined ^an^lion cell- are present. In no case, ho\v-
ever, does any eye j;raft of thi- a;^e -how more than the primary
differentiation into pi^mcnted ta|>etal cells and implemented
retina-forming cell-.

The Drafts of the optic \e-icle pin- the lens ectoderm of the

281

LEKiH HOADI.I.V.

thirty-six hour embryo will be considered next. At this stage the
optic primordium possesses great power of independent growth
and differentiation. Many grafts were obtained and sectioned,
a number of the most successful of which will be discussed here.

FIG. i . Graft of an eye from a forty-eight hour chick embryo in alcohol before
section, 19 E 13. (X 10.) For explanation, see text, page 290.

All of the cases which show any growth at all, prove, on micro-
scopical examination, to have a primary differentiation of the
tissues into a pigmented tapetal portion, a non-pigmented retinal
portion, and a group of lens-forming cells. The results vary all
the way from grafts in which there is no great differentiation of
the retina and lens, to the more successful ones in which the
retina is differentiated and the whole transplant appears approx-
imately normal.

The more successful grafts show a high degree of differentiation
though great variation in the extent to which the subordinate
parts of the organ have differentiated. It is an interesting fact
that the basal layer of the retina often shows many cells in active
mitosis, i.e., the graft is still in active growth. In the best graft
obtained (Fig. 2), there is a very clearly defined retina with t he-
layers developed sufficiently to warrant the statement that it is as
far advanced as the normal; the lens is perfect, showing the
anterior epithelial layer and the lens nucleus; the anterior and
posterior chambers are present; the tapetum and tin- i artilage

CHORIO-ALLANTOIC GRAFTS.

287

of the eye are typical of the normal of the same age. The ciliary
processes show around the lens. This case, 9 .4 I, is therefore
selected for detailed de-< -notion.

Alt

O.Se

A.C.-

p Alt

Oh.

I- 1-.. -•. Cl n from thir ur primonliiun.

.1 i. (X ao.)

i ... •.:.• .liranc. Ch.. • •• : mem-

.li.irv proc 0 St., ora serrata. P..\:;.. p.. in; ..i .UM. him nt

el tin- K'-'it i" t!"- 111.1:11 ; the membrane. P.C.. ; Ret.,
.in.

On Man h j \, 1'ijj. ! he 1 1| i tic vesicle of a thirty-six hour chick
emoved as free from adjacent part- as po^-ible .ui<l was
placrd in-.tr tin- juncture « •!' t \\ • • « -f the larg«.-~t l>lc.. •.! \i-d- on tin.-
< li..ri()-,ill.iiitnic TiU'inlir.nii- <>t~ ,t chick of nine and uiu-half days'
iiH-iili.itinn. Tl . tlu-n rrturiu-il i<> the incubator \\ '.

dr\rlo|uiu-nt ] >i< •• rcd.-d until April fourth \vlu-n the egg \va-
I'pi-nrd and tlu- .^rat't. which \va- lar.^r. n-ni' >\-rd. In m\- n<
taki-n at the tinir, I notice tin- remark that the ^i.tl't \\.i- hea\il\-
|ii-inented.

1'pon inicro-copical examination of t he >ei tion^, the de-ree of
differentiation of th< gl it ]>ro\ed to he (|uite remarkable. In
only a few place- i- mito-i-, in the ba-al retinal la\ er c..mnion. .
at place- where there are lar^e folds protruding into the po-terior
chamber. B\ far the greater jiart of the ti—ue h ised

pri'liferatini; and is differentiating. It may be well to rounder
this differentiation under four heading-; the chambers, the
tapetum and retina, the lenticular region, and the surrounding
tissue.

LEIGH HOADLEY.

The chambers in this case are well developed. A definite
anterior and posterior chamber are present separated by the lens,
the lenticular zone, and a thin layer of mesenchymatous tissue
which is covered with an endothelium. The entire anterior
chamber is lined by this tissue. The posterior chamber is large
and is lined throughout by the retina save at the anterior border,
where the lens and ciliary processes occur, as mentioned above.
It is not regular, however, for at intervals, as can be seen (Fig. 2),
the retina has proliferated to such an extent that there are folds
entering the cavity. In no place do these obliterate the chamber
as in some of the grafts.

The retina and the tapetum are well formed, though where con-
volutions of the retina occur, the tapetum does not always follow.
On close examination, the tapetum appears to be formed of three
or four layers of heavily pigmented cells arranged very closely.
All of the layers of the retina are present (Fig. 3) but the layer of

Mesen
— Memb.ejt.tv.m
St Op.

In Nuc
Out.Mol
. Nuc.

FIG. 3. Cross section of retina from same graft as Fig. 2. (X 180.) B.C.,
blood corpuscles. B.V., blood vessels. Cart., cartilage. Gaiig.L., ganglionic
layer. In.Mol., inner molecular layer. In.Xuc., inner nuclear layer. Memb.
Ext. Lint., external membrana limitans. Mesen., mesenchyme. Ont.Mol., outer
molecular layer. Out. Nuc., outer nuclear layer. St.Op., stratum opticum. Otlu-r
abbreviations as in Fig. 2.

rods and cones shows but little differentiation, a condition cor-
responding to that found in normal eyes of chicks of the same age.
Bordering the cavity is the very thin membrana limitans interna.
This separates the cavity of the eye from the stratum opticum
which lies just beneath it. This is clear an<l ea-ily distinguished
from the sparsely nucleated outer ganglionic layer. Bet \\een ihi-

CHORIO-ALLAXTOIC GRAFTS. 289

and the inner nuclear layer is an inner molecular layer, very
granular in appearance and containing only a very few scattered
nuclei which have failed t<> associate as yet with either of the
adjoining layers. The inner nuclear layer i> nearly three times
as wide as the molecular layer and is very heavily nucleated.
Separating thi- from the miter nuclear layer is the outer molecular
layer, also free of nuclei. A- mentioned above, the rods and c< >ne-
have not differentiated at this stage. At place-, e-pecially in the
•MI of ihe retinal fold-, there are connect ive tissue cells which
fill in the spaces between these folds and the tape turn.

The differentiation occurring in the lenticular region i- quite
strikin. . The tapetum continues anteriorly to a point

0. Se

Ret

ens

Cilp-

T^- Hnt tp;tn.

-Tap.

H rieserv- ^ . - :'."• ' '

I-n.. .}. Leu an.! 1.-:. >amc gi. ml .;.

:lu-liiini II. M, i >:.. li < HluT

:

ju-t nie<liad \t> the margin "t" the leu- wlu-re it douMe- <m it-elf in
a pi^nieinle— la\er <-l «'ell>. The <HM -errata, ciliar\' pr< -
and the iri> are pie-ent. The leu- it-elf i- >uppi«rtt-il l.y the
SUSpen-ury lii;anient .

Ill it> differentiation, the len- i- (|iiile typical (Fig. 4). It i-
i'\-.il in shape, heiu^ coiup<>r-ed <>f a wide nucU-u- nf liln-r- with
their niedialK- located nuclei, and an anterior epithelial la> « r.
Both are identical with the same part- of the normal of the >aine
age. The len- surface i- toward the ectodermal portion of the
chorio-allantoic membrane. The incorj)oration of the

LEIGH HOADLEY.

within this membrane was so complete that no injury was done
to these parts when the membrane containing the graft was
removed from the shell of the host.

Very little tissue save those already mentioned occurs in this
graft. The sclerotic coat of the eye is present throughout. To
one side there is a small mass of brain tissue containing typical
nerve cells and processes. This is heavily infiltrated with blood
vessels and not easily identified. Adjacent to this region, the
tapetum and retina are very much convoluted and disorganized,
and a heavy infiltration of blood corpuscles among the folds
makes the identification of definite structures impossible. The
optic nerve and pecten which would be located in this area under
normal conditions cannot be found. The disorganization and
the infiltration of the blood cells may mask their position as
degenerative changes are apparent. The entire graft is sur-
rounded by the mesenchymatous tissue of the host.

In not all of the grafts is brain tissue present. A graft of the
optic region of a thirty-five hour chick, case 14 A I , when sectioned
shows a very evenly shaped posterior chamber lined by well dif-
ferentiated retina. This eye shows the same differentiation of
retina and tapetum as graft 9 A I above, though there is a total
absence of brain and cartilage tissue. The various layers of the
retina are well defined. The cut edges of the excised optic vesicle
have drawn together and the point at which they have fused is
marked by a thickening and by slight convolutions of both tape-
tum and retina. No lens tissue is present in the sections, prob-
ably due to the failure of its primordium to be incorporated. The
grafted tissue is surrounded by an extremely thin layer of host
mesenchyme. In no place is there any sign of brain tissue, optic
nerve, or pecten, though the point at which they would normally
occur is definable.

Graft 19 E .13 from a forty-eight hour chick embryo, is some-
what similar to 9 A I above. This case, after seven days of
growth, a total of nine days, showed a graft 4 mm. x 4 mm. in
size. A drawing of this graft as it appeared in alcohol before
sectioning is shown in Fig. I. The heavy'pigrnentation of the
greater portion of the surface will be noted, together with the
absence of pigment from a portion of the tipper face. It w.i-

CHORIO-ALLAXTOIC GRAFTS.

291

thought on macroscopic examination, that this would prove to be
the lens, but on section, a very different picture presented itself.
The implemented portion is the point at which the optic nerve
leave- 1 he retinal area and runs out into the mesenehyme (Fig. 5).

o c

:3S-Ret

.

•

'

:pfl

• ,•';-" xjvrjVS

- * ^>K»^-

'•'•/t-Tap

•• / v$W-

,~r

--Mesen
•. .. '

--
, - '

^

' -r;.*5»'-~ -

. ;, i nerve in graft, i, > . i.

( X 140.) flr.. l.ram. i :.itje. Opt.\.. • .

( >ilu r al'tii'

The |M •« ten i- |>n-t m though irregular in -hapr, due no doubt to
tin- nuvli.niiral Stress in t hr growth after Iran-plantation. 1 r<>in
this point the optic nerve gri >\vs out between the tapet inn and the
eriddennal wall of the chorioiiic membrane for c|uite a di-lance.
Tlu- otluT white area in the ^ruft is brain ti — tie whifh was ti.m--
planted \\ith the optic cup. Only in the region of the ner\e and
nervous ti--ue i- there any mesenclu me or cartilage in abundance.
In the re-t of th( .the tapet uni comes uj> cio-e to t he border

ot the chorioni* or the allantoic nieseiu'lu me of the nienibraiie.
The l.i\-ei> of the retina are all well differentiated. The ora ser-
rata appear- near the margin of a much di-torted leii-, ending «\\
it- lenticular border in a much convoluted and di-or^ani/ed
ciliary process. The mi--hapcn appearanci- of the len- i- ap-
parenth- due to the fact that it lie- next to the -hell membrane and
20

292 LEIGH HOADI.hY.

is not completely incorporated by the chorionic overgrowth. At
both ends of the sectioned material, where the stratum opticum is
cut nearly transversely, the nerve fibers of this layer show very
well. The whole graft compares \vry favorably with the normal
eye of the same age.

In the experiments performed on the forty-eight hour chick, as
in those cited for the thirty-six hour material, the degree of
regulation seems to be dependent on the amount of injury to the
engrafted tissue, both during manipulation and during the growth
period which follows. My data show many cases where a much
convoluted retina and tapetum are present with no posterior
chamber showing. All degrees of lens malformation may occur.
The retinal cells are always definable as such, though the degree
of differentiation of the retina as expressed by zone formation
varies considerably.

Five successful grafts were obtained from thirty-six hour pri-
mordia from which the lens ectoderm had been dissected away.
The data on these cases are to be found in table I. In only one of

TABLE I.

GROWTH OF THIRTY-SIX HOCR OPTIC VESICLES WITH THE LENS ECTODERM

REMOVED.

• Case. Control Age. Remarks.

15 A i 83 days Posterior chamber, retina and tapetum present.

No lens tissue.
15 B i 8\ days Nervous tissue, cartilage, much convoluted retina,

tapetum, small amount of lentoid tissue.

15 C i 8£ days Very little area of much convoluted retina and

tapetum only.

16 B i 8 days Nervous tissue, retina and tapetum. No lens

tissue.

16 B 2 8 days Cartilage, retina, much convoluted tapetum. No

lens tissue.

these grafts are any lens cells present. It is more reasonable to
suppose that these are the result of the proliferation of some cells
of the lens ectoderm which were not removed in the operation,
than that they arose by a redifferentiation of the cells of the
retina or iris (see Fischel for amphibian larva?, 1900) or by a stim-
ulation of the chorionic ectoderm by the optic cup (see Lewis i<i<>7
and Spemann 1912, and other papers on the amphibian <.-y(A
Spemann (1912), Lewis (1907 a, b, c,), and other w<>rkrr- have

CHORIO-ALLAXTOIC GRAFTS. 2Q3

found that the optic cup of the amphibian embryo has groat
power of self-differentiation when transplanted to other parts of
the same or another eml >ryi », while the lens ectoderm is dependent
on contact with the optic cup for its development. The speci-
ficity of the ectoderm which forms the !• ries with the species
employed. Lewi-; PC 7 otes that the various parts of the cup
develop independently "in marked contra-t" to the lens which
will not form unle— the optic cup is present. If brain also is
piv-ent , tht TV may «r may not be an optic nerve developed. The
tissues -rrm to coniinuc t lu-ir own specific type of differentiation
independenl "t" one .mother.

Tin- f.id that in • ase 14.1 i where no brain ti— ue i- pn-sent,
tin- optic ner\e i- also Kicking, while in case 10. /•." i;, both
pn eems, th< -erve more than pa— ini^ noti

In the lir.-t meiit i< >n< • , the point of fusion i>f the cut i

the primordium i- definable. There is no indication of the optic

n\ \\here in thi n. In case 9 A i. while tin

-mall .iiiioimt i .1" br.iin ti--ue present, the po-itimi of the ojitic
iiciM- (.iniioi be found. It may be obliterated by the 14:
numbt-r »t dr^nu-i :i\olutions in the retir i t.i|)ctuin

adjoining tin I -e 19 E 13, however, bot h optic in i \ r

and pi ( ti-n ai ilrtinitely located. The position of the optic

nei\c i- tin- -.inn- .1- th.it of the optic stalk which connect- the
ic cup \\ith the brain in the early stages of de\ elopnieiit .
When thi- i- removed, the optic nerve fails to dc\dop. The
fibers oi i he < ell>, in-te.;d < •!" finding exit by this path, .ire retained
in the -tr.itum opticiim. It would seem that the-e -trucuin •-
we'll as the p.irt- of the retina and lenticular re^io- from \

delmite loc.n ion- in t he |>rimordium.

The experiments show very definitely that the \ariou-;
the e\e | H iniordiiiin have the power of independent self-differenti-
ation to quite a remarkable extent. Thi- mnrlu-ion i- >tren-th-
encd l.ty the findings of Lewis, Spemann and other- for the amphib-
ian eye. The regulation of the part- to form a more or le-- ;
feet -tructun -'.ins, on the other hand, to be dependent on
vari< >u- mechanic.il factors involved in the experiment.

29-J LEIGH llo\]i| I Y.

B. The XasaJ Region.

The material for the grafts of the nasal region was taken from
chicks of thirty-six, forty-eight, and sixty hours of incubation.
As in the case of the eye tissues, only a few grafts were made of
the sixty hour material. Inasmuch as it was desired to note the
growth and differentiation of the nasal epithelium especially, no
twenty-four hour material was grafted, for at this age, the medul-
lary folds only are localized.

When grafts were made of thirty-six hour chick primordia, the
anterior end of the head as far back as the optic vesicles was
transplanted. If the material was taken from an embryo of
forty-eight hours incubation, the tip of the head was cut away so
that both nasal pits were transplanted in one piece.

A brief account of the normal development of the olfactory
organ based on Disse's account (1897) will form a useful intro-
duction to the study of the grafts of this region to be described.
After three days of incubation, the sensory epithelium of the
chick nose is represented by two thickenings of the epithelium in
the nasal grooves. As the embryo approaches the fourth day,
the epithelium becomes differentiated into columnar epithelial
cells and round cells, more transparent in sections stained with
carmine. These cells are isolated in groups which are more or less
distinct from each other. It is from these that the cells migrate
out, forming the olfactory nerve and the ganglion-like group.
At first a chain of cells appears to lead from the nasal epithelium
to the brain. As late as the sixth day, this nerve primordium
contains cell bodies but they are most numerous at the extremi-
ties. In the eight day chick, the stage which serves as the con-
trol for the grafts to be described below, Disse finds the nerve
formed and two types of cells appearing in it at rare intervals.
There is a pear-shaped unipolar cell, and a bipolar cell that is
spindle-shaped. The greater part of the nerve is composed of
nerve fibers. The cells are most numerous at the ends of this
fiber bundle. At the distal end, near the nasal epithelium, the
grouping of these cells and their shape led Heard (1885) to
describe a ganglion-like mass which Disse describes merely .1- ,i
group of cell bodies. Disse (confirming His, 1889) concludes that
the nervus olfactorius arises from certain cells of the nasal
epithelium and adds that it lacks a ganglionic portion.

CHORIO-ALLAXTOIC GRAFTS. 295

The cartilages of the nasal capsule of the chick have been de-
scribed by Colin i I'jo.v. The prenasal cartilages and the nasal
septum an- pre-ent on the eighth day. The primordia of ihe
three turbinal> arc- formed. These are the only cartilages which
arise from the mesenchyme transplanted in the graft-.

Graft 41 B i i- a growth of the nasal region . .f a chick of thirty-
six hour- of inrnb.i; ion. The graft was removed after -i\ days of
growth on tin- chorio-allantoic membrane of the ho-t. a total of
• ii and one-halt da\ -. and was 3 mm. \ 3 mm. in -i/e. ( )n
-ection, the varioti- j. • the nasal region are found to be

present. Tin •> compare 1 ,\»rably in their development with
the normal region in the embryo of eight days of incubation.

The i;r.ihed dssues with the exception of a portion of tin- ecto-
derm are well -u r rou u< led by the mesenchymatous t i — in- of the ho- 1
membrane. In it. Ion; - areas are definable, the nasal -

the na-al < a i • tin- l»tebrain, and the epidermi- of the head.

With the exi epiion of the epidermis, these have maintained ap-
proximately their normal relationships. The section- < nt the

on transverse!) so that at one end the nasal
-epaiat< -d by the cartilage of the nasal septum. As the -cction-
an rly, the sacs become surrounded !>\ a <

tilai^inou- -heath, the prenasal cartilages. As the na-al sacs
di-appear | >«.~T eri< irlv . the forebrain appears. It i- hea\ily
in til (rated \\ it !: bli-.d D 'rpn-ck-s \\hich do not seem t" be located

in an\ organized vessels Between these structures themseb

and -epaiatin;^ them from the chorionic and allantoic boi-der- of
the lio-t membrane is 1 growth of me-eiich\ me. A more

detailed de-cri] >t ion . .f these parts follows.

The na-al sacs are t \\'o in number and di-tinct I Tlu-\

end blindly at both ends and an- com pie t el \ -urroundi-d b\ - me-en-

ch\ me. save at one place where the left -ac retain- it- connection
anteriorly \\ith the epidermi- of the he..d. Po-teriorly t lu-\-
\\ iden out forming cavities of some size. 1 he epithelium which
liiu-- the cavitie- i- columnar in t\|n . Mito-i- i- fairK i ..mnion
in the entire region but i- more commonly found near the cephalic
border where the epithelium i- noticeably thickened. I'or the
nio-i part thi- la\er i- -harply detineil from the -urroundin^
me-ench\ inc. SO that it appear- a- if a membrane were pre-ent

296

LEIGH HOADLEY.

surrounding it. In the thickened .ire. is, however, the boundary
becomes indistinct, and in many plures, cells may be seen which
appear to be migrating from the epithelial layer into the mesen-
chyme. These areas represent the sensory epithelium and from
them the olfactory nerve arises. A short distance from the nasal

FIG. 6. Cross section of graft of nasal region of chick embryo of thirty-six
hours of incubation, 41 B i. (X 140.) All., allantoic border of membrane sur-
rounding graft. B.C., blood corpuscles. Br., brain. Epderm., epidermis. Mesen.,
host mesenchyme. N.Epith., nasal epithelium. .V.5., nasal sacs. Nas.Cart.,
prenasal cartilage. O//..V., olfactory nerve.

epithelium may be seen a small group of ganglion-like' cells
(Marshall, 1879) and toward these the nerve cells appr.n to be
migrating. From the ganglionic group, nerve hlu-r bundles o>n-

CHORIO-ALLAXTOIC GRAFTS. 297

taining a few cell bodies which correspond to the pear-shaped and
spindle-shaped nerve cells of Disse, run to the forebrain where
their course is lost in the disorganization present there. The
«• round primordial nerve cells of the nasal epithelium occur
in the liber bundle- at rare intervals. The development of the
nerve com-pond- '" thai "f the embryo . .f seven >f incuba-

tion as de-cribed b;. 1

The l"<irebr..in of the -raft has no limiting membrane. The

character of the cells is undoubtably that of brain cells but they are

unorgani/ed. Tlii- rs to be due to the fact that thi-ti-- lie

Pensively intiltr.ited with blood ve--el-. In places there

i to In filled with bl 1 cell-. Ntimer-

-ii). ill nerve-lik. found in thi- region which

formed by union of the groups of the nerve pi They end

bluntly and blindh M netrating the nu-eiichyme for -hort

di-' . inbliii^ -imilar structures found in graft- of the

im 'haloii which will be discussed below.

The cartilage \\hich -tirroiinds the nasal sao has been men-
tioned above. Th: ge is in a single large piece bin at
interx al- in t 'u-re are breaks in its continuity. The
na-al -eptum i- pie-eir I'-shapcd structure, i he i op. ,| tin 1"
forminj it of the iloor of the region. Forming a roof over
the na-al nt--haped prenasal cartilage. Thi
broken at a -in;Jc place .-n one >ide to permit the e\it ot a b!

vessel. The cartilagt does no( jx'rsist through the entire graft
but disappear- -hortU" be\ oiid the posterior end ot the na-al -

The epidermi- of the head i- present in the .interior end of the
ft as an extensive, mon -s cornitied m..-- <-f ti--in- imbi-d-

ded in me>ench\ me. A- it i- traced posteriorly it approai hesand
tinalK', at the p<>-teri<>r le\ el of the olfacto: 5, reaches the

chorionii- surface "f the membrane where it i- unincorporated by
the h«»t. A -ingle lube-like growth of thi- epidermi- ha- come
to lie between the tip of the left olfactory sac and the forebrain.

A- far as I have been able to di-c<.ver. the only data oil the
differentiation of the na-al epithelium when i-olated either com-
pletely or together with the immediate region, .ire furni-hed by
Lewis (PC-; ,; . This author record- two experiment- which
-how in the lii>t place that t he epithelium develop- indeitendently

298 LEIGH HOADLEY.

of the forebrain, and secondly, that the nasal pit is capable of
independent differentiation. Both sets of experiments were done
with the larva of Amblystoma. In the first type of experiment,
though the forebrain primorditmi was removed, the nasal epi-
thelium sent out nerve processes into the adjacent mesenchyme,
developing normally. The second type of experiment involve the
transplantation of the excised nasal pits to a position beneath the
ectoderm dorsal to the eye. Lewis (page 464) states that
". . . differentiation of the organ continued and nerve fibers
were sent off into the region between it and the ectoderm, . . .
but extend only a short distance."

In the grafts obtained from the transplantation of the nasal
region of the chick embryo, the different parts of the organ show
a high degree of specificity in the course of their differentiation.
In comparing them with the control and with the description of
the process of nerve formation by Disse, it was found that they
show a condition comparable to growth in a normal embryo of the
same age. The cartilage which is present, is, without a doubt,
that of the transplanted part, and not of the host. The nasal
sacs are present in nearly their normal relationships to the rest
of the parts and have a differentiation closely resembling that of
the same parts in the normal organ. The nasal epithelium gives
rise to the nerve which is also present. The nerve cells of the
forebrain differentiate nerve processes, the abnormal appearance
in the sections being due to excessive vascularization and the
absence of the limiting membrane, both of which are mechanical
factors controlling the limits of differentiation to a great degree.

C. The Otic Region.

Grafts of the otic region and the otic vesicle are much more
difficult to make than those of the nasal region. The auditory
pit appears about the thirty-six hour stage, and the vesicle is
closed about the forty-eight hour stage; the material grafted
was taken from embryos of the last mentioned age in most cases.
The first experiments were made by shelling out the otocyst and
transplanting it, but it appeared very soon that the oto< \ -t .ilone
does not persist well in grafts. The mechanical injury due to the
pressure of the egg itself, together with the adhesive quality of

CHORIO-ALLANTOIC GRAFTS.

299

the shell membrane due to its absorptive nature, resulted in no
incorporation in the membrane of the host, and consequently no
growth. I then tried transplanting the otic vesicle with the im-
mediately adjacent mesenchyme, a slightly larger piece of tissue,
and found that in -uch operations, though incorporation may not
be complete and the mechanical injury more or k--- extensive,
nevertheless, growth en-ued. The investing tissue trail-planted
at the forty-eijit hour -tage includes only me-enchynie and
somatic ectoderm. In .'.11 the operation-, the brain wall •
removed from ilie trail-plain before it wa- |)l.iceil on ihe mem-
br.me of ho-t .

Table II. is a i' • "t 'I of ihe -uccessful growths of the auditory

TABLI II.

1 ' THE OTIC Kl •

Case.

-•-»/' 7

<'> In

in crescent ;ii««un«l fpitlu'liiiiii nt "tu-

nun ilittiTfiuiat<-i|. Ino'in-

•<• incorporation.

hi

inclept-ndont oi 'lui. n-nti.it.. :

ilii-lium. Ini-oriiji]. it ii.

in.

latcd epitlii-liiim in tv. um-

> artilaK' . In. ..inpl.-t,- in-

•••ration.

lir.

near though n I..IMI-

incorporalfl. iliit'<-rciiti.iii-'| .-pith. -

lium.

iir.

near cpitln lniin. Small gi

iiuin dirTcn-i

hr.

< »nly ~inall am

tin 1 111 III wllicll :

\oB6

In.

antl cpithi-litun al'.tn- pn-i-m.

thi-lium differentia'

vehicle. It will be noted that one of the-e i- from a donor of
thirtv--i\ hour- incubalion, but the embryo >ho\ved .1 more
a-l\anced Stage of dr\ elopmetlt than i- u-nal for (hick- of that
age. It corre-pomU more nearly to the fort \ -foiir-hoiii ,nd

( oiiM'(|iientl\- a|)|)roache- the u>llal don, ir age, forty-ei^ht hour-.

T\\o criteria are einp!«'\ed in tin- identification of the par
growths obtained in grafts of the OtOCVSt. The illorph, .logical
arrangement of the \ ari«ui- -tructure- aid- -reatK". but owing to
the di-or-ani/atiot) j.re-i-nl in every case, this is -upjtlcmented b\-

300 LEIGH HOADI.KY.

close histological examination. In normal development, the
cartilaginous otic capsule surrounds the sacculus, utriculus, and
their derivatives, while the ductus and the saccus endolymphaticus
are free in the mesenchyme. The histological differentiation of
the several parts of the membranous labyrinth in the grafts is
essentially the same as that of the control so that identification
can be made positively by histological structure; this is supple-
mented as stated above by the general morphological relation-
ships present.

The differentiation attained by the grafted tissues is of two
kinds, cartilaginous and epithelial. Xo ganglion cells or other
nervous tissue of any kind can be found in the grafts. This fact
emphasizes the isolation of the primordium before its transplan-
tation, for the ganglion and the otocyst are very closely associated
at that early stage. The cartilage partially surrounds the epithe-
lium, the latter forming the lining of cavities of various sizes.
The relationships of the various parts of the organ is not normal.
I have selected for description two cases which show quite an
extensive specific differentiation of the membranous labyrinth.

Graft 32 A 5 is a growth obtained from the transplantation of a
forty-eight hour primordium which has remained in the host
membrane for six days, a total of eight days growth. Completely
surrounding a large epithelial sac is a capsule of cartilage which
is discontinuous in two places. Through one gap, small blood
vessels enter a narrow group of mesenchymal cells which separate
the epithelium of this region from the cartilage surrounding it.
The other break in the cartilage sheath is in a place Avhere an
incompletely incorporated portion of the epithelium makes a
connection with the epithelial sac mentioned above. The car-
tilage, from its general relations, is undoubtably the otic capsule.

The epithelium of the membranous labyrinth shows a high
degree of specificity in its mode of differentiation. It may be
divided into two main parts, that incompletely incorporated and
outside of the cartilaginous capsule, and that within the capsule.
The epithelium which is outside of the cartilage is histologicullv
identical with that of the normal saccus endolymphaticus, the
part connecting the saccus endolymphaticus with the epithelium
of the capsule, the ductus endolymphaticus. These two portions

CHORIO-ALLAXTOIC GRAFTS. 3OI

are surrounded by mesenchyme only, as in the control. The
parts within the cartilaginous capsule show a high degree of
differentiation also. In the region of the vascular foramen refer-
red to above, the epithelium has the characteristics of the normal
utrirulu- of the ei.^ht day chick embryo. This determination i-
further confirmed by the appearance in the epithelium of three
thickened which represent the cristae acusticae of theampul-

lae of • raicircular canals. No ganglion cells ai ociated

with thi in the controls. There are no canals or ampulla-

forim-d though at a point o.cupied by one of the cristae I ig. 7.

~< • ;

C-^:
'

I ic vesicle of forty-*-:.

lit days growth. 32 AS- (X 23; /•' I.

:ng beginning of can. i Cart '.la^c

•i. . .i|i-i:N. M .chyme containing M I v-

.'u-liiim.

Compare Fig. s there is a slight evagination which suj the

inning i-l" -in h i"rmation. It is very limited in extent,
appearing for only a l\-\v sections. Thn 'ii^h the -paci- formed by
tin- second break in the cartilage, the 1 C liiu-d b\- utricular

epithelium i- ronm. led to the endolymph.itic derivatives men-
tioned abo\e. A few -ertions be\'ond this the wall of the sac
-li«>w> a differentiation into ., -tratilied -arcultis-like an<l a thinner
utricular epithelium. The ductus cochleari- and the lagena
which normally ari-e from the sacculus are absent in the craft.
It may \ery well be th.it the presence of the cartilage sheath has
arted a- a mechanical obstruction to t heir formation.

- 2

I. HIGH HOADI.KY.

The same definite differentiations of the epithelium show well
in case 40 B 6. Here, as in the case cited above, both cartilage
and epithelium are present. The cartilage is not so extensive,
however, but is located near, though not surrounding, the epithe-
lial tissues. The relations of the parts of the membranous
labyrinth are very much disorganized in this graft, the tissue

S.Ep

G.c.

. mv-

.....

;<7 •:* * 0 *.*>•,•• . - . -•

FIG. 8. Section of crista acustica of one of the ampullae of a chick of eight
days incubation. ( X 235.) G.C., ganglion cells. X.T., nerve tract running
from crista to brain. Other abbreviations as in Fig. ~.

appearing in two places, separated by the host mesenchyme.
The character of the cellular differentiation is distinctive. In
one of the groups, the differentiation is like that of the ductus and
saccus endolymphaticus. These structures are connected, the
ductus appearing as a tube leading from the saccus. Xear these,
a canal-like formation appears for a few sections These parts are
not widely separated from the other group of structures which
are adjacent to the mass of cartilage. The parts in this group
are in the form of distinct vesicles and tubules, resembling
histologically, utriculus, sacculus and a small portion of a canal.
In the utricular wall are two areas interpreted as crista? acu>ti-M-.
one of which is associated with the canal-like portion already
mentioned.

It would seem therefore, from the data given for the a!»o\e
cases, that the differentiation of the constituent parts of the otic
vesicle is very specific, and that it takes place independent of i he
relationships of the various parts, which depend on mechanical
factors.

CHORIO-ALLANTOIC GRAFTS. 303

Streeter (1907, i<H4 ', who has experimented extensively on
the development of tlu- i-.tr vesicle in the amphibian larva, has
noted the ability of the vesicle to develop in a nearly normal man-
ner v. hen transplanted to other portions of the head either of the
same or oppo-ite -ide. In most of the • ses, it reorients itsel
to dorso-ventrality and retains its definite character as from the
right or the li-fi -idr. In Spemann's experiment- I'lio), the
inverted otocy-t di<l not orient itself in every ca-e, <lue. in
Si reeter's "pinion, to the fact that in making the large inci-ion for
the rotation of the vesicle, the adjacent ti->ue wa- mneh dis-
turbed and al-o to the f..ctor of mechanic..! pre--ure which was
applied to in-nre the closure of the wound. Streeter eliminated
the-i- !'..< tor- by makiiu -lit and stretching t hi- for the opera-

tion, ("lo-un-aiid hi-alin^ t hus took place immediately \\ it hont
the applie.ition of prr—tire. Spemann's results do not otherwise
seem to be in conflict with those of Streeter. The n-ult- of both
1. 1 the-e iii\( . -i/e the high tie. ' independent

>elf -differentia i ion obtained by the organ. Street < r notes i-u i.
' All our e\ idence points to a high decree of differenti-
ation <.f the cell- of the \e-icle, and it is conspiem m-ly jiro\eii by
their i on o| l.iier.ility, . . ."

It will be of interest to note also a case cited b} Lewis [907 .
in -pe.ikiiu .-f \\hirh he -t.:tes that an otic < .tp-uh- of the \\»-[.
Ainfilystonin. farmed .iround the membranoii- labyrinth .iri
fn.m tran>pl. nited ti—ue from Rana sylr<iti«i. Hi- .nld- that the
character of the i artilage • 'ell- i- -uch .1- to determine them as "f
.\nibl\-toiu.il origin. !.e\\i- eon-ider- that tin- -how- that in
\\.i\, the otir \\--icle -tiniulate- the formation ,,f (he .
n- ca|»-uli- around it. He says (pagl I |5 . "It t he •

tilaginous capsule about the otic vesicle is dependent upon the

inlliu-nee of the latter for it- origin it will |>robabl\ be found the
'ilai^e of other tt-i»ii- of the einbr\'o i- like\\i-e de]-eiident on
•ain inthience- in the neighboring -tr net lire- for it- origin Iroin

the ine-ench\ me \ piece of trail-planted brain or a

trail-planted e\e, for example. e\ en when I'lo-e to the cartil
•iiini; about the otic \e-icle or central ner \ou- -> -lein do,-- not

stimulate the formation or growth of cartil .• 'in it-elf."

In the i;raft- of the otic ve-icle de-cribed above, the c'artilage

304 LEIGH HOADLEY.

is identified as that of the otic capsule but it is very definitely the
result of the differentiation of the cartilage-forming tissue of tin-
graft and not of the host mesenchyme stimulated by the mem-
branous labyrinth. The very fact that it only partially sur-
rounds the parts of the vesicle proves this. It will be recalled
that a fringe of mesenchyme was transplanted together with the
otic vesicle. This would easily account for any cartilage present
in the grafts.

The specificity of the cellular elements of the otic vesicle is very
marked (see also Streeter, he. cit.). Utriculus, sacculus, endo-
lymphatic duct and sac, and the canals are recognizable by the
character of their histological differentiation. Though the canals
have not formed in any of the cases, their general position is
indicated by the occurrence of the cristae acusticae of their ampullae.
In no case are they accompanied by ganglion cells. Neither is
the auditory nerve present in any of the grafts. This indicates a
complete isolation of the primordium, for the eighth ganglion is
closely associated with the otocyst even at the stage used in the
operations. There is therefore no possibility of a stimulation of
the epithelium of these sensory areas by nerve cells. These areas
are localised in the epithelium itself. By means of various graft-
ing experiments, Harrison (1903) tound a similar differentiation
of a sensory area, independent of nervous stimulation, to exist
in the development of the lateral line organs of Rana sylvatica and
Ran a palustris.

In spite of the much altered relationships of the parts of the
membranous labyrinth and the surrounding mesenchyme as they
develop in the grafts, they differentiate so that they are easily
identified in the sections. The amount of adjustment of the
relationships of their parts is, on the other hand, very variable,
being dependent on the amount of mechanical injury and the
position of the grafted tissue at operation, together with the
factors of pressure, and stress and strain during the following
growth period.

D. The MesencepJiahn.

As has been stated above, the selection of the mesencepliulon
in making grafts of brain tissue was made because of the- fart th.it
it is here that the correlation centers of the eve have tlu-ir seat.

< BORIO-ALLANTOIC GRAFTS. 305

In the adult bird it form- the large corpora bigemina which give
off ii" peripheral nerve-.

In the embryo, at about the forty-ei-ht hour stage, the brain
wall in this region i- -tilticiently independent of the ectoderm to
ble the i-olation of thi- part of the brain with very little or no
mechanical injury. It is comparatively :vmove all of the

meteiicephalon and the diencephalon from Mich pieces, so that
the Iran-plant.- contain only mesencephalon. Since Mich
material \ ield- ^r.ift - fa \orable for this study and M'IUV the injury
at an earlier levelopment is of ile« e— ity much greater,

the entire 'i-riment- with this priniordium wa- n

with t i --ue from chii k t hours of incubation.

I or the -ake of i oinpari-on, a brief description of the ine-eti-
i i ph. ilon of tin- ei^ht-day chick will be given here. The whole

•in i- di\ ided into \\\,< portions, a basal and a tect,:l. T
Mirroimd three < onnei ted . a\ ities the larger one- bein^ tin- u\-

md the -in, die' the aqueduct of Sylviu-, which is

\\ it hin t lie IM-. 1 port ion of -the region. The differentiation of the
wall of the (oipoia bigemina, the tectum lobi optici, i- \erv
del:- It \\e exami- -ectioil of part of thi- region, we

find that tin: -rtical layers presti > . T

will be de-'i ribed in tlu- order in which they are toiind on exami-
nation in the ependyma to the periphery.
Next the UK an ependyma which is densely nucleated
.md fairly \\ide. IP 'in it, cell bodies appear to be mi.
toward the per:: The next layer is an inner liber /one
which -tand- in j trast to the nu- ; /one ju-t nieii-
tioned. It i- i.itlu-r wide and contains cord- <ompo-ed of the
mui a ting • e!l bodie- and their processes. The mi-ration of t!
cell bodie- end- in a nuclear layer which lie- just peripherad to the
inner fiber la\ i r. The processes of the-e cell- extend -till further
out, formin- an outer fiber layer which i- much narrower than the
inner liber layer and < -mtains no nuclear -trand-. Bounding the
outer liber la\ er i- a membrane, the external limiting membrane.
ontheouter -ideof which is a thin layer of mesenchyme contain-
ing blond vessels.

The outer liber layer contain- processes of thi <-ll bodie- of the
outer nuclear /one. The-e are -hown very clearly in tangential

306 LEIGH HOADLEY.

sections of the lobe, in which fiber tracts appear, running for some
distance across the field. Longer fiber tracts appear in the inner
fiber layer. These run into the base of the lobes, through which
all of the innervation of this region finds its exit. The basal por-

v.s.

Mesoc.

FIG. 9. Cross section of a portion of the tectal part of the mesencephalon of a
chick of eight days incubation. (X 180.) E.L.M., external limiting membrane.
Ep.N.L., epenclymal nuclear layer. I.F.L., inner fiber layer. Mesoc., mesocoele.
M.N., nuclear strands. O.F.L., outer fiber layer. O.N.L., outer nuclear layer.
V.S., vascular sheath in the mesenchyme.

tion also contains fibers of the ascending and descending nerve
tracts. No peripheral nerves are present. The vascularization
of the region is accomplished by small vessels which enter the
nervous tissue through the external limiting membrane.

Bearing this general condition in mind, an examination of the
sections of graft 38 A 4 shows, with a few exceptions, a genc-r.il
conformability of its development to that of the normal. The
two greater cavities of the control are represented by one large
mesocoele. This is bounded by the basal and tectal portions of
the grafted tissues. The basal portion gives a rather compart
appearance with many fiber tracts. In the cortical n-gion tin-

CHORIO-ALLANTOIC GRAFTS.

307

differentiation is very striking. The two nuclear layers are
present and separated by the inner fiber layer which contains the
nuclei in definite strands. In tangential sections this layer also
show- many fiber tracts. In this case, this fiber layer also shows
a very heavy infiltration of Mood vessels, a condition not present
in the normal. Such a condition seem- to be correlated with

Al M

^/;

.

\'.: '££

\ ': . • ' ~TT

- 4^-f-BC.

• .

_^-«^Al M.

•• ,' .--;
- $ ~^3l

H M
B.V.

— E.L.M.
3.FL.

~^ON.L

1 ' ' '~-V^-li^ '

'

Mesoc.

l- 1. xlit

:i.il limiiinn 1 : : So-> ••!'• ^ •

alUint.'ii- In : .. blood vessel.

II. M. tit., I. '•* as

. 11. (i .alon of a forty-

i-i^h- il limiting • :il. 18 B 5. (Xl8o.)

i-ertain oilier dil'tereiu r- in the dc\ elMjunent of the parts of the
primordium. the mo-i evident of \\ hich is in the outer fiber layer.
In thi- i^raft (Fig. [0 . t he outer fiber layer is not bounded by an
external limiting membrane and as a result the libers penetrate
the adjacent me-i-nch\ me. The-. -til.. rown out into the

meseiu^hyme, or the me-eiich\ niatoii- ti— ue of the host has
advanced into the nervou- ti--ur of the i;raft ; the width of tin-
zone in which libers alone are prc-ent i- very variable. The
outer layer may even be lacking in many place-, the outer nuclear
layer bi-in^ in immediate contact with the loose mesenchyme
21

JO8 LEIGH HOADLEY.

cells of the host, between which the nerve processes make their
way. At numerous points along the periphery these processes
are united into nerve-like masses which end bluntly at varying
distances from their point of origin. Some of the cell bodies of
the outer nuclear zone appear to have migrated through the
immediately surrounding loose mesenchyme to a more definitely
organized envelope of mesenchymal tissue which surrounds the
graft and in which run the blood vessels of the host. The struc-
ture of this sheath resembles very closely that of the correspond-
ing layer in the normal embryo. It is impossible to say whether
it arose from the mesenchyme of the host or of the graft. The
blood vessels are those of the host and they show a most extensive
development. In places, large spaces filled with blood are
present between this layer and the developing mesencephalon.
Apparently because of the fact that the external limiting mem-
brane is not present in such places, the vascular infiltration of the
fiber layers as well as the nuclear layers is enormous. Even in
the mesocoele there are large groups of blood cells though no
organized vascular walls appear around them. In the nuclear
layers there is a crowding of the developing nerve cells due to the
presence of small vessels. In spite of this fact, the layers are
distinct. The individual cells of the areas seem to be differenti-
ating in a normal manner.

The significance of the absence of the external limiting mem-
brane becomes evident upon examination of another graft, 18 B
5 (Fig. 1 1), in which this membrane is present around a majority
of the cortical tissue. Where it is present, the differentiation of
the layers of the cortex corresponds exactly with that of the
control. In the peripheral fiber zone, the fibers form a clear band
and do not extend into the mesenchyme. As a result, none of the
nerve-like processes are found. The vascularization of such
regions is much less extensive than that of regions where the
membrane is absent. As a result of this, practically no distortion
of the nuclear and fiber layers occurs. These are normally pro-
portioned and contain cells which are in the process of normal
differentiation.

In other portions of the same graft there are areas lacking the
external limiting membrane. As wre should expect, the outer

CHORIO-ALLAXTOIC GRAFTS. 309

fiber layer is reduced or absent ; processes of the nerve cells
penetrate the mesenchyme, forming nerve-like structure:5; and
the layers of the cortex are very heavily vasculari/ed.

This graft shows a single large mesocce!' < - the other • -<

de-cribed. At one end of the sections of both graft-, where the
roof of the mesencephalon is cut tangentially. -ive fiber

tracts are visible in the inner fiber layer. The mesoccele in both
case- ;- iompletely eiicl..-ed by nervou- ii--ue, the brain having
fu-fd .it both ends. This was aided in all probability by the
apposition of the cu --tion. T; - »s morphology of

the organ ari-ing fn 'in -u< h a ^raft seems theret'i >re to i • . led

io ijiiiit: a con-ideral'' it by the en\ ironmeiit though the

differentiation of it- cellular constituents is specific.

The growth and differentiation of the mesencephalon i

!e ii|)on the chorio-.ill.mtoic membrane of the chick embr\o

brings out a number of "N cry interesting far* s. Phe basal and the

iei lal portion- retain their individuality though the two cavities

Mended into .me large nie-i >( ofle. The ba-al port ion ( le\ el"| is

more or U- UK ^ularK . dependent apparently on the mechanical

in\ol\ed .UK! tin- ab-ence of ascend: .-nding

L- when i-Maicd. T: .il portion, ho\ve\er, -ho\\ - two

type- of reaction, both of which have already been dc-cribid.

The iniere-i:: • about the development of tin- part i- toiind

in the form taken b\ the outer liber la; :dent on the

presence or absence "I t' • rnal limiting membrane. This

a !-• • tS the other la\ er- illa-UUH 1. a- thi- membrane appt

to regulate the peiiet rat ion <>t blood from the host

Mipi . Ph< growth and the differentiation of the various lay-
ers i- very -pctilic in both < I he nuclear cords and the
\\idth of the layer-, \\ith the e.\cepti<'ii of the outer fiber layer,
as ha- been mentioned above, are appro\imati-I\- normal.

It was mentioned above that in >ome of the graft- of the
mesencephalon and the na-al re-ion, thi-re are -mall peripheral
nerve-like bundle- pre-ent which do not correspond to any Mich
formation in the normal. The-e outgrowths of ner\'e are the
n-ult of the differentiation of the individual liber- of the nerve
cell-, but, owing to the fact that the membrane which furni-hes
a mechanical ob-tructioii to indefinite outward growth of

3IO LEIGH HOADI.KV.

these is absent, these fibers have continued to grow into the
mesenchyme, forming, in places, nerve-like bundles. Lewis
(1907 d) describes a similar outgrowth of nerves from the brain
tissue of Rana sylvatica from which he had removed the optic
primordium. He states (page 464) : "These nerves extend from
a region of the brain which, under normal conditions, never gives
rise to peripheral nerves, or at least to nerves that leave the brain
in the region to run into the mesenchyme. . . . The injury in
the brain in these experiments was done long before the nerves
normally appear." The data presented here seem to indicate
that the regulation of the growth of the elements is dependent
to some degree on the structural environment, and that the dis-
organization produced by the injury during operation is sufficient
to modify the gross structure obtained in development.

IV. DISCUSSION7

The independent power of growth and differentiation of other
embryonic parts has been studied to a limited extent by this
same method. Danchakoff (1921) has investigated the growth
and the differentiation of the blastoderm of the "primitive streak
stage" of the chick embryo after transplantation to the "allan-
toic" membrane. Blastoderms of "ten and more somites" from
which the posterior portion to the eighth somite had been
removed, were also grafted. In the blastoderms transferred to
the allantois at an early primitive streak stage, no differentiation
of organs appeared; the implanted parts remain as "islands of
undifferentiated tissue." Growths from blastoderms of a late
primitive streak stage showing a slight indication of the head
process, show groups of cells which "grow well and reach a high
degree of differentiation." Eye, nervous tissue, and notochord
are present, and in the cartilage which forms around the cord,
different parts of normal vertebrae are recognizable. Dancha-
koff concludes from her results, "the earlier a whole blastoderm
is transferred to the allantois, the less growth and differentiation
is obtained." While the principles involved in the growth ;uul
development of such transplantations are probably the same as
those applying to the independent differentiation of the isolatnl
organ primordium, the situation in grafts of such large portions

CHORIO-ALLAXTOIC GRAFTS. 311

of the embryo is very complex and for this reason an analysis
would be difficult.

More n-cently Atterbury (1923) has transplanted the metan-
ephric "anlage" of the chick in the same way. The material
u-ed in the experiments was taken from seven day chicks
trail-planted to the membranes of chicks of the same age. The
primordium of the kidney consists of an ureteric channel with
< ral tubule- and metanephrogenous tissue at the time
trail-planted. Tin- differentiation obtained from such grafts
after variou- period >i\\ th was similar to that of the control

melanephro-. The i-ohitcd metanephric primordium of the
chick embryo ha-, thu- the power of independent >clf-differen-
tiation as have the • idied in this report.

A genera] sun power of growth and differentiation

of isolal ' was made as a preliminary to

the piv-i -lit -tudy. The Drafts obtained from implant- of ci

n and the primitive streak confirm

the re-ult- t! M hakoft has obtained for the " bla-toderm in

tot embryo. Cartilage, muscle, ncrvi >u- ti--

-ue, nepi • ''ii- ti— tie, and epidermis differentiate well.
Some of the im somitic region show \cry definite

striation-. The cartilage suggests the form* it would have
Mimed in the normal embryo and the spinal nerve- ha\ rial

iiunt iml segmentation in the cord.

If the ii \\ith more or less extcn- m-

pK I with the results obtained in the invi ion

of the . 1 prim< .rdium of the single organ, i: r- that in

both i he differentiation of the graft i- • ilic

nature, after a < eriain amount of physiological dilferentiat:
taken place in the transplanted. Thi> physiological dif-

ntiationof the ct 11- of the organism h a tailed "i>roto-

plaMiii. :ali/ation" by Child (1921). In ^peakin.^ of "-elf-

di I tereiiti.it ion "in isolated parts of annelid and nmllu-k em In
he- 1. 120 "... protoplasmic specialization has advai

so far that the regulatory reaction- to isolation of part- are
limited, and the behavior of the part after isolation i- th«
much the saiiu- as before." Just when the speciali/ation of the
cells c>f the oixan primordium is est.ibli-hed in the chick embryo
matter re<|iiirin- further iiiv tion, though the n--ult-

312 LEIGH HOADLKV.

the present study indicate that it is at a very early stage. It is
conceivable that, for the eye, for example, specialization iiuy
occur after the earlier stages used by Danchakoff, for in her
experiments, the entire embryo with a part of the blastoderm and
not the isolated primordium was transplanted, so that physiologi-
cal isolation was not complete. The time of such organization
differs for different animal classes as has been shown by the
experiments on the independent development of isolated blasto-
meres, some cleavage being determinate from the very first with
no post-regeneration. The antero-posterior course of embryonic
differentiation in the chick indicates that the time of this speciali-
zation varies for different parts of the same embryo.

The results of the study of the organs described indicate that
the cells of their parts possess a specificity which is histologically
equal to that of the normal. That such a specificity is an innate
property of the cells of the embryo at certain stages has, moreover,
been shown by numerous investigators by the methods of tissue
culture, where the cells behave more or less as independent units.
The same principles apply to the development of the subordinate
parts of the organs grafted. This is emphasized by the dif-
ferentiation of the cristae acusticae of the membranous labyrinth,
the lens and the lenticular zone of the eye, and the various parts
of the nasal region and the mesencephalon. The development
of a subordinate part when isolated as such, is a problem, the
results of which must be established by further investigation.
That such parts are capable of differentiation without the stim-
ulation of external factors is proven by their behavior when
isolated in the primordium of the organ. Xo nerve stimulation
of the cristae acusticae is necessary for their origin, for example, a
condition closely resembling that found to exist in the formation
of the lateral line sense organs of the amphibian larva by Har-
rison. The results which have already been obtained indicate
that the same general principles are involved in the differen-
tiation of the parts of the other primordia which have been
employed in the transplantations.

The limits of development of the isolated parts seem to depend
on the mechanical factors involved rather than on the physiologi-
cal isolation as such. In none of the cases is this more marked

CHORIO-ALLANTOIC GRAFTS. 315

than in those where the otocyst and mesencephalon are used in
transplantation. Despite the disorganization produced by the
manipulation and by the forces acting on the graft during its
growth on the host, the results of the isolation are not marked,
differentiation occurin;sr in an approximately normal manner.
The absence of the external limiting membrane of the mesenceph-
alon, for example. doe- not alter the differentiation of the parts
•ich. but doe- modify the type of development obtained by the
on. a who'.

Be ring thea point- in mind, we may say that the-e iv-ults

empha-i/e the mosaic type of the development of the primordia

mined, after their origin in the embryo. The time of their

in varies, but once they are physiologically con-tituie<:

Mich, their i-olation from the adjacent structures has little i-liect

on the unfolding nature of the reali/ation of the potentialities

of their coii-tituent part-. The limits which are impo-ul on the

(le\elo|iinent of the-e transplanted organs are dependent <>n

en\ iiouineiit.il condition-, apparently largely of a mechanical

nature. I • nature of the development of the different cell- i-
app.itvntlv v< cific, but to what degree, remain- the subject

f..i further iiiNe-iuaiioii- which are l>eing made at the pre-eiit
time.

V. LITERATURE CITED.

Atterbury, R. R.

'23 1> Mctanepliric Anlagc of tl. in Alia:

Am. JOUMI. Ai: KI.

Beard, J.

'86 1 liial Sense Organs and tlK-ii A Ganglia in

1. lain. ; :t. Journ. Micr. Sci., XX\'I.

Child, C. M.

'21 1 .1 uii'l I )i-\ i-lupment of the NVt tem. I aiv. of Chi.

Pi
Cohn, Franz.

•03 lite des i es Htthnchens. A;.h. f.

Mik:. A:. .f... l.XI.
Danchakoff, V.

'16 K«iuiv.il'-ii' •• "' Hemopoietii An] -. .Journ, '•. t.. XX.

'22 Grafts in the Allantois of Embry^ni. An'.a^. •-; ol the CMiick. Anat. !•'

XXIII. Abstract.
Detwiler, S. R.

'22 I • \|>.-i imrnts on the Transplantation ut" Limbs of A mblystoma. Journ.
Ex| XXX\.

314 LEIGH HOADI.I.V.

Disse, J.

'97 Die erste Entwickclung clcs Riechnervcn. Anat. Helte IX., Abt. I.
Fischel, A.

'oo Ueber die Regeneration der Linse. Anat. Ileftc XIV.
Harrison, R. G.

'03 Experimentelle Untersuchungen ueber die Entwickelung der Sinncsorgane

der Seitenlinie bei den Amphibien. Arch. f. Mikr. Anat., LXIII.
'oya Observations on the Living Developing Nerve Fiber. Soc. Exp. Biol.

and Med., IV.
'oyb Experiments in Transplanting Limbs and their Bearing upon the Problems

of the Development of Nerves. Journ. Exp. Zool., IV.
"10 Outgrowth of Nerve Fibers as a Mode of Protoplasmic Movement. Journ.

Exp. Zool., IX.
'12 The Cultivation of Tissues in Extraneous Media as a Method of Morpho-

genetic Study. Anat. Rec.. VI.
His.

'89 Ueber die Entwickelung des Riechlappens und des Riechganglions und
iiber diejenige des verlangerten Markes. Verb. d. Anat. Gesells. a.d.
dritten Vers. in Berlin.
Kiyono, K., and Sueyasu, Y.

'17 Ueber die Implantation der Embryonalgewebe verschiedener Tierspezies in
die Hiihner und Entenembryonen. Organ der Med. Gesells. zu Kyoto.
XIV.
Lewis, W. H.

'073 Lens Formation from Strange Ectoderm in Rana syhatica. Am. Journ.

Anat., VII.
'O7b Experimental Studies on the Development of the Eye in Amphibia. Am.

lourn. Anat., VI.
'070 Experiments on the Origin and Differentiation of the Optic Vesicle in

Amphibia. Am. Journ. Anat., VII.
'oyd Experimental Evidence in Support of the Theory of Outgrowth of the

Axis Cylinder. Am. Journ. Anat., VI.
'076 On the Origin and Development of the Otic Vesicle in Amphibian Embryos.

Anat. Rec., VI., with Am. Journ. Anat., VII.
Lewis, M. R., and W. H.

'na The Growth of Embryonic Chick Tissues in Artificial Media, Agar and

Bouillon. Johns Hopkins Hosp. Bull., XXII., No. 241.
'nb The Cultivation of Tissues from Chick Embryos in Solutions of NaCl,

CaCh. KC1, and NaHCO3. Anat. Rec.. V.
Lillie, F. R.

'19 The Development of the Chick. 2d Edition, Henry Holt and Co.
Marshall, A. M.

'79 The Morphology of the Vertebrate Olfactory Organ. Quart. Journ. Micr.

Sci., XIX.
Minoura, T.

'21 A Study of Testis and Ovary Grafts on the Hen's Egg and their Effects on

the Embryo. Journ. Exp. Zool., XXXIII.
Murphy, J. B.

'13 Transplantability of Tissues to the Embryo of Foreign S|«virs. Journ.
Exp. Med., XVII.

CHORIO-ALLANTOIC GRAFTS. 315

Roux, Wm.

'88 Ueber das kiinstliche Hervorbringen halber Embryonen durch Zerstorung
einer der beiden ersten Furchungskugeln, u.s.\v. Yirchow's Arch., CX\"I.
Spemann, H.

'10 Die Entwickelunj; des invertierten Horgriibchens zum Labyrinth. Roux's

Arch.. XXX.

'12 Zur Enuvickelung <]a \YirbeItierauges. Zool. Jahrb., XXXII.
Streeter, G. L.

'07 Some Factors in the Development of the Amphibian Ear Vesicle and

Further Experiments on Equilibration. Journ. Exp. Zoul., IV.
'14 K.\|» rimental Evidence Concerning the Determination of Posture of the
Membranous Labyrinth in Amphibian Embryos. Jour. Exp. Zool., XVI.

I wi*h hen.- t-i • ay thanks to Mr. Kenji Toda for the preparation of

the figu:

CYTOLOGICAL CHANGES IX UNFERTILIZED
TUBAL EGGS OF THE RAT.

MARGARET C. MAXX,
LABORATORIES OF ZOOLOGY, UNIVERSITY OF CALIFORNIA.

INTRODUCTION.

Recent work on artificial parthenogenesis in vertebrates has
reopened the question as to the possibility of some degree of
parthenogenesis in mammals. Many authors have described
division comparable to cleavage in eggs of atretic follicles, but no
thorough-going study of the cytological changes in the tubal eggs
of mammals has been made until recently. This lack of obser-
vation was natural as long as ovulation was thought of as occur-
ring only at copulation, since under these circumstances the possi-
bility that fertilization might have occurred could not have been
excluded. Other work indicating that ovulation occurs at
regular intervals in several mammals including the rat quite
independently of copulation, lead to the question as to how far
the cytological changes of the unfertilized tubal egg of the rat
might parallel normal development. It \\as of interest to com-
pare the changes in the unfertilized eggs in atretic follicles, as
well as with eggs which have been artifically stimulated to par-
thenogenetic development. The conditions which have been
described in these three cases are so much alike that one is
inclined to believe that similar physiological changes are going on
in each. Whether these processes are wholly disintegrative in
the tubal rat egg; or represent an abortive beginning of partheno-
genetic development appears to be a matter of interpretation.
Several authors have stated, and this work corroborates the
observation, that mammalian eggs disintegrate in the second
polar spindle stage if fertilization does not occur soon after
ovulation. Many other authors have noted the occurrence of
divided eggs, simulating, at least, two, four, or more cell stages,
in the lower portion of the oviduct and uterus of mammals. It
is the purpose of this paper to trace the cytological ch.m^rs which
occur between ovulation and such apparent cleavage stages.

316

CYTOLOGICAL CHANGES IX EGGS OF RAT. 317

MATERIAL.

For the material used in the preparation of this paper I am
indebted to Drs. J. A. Long and H. M. Kvans. I also wish to
express my thanks to Dr. Long for helpful encouragement. The
material was prepared for a study of ovulation in the rat, Mits
norwegicus. The females from which the ova were obtained
were isolated before parturition and kept from males thereafter,
thus preventing the po— ibility of ova being fertili/ed. They
were killed at different intervals after parturition. The ovaries
and c\ id in t- \\eiv fixed iii Xenker's or in Benda'- fluid and .stained
in cannine, iron h.eniatoxylin, or phosphotungstic acid luenia-
\lin. The l.itit-r was by far the most useful -tain. The
I ;• nl. i niateri.il v illy valuable for study of the c\ \< >pla--

mic (.•lenient-, but it w.t- less useful than the Zenker preparation-
l"r detail- of nuclear -tincture, because the darkly -tainin:;
granule- .ind niitoc hondria tend to obscure the chromoMime-
e-pri ially in the period of scattering. In the fifty oviduct-
which were -tudied eggs were found in the di-tal portion in
t\\el\i in the middle upper portion in ten, in the lower-

middle portion in -r\eii, and in the proximal portion in nineteen.
In all two hundred and fifty-two eggs were examined. As many as
nirit .uixl in some oviducts, but three, four, ,md h\e

Were n- ••• d nruch nnue commonly.

OBSERVATION

\t o\ ulaiion the -ei mid polar -pindlr lie- parallel to the periph-
ti\ of ilu- egg \\ith tin- d\ad- arranged upon it in t\-pical final

prophase position. In any one oviduct all of the ova are al about

the -aim- Stage. fable I. demonstrate- the fact that the com-
moner stages -ho\\ a natural -e«iiieiice \\ith re-|u-ct to po-iiion in
the o\idtict be^inninv; \\ith di-tortioti of the -pindle and 5i
teiin- of the chi Mini •-• >me- and ending with the imei|iial fragmen-
tation of a multinucleate cell. It i- e\ident that \\ e are tlui-
pro\ idi-d with a natural -eriation of c\ tolo^ica cat

value in interpretation of the re-nit.-.

Stage I. — It v ill be noted from Table I. th.it by far the great
number of ova found in the distal third of the oviduct \\ ere in the
• nd polar- spindle stage. The -pindle may, however, be
-li-htly turned or even at right an^le- to it- original po-ition

318

MARC.ARET C. MANX.

parallel to the periphery. In a few cases some of the dyads have
divided and the monads have moved apart from one another on
the spindle fibers. First polar bodies are often found, but no
evidence of second polar body formation was noted. The cyto-
plasm shows no marked differentiation, consisting of a rather
coarse reticulum enclosing a few large yolk globules and many
small granules. Some of them stain deeply and are found in
chains. These may be mitochondria. The zona pellucida is

TABLE I.

SUMMARY OF CYTOLOGICAL DATA AS RELATED TO POSITION IN THE OVIDUCT.

Position

Num-

-•

Series.

in

ber of

Cytclogical Notes.

Oviduct.1

Eggs.

318

Distal

3

3 with second spindles slightly rotated in all.

345

* *

9

2 with second spindles parallel surface. I perpen-

dicular; 6 with chromosomes scattered.

383

« t

4

4 with spindles parallel surface.

337

*'

3

i with spindles parallel surface; i perpendicular;

i with chromosomes condensed.

227

• t

3

2 with spindles parallel surface; slightly rotated

in i.

10

"

4

4 with spindles parallel surface.

I.

103

< I

7

Second spindles show some scattering, periph-

eral cytoplasm vacuolate.

no

4 1

6

4 with spindles parallel surface; 2 show some

scattering.

III

1 1

4

4 with spindles parallel surface.

18

2-5

2

2 with blunt spindles parallel surface.

33

2-5

3

2 with 2d polar spindles parallel surface; i con-

tains i large and i small spindle.

"3

3-5

6

5 with 2d polar spindles parallel surface; i

central.

102

3-5

i

i with 2d polar spindles parallel surface.

109

3-5

3

3 with 2d polar spindles parallel surface.

115

4

9

5 with 2d polar spindles parallel surface; frag-

mentation in others; distortion.

35

4

3

i with 2d polar spindles parallel surface; 2 with

spindles and vesicles:

26

4-5

i

I with chromosomes scattered, vesicles forming

about them.

II.

22

4-5

2

2 with chromosomes scattered.

367

4-5

3

3 with chromosomes scattered.

107

5

3

i with chromosomes scattered; i second spindle

intact; I fragmenting.

8

5

3

I with chromosomes scattered, broken spindle;

2 second spindles.

5-25

i

i with chromosomes scattered, distorted spimllr

114

5-5

5

4 fragmenting, multinuclcate; i with 3 large

vesicles.

347

5-5

7

6 with chromosomes scattered, I with I large

group; i fragmenting.

4

6

2

2 multinucleate.

CYTOLOGICAL CHANGES IX EGGS OF RAT.

319

•ion

Num-

Stage.

Series.

in

ber

Cytological Notes.

Oviduct.'

III.

325

7 I

i with 4 equal uninucleate cells.

308

7 I

I multin . :ragmenting.

17

7o 4

4 multinuclt

-47

8-5 3

3 multin

403

8-5

9 multin lulling irrespective of nucleus,

ming.

12

8.5 3

2 multin ; >indle parallel surface.

354

8-5 5

3 multinucleatc and fragmenting; 2 with scat-

8-5

:iu-nting irregularly.

•• or more nuclei; ir-

9 I

I inulti:

9

.ting with u: ijularity. 2 4 cells.

9

M.i ate fragments.

3

crating.

2

i tin: :p of 4 equal cells in a

9 I

I \\ uh 4 multin;.

9

5

1 I In- : i the oviduct; thus stage I.

in> li. •. i.*.. from 1-3.5. while gc is the

•

Ill

1

U

!U.

Cells.

Fr.i .

in i '•

Central
Condensed.

;">

I-}, v •

4

2

15

O

o

o

2

J-C .

6

•»

i

o

i

i

7

i

i

o

2

O

6

22

2

Total

-Ji

4

16

20

22

2

2

r\ i.lcnt in .ill of the- Hnid.i .md in -(-me of the /mki-r preparations
.1- ,i \\iili- In mil '-run Hi- l.ivrr. \\'hy it should have been pn--
served in some <>f the /rnki-r pri-p.iratioiis .md dissolved off in
dilu-r- i- nut undiT-tixid. The- fdllick- cells appear to be fjuite
normal.

All of the eggs in twelve oviducts contained -pindlcs either
p.ir.tllel to the peri])her\- or ><mie\\ h.it perpendu nl.ir to it. In
all, 47 e^s were found to be in this stage and of ilu-<- forty were
in the upper half of the oviduct.

J20 MARGARET C. MANX.

Stage II. — In the upper middle portion of the oviduct the
spindle is beginning to break down and the chromosomes are
becoming scattered. In most of the unfertilized eggs studied the
second maturation spindle remains parallel to the surface of the
egg and there breaks down, the fibers splitting off and carrying
the dumb-bell shaped chromosomes with them. This process
varies in completeness in different ova. The spindle may be
evident after most of the chromosomes are vesiculate, or it may
disappear almost completely as the chromosomes scatter.

Eggs were found in the upper middle portion of the oviduct in
ten cases, and in all forty eggs had broken spindles and scattering
chromosomes. Thirty-one of these were in the upper half of the
oviduct. The breaking down of the spindle results, at times, in
a simulation of multipolar spindles but their irregularity and the
fact that the chromosomes are usually widely scattered indicate
that they result from the breaking down of a single spindle.
Sometimes the spindle, when quite degenerate, is no longer
fibrillar but appears as a dense homogeneous mass of material of
a very distorted shape.

In two oviducts eggs were found in which the spindle was
central, and the cytoplasm about it unusually dense. The
spindle appeared shrunken and the chromosomes had coalesced
into a compact mass. Such eggs were also found well down in
the oviduct.

Stage III. — The vesicle formation which began in Stage II.
proceeds until multinucleate cells are formed. A liquid gathers
about a single chromosome or a group of them. The chromo-
somes then swell and break up into large and finally into small
granules so that the nuclei assume the typical resting condition.
They vary greatly in size and number, different cells in the same
oviduct containing from one to ten nuclei which differ greatly in
size. The appearance of the nuclei differs with the fixative used.
In eggs fixed in Benda the chromatin lines the inside of a clear
vesicle, most of it being in a mass on one side. In eggs fixed in
Zenker the chromatin appears as granules on a net. Central
spindles in different stages of degeneration were found in three
oviducts. In one of these one egg contained a central spiiidK-
while the other three resembled typical resting cells. In the

CYTOLOGICAL CHANc.F.S IX EGGS OF RAT. 321

other two, five eggs contained degenerating r-pimlles while the
rest were in an advanced stage of fragmentation.

Crai k- often appear in the cytoplasm but the zona is >till
u-ually intact at this stage. The follicle cells are often grouped
into small syncytia containing many small crescent shaped nuclei.

The eggs of stage three were mostly found in the lower middle
portion of the oviduct. In all thirty-six ova were multinucleate.
Of these sixteen were unfragmented, and twenty were beginning
to fragment.

Stage IV. — I-"ra;.;mentation U-^in- almost simultaneously with
• K- formation but i- u-ually completed in the most proximal
portion of the o\ id in i. Tin- fragments vary in number from two
or three to main . Tin \ may b- < >ximately equal in size or

very tmci|iia!. A fragment may contain one to several nuclei or
none. I Lamentation occur- quite independently of the vesi
when mai 1 may take place either by a process

of lobin^. « .r. m« >re > . ,mnn >nly. by the appearance of cracks in the
c\ topla-m.

- 'me of i he o\ .1 iii i In- IOUIT portion of the oviduct resemble 2,

nd 4 cell ely. Three two, one three, and

loin- cell 'iind. In oviduct 32 only one egg in live

r« -mil. led a nonn.,1 condition, the rest being multinucleate and
more or less fragmented. In o\iduct 37 two two and one four
cell stage \\<i« found, while two i-^gs were multinucleate i
.mcntcd. Nuinbi-i .ntained very degenerate Ira^mnits

and one four tdl 1 he c\ topla-m is very degenerate in

in, -i of ti • iirntii'. \ /oiia may or may not be pre-ent

but if it ap|n-ar- at all it i- nearly always ruptured. The !
meiii- are usually (|iiiie d»-\oid of (o\erin- far down in the tube,
and it i> ini|>o->ibU- to tell how many ova they come from. The
follicle cell- are now empty -hell- devoid of protopla-m.

romplc-te fragmentation was the rule in nineteen ovidu-
All of the-e \\ere found in the lower third of the tube. In all

twenty-two ova had n 'but this number does not

repre-ellt the total -ince in -e\'eral ca-es degeneration had pro-
, ' eded SO far that the figure- for them had to be omitted.

322 MARGARET C. MANX.

LITERATURE AND DISCUSSION.

The observations upon atretic and unfertilized tubal eggs of
mammals agree essentially but the interpretations are rather
various especially with regard to atretic eggs. This is easily
comprehensible since no natural method of sedation is provided
in atretic material. On the other hand one can easily determine
the position of tubal eggs in the oviduct, and as these data show,
can then demonstrate that such eggs pass thru a regular series of
changes as they descend the duct.

In eggs which are liberated into the fallopian tube the second
maturation spindle sometimes rotates to a position at right angles
to the surface. The dyads also divide occasionally so that the
second anaphase really appears to begin in some unfertilized
eggs. In by far the greatest number of eggs in the upper third of
the oviduct, however, the second polar spindle is parallel to the
periphery and intact, with the dyads still in late prophase posi-
tion. The spindle may or may not be somewhat broken. Xo
cytoplasmic changes were noted at this stage. In eggs in two
oviducts the spindle was central in position.

Three of the facts cited above might be interpreted as indica-
ting that the cell processes may not all be wholly degenerative at
this stage. Firstly, the spindle may rotate as if in preparation
for the second polar division. Secondly, some of the dyads may
divide, and thirdly, the spindle may be found in the center of the
egg- Eggs with central spindles all showed one of two degener-
ative conditions; the chromatin simply clumping into a single
mass and the cytoplasm shrinking and becoming denser, or in
other cases the cell wall may break and the very vacuolate
cytoplasmic contents scatter. There is no indication in the tubal
rat material that normal cleavage ever occurs in these eggs altho
it may of course be possible that more material might include
such stages since a few eggs give the appearance of approximately
normal early cleavage stages. In by far the greatest number of
eggs degeneration begins with the failure of the second polar
division. It is of some interest that the formation of but one
polar body is a characteristic feature of development of nn>-t
naturally parthenogenetic animal eggs.

Degeneration has been shown to begin with the breaking "I the

CYTOLOGICAL < EANGES IN" EGGS OF RAT. 323

spindle and the consequent scattering of the chromosomes
throughout the cytoplasm. Kingery ('13) described the same
type of degeneration for eggs of the mouse. The cyto-

plasm and diromosonie- inter that the latter at first appear

to he surrounded by a « lear liquid. They then swell and form
little \e-ides resembling nuclei. Where several chromosomes
guous lar-e V( -ult. ("harleton ('i/) described

nueleu- "f tubal mon-' farming from the entire

chrom- Mp, which th- to a multimicleate cell

liy l<.l»inu. Tin- -eriation ht •:• nU-d demonstrates that this

in the- r.it. It" 'ins nuclei were

fu-ing, picture- -ueh a- ( 'harleton ! rihed as due to lobing

\\< .uld be <.bt. :iiu-d. \\ i M -ribed syncytia in

artificially p.irt hem i^eiiet i< Poxopneusl _s as resulting from

.i failure «.i" cytokinesis following nuclear di\i-ion by a mitotic
proo I • ommon a] broken spindles and scat-

tin in the upper p.irt < .1 the oviduct as well as the
different e in ~i/e d" the \ esi( !e~ indi y that this is not

the case in tub.i!

\o true multipol.n- -pindle-. inn; or cytasters such

were de-.-riU-d b> Ht-niii . 5) for atretic rat eggs, by

Newman i; F the armadillo, and in artificially

I l-\ Wilson (*oi) were found.

( )n, in., p pindles at opposite poles of the cell.

I. ,nn- 'i ; :.nier|ireted .1 -imilar condition in a tubal egg of the
-ui hat tl polar body had no: I

dct. tched from i . but note- that a similar case could not be

M. inierprete.l -inee the fn>t polar body was clearly present. In
\ie\v of t1 1 i>olar spindle of tubal rat «

r-oi ue time- -n.i|)- in tu I elongation and di-tortion it

seems best to coii-iiler -uch picture- as originating in thi- manner.
1 )uring the MMtterin^; of the chromo-omr- the cytopla-m In-, omes
more co.n>ely ^r.inuKir and the nu>he- more opi-n. \\il-on
i ibi-d the c\ topla-m of //;•/;/;• artificially parthenogenetic
ser b\- clumpin- of micro-onie- preceding
a-ter formation.

< >nc cell was found with nd one with two re-ling nuclei.

They are probably the result of chan - of chromi

324 MARGARET C. MANX.

into one or two rather than into many masses.1 When few nuclei
are present they tend to occupy the center of the egg. Eggs
containing resting nuclei were lower in the oviduct than any
of the eggs containing spindles, hence it is probable that spindles
never form following the disintegration of the second polar
spindle in tubal rat eggs. Newman ('13) interprets his atretic
armadillo material as indicating that new mitotic figures with
asters form following a return to the resting condition after the
failure of the second polar division, but none of the rat eggs with
central spindles showed any indication of astral radiations.

There is no indication in the rat material that fertilization by a
polar body ever occurs. The first polar body is completely
separated from the egg before ovulation takes place.

Fragmentation occurs almost independently of the newly
formed "nuclei," so that some of the resulting cells contain nuclei
and some do not. Newman ('13) and Charleton ('17) describe a
similar cutting off of anucleate portions of the egg and consider
it as a form of deutoplasmolysis. Newman believes that such a
process would be advantageous to a cell system governed by a
haploid nucleus. Since in the rat eggs the distribution as well as
number and size of the "nuclei" is haphazard in the fragments it
seems improbable that the process should be considered normal.
In the tubal rat material it appears that the fragmentation
process is less abnormal when few nuclei are present in the egg,
and that it may at times be fairly normal is indicated by the fact
that a few normal-looking 2, 3, and 4 cell stages were found. In
his observations on unfertilized eggs of Toxopneustes Wilson
('oi) states that most of the blastomeres resulting from a simul-
taneous cleavage of syncytia contain nuclei but that some do not.
Here also the fewer nuclei present, the more normal the division.
Thus in tubal rat eggs as in artificially parthenogenetic one>, .1
few may have more vitality than the rest, and these may possibly
be considered as possessing a very limited parthenogenetic
capacity.

In the lowest portion of the oviduct all of the ova were obvious-

1 The possibility that sperm may live in the oviduct for several days and then
be capable of fertilization should, however, be taken into c<mM<ln.iti<>n in inti-i-
prcting these occasional normal looking cases. \ViU«>n (,'oi) also found some un-
fertilized eggs which resembled fertilization stages.

CYTOLOGICAL CHANGES IN KGGS OF RAT. 325

ly degenerate. They might be expected to enter the uterus, if at
all, as mere shells and scattered cytoplasm. The zona is either
broken or absent at this stage.

CON< i.i -ION.

I "Howing ovulation most of tin- changes undergone by tulul
rat eggs are distinctly abnormal. The central position of the
spindle, the occurrence of three uni-. and one binncleate cell, the
coar-enin- « .f cytopla-mic r-truct tire, the < >1 »MT\ ati< >n that division
is more or le— « regular acn.rdiiu '" the number of " nuclei" which
are pre-ent, and tin- discovery i-j" a fe\v two. three and four cell
Stages may be thought t<> indicate -••me tendency towards normal
de\elopniein in certain of the nnt'eriili/ed eggs. < >n the con-
trar\ , the t.n I that the central -pindle i- ran- and always degener-
ate, that only three uninucleate cell- were found, the i:reat pre-
ponderain e of multiniH leate cell>. and the fact that only very
enerate fragment- are found near the uu-rii-. make it neces-
Bary to < onclnde that although certain auto-re^ulati\ e processes
may OCCtir in a few ova, a <letmiti- t\pe of degeneration is the

rule.

lUMI.I. x.KAI'UV.
Athias, M.

'09 I-i-- |.lii'ii'iiin-ii ' .;aaf en \

il';itii--i-- i \\>-f. !•• I • .34: 1-23.

Barfurth, D.

'96 \ViNtnh iiln-r ilif I : .1. i. Entwickl..

Bonnet, R.

'99 l.i'.t Ct bel \\ :: '•••Ilirrrll I' : ' it. U. Entwickl..

9: .v
Bischoii, T. L.

*44 Sur la m.itur.itinu <-t la > lain- jx-rt.»li<, . noinme et des mam-

iniirn'-i iniIi-|M-ii(aiiii-i!.' Ann. Sci. Natur.. 3:

p. ; i ••.'. > : U
Charleton. H. H.

'17 i-.it'- ••: the Unfertili* M BIOL. BULL.. 33:

•«-338.
Flemming, W.

'85 i"l>iT<li«- Bililunn \iin Ri. hti. - rn t..-im I'tuergang

Gra.iti- II-T Fullik<-l. Ar.h i. An.it. u. I-.'iitwicklutu-^ . I 1885.

221-244.
Hartmann, C. G.

'16 Development of the Opossum. Jnurn. Morph.. 27: 1-83.
Heape, W.

'05 Ovuluti'in ami Degeneration of the Ova in tiie Rabbit. Proc. Roy. Soc.
LonMun, ~(> b.

326 MARGARET C. MANX.

Henneguy, L. F.

'94 Rt'cherches sur 1'atresie des follicules de Graaf chez les mammiferes et

quelques autres vertebres. Journ. d. 1'anat. u. physio!., 30: 1—39.
'95 Lccons sur la Cellule. Paiis, Georges Carre, Editeur. Trentieme Lecon,

466-478.
Hensen, V.

'69 Uber eine Zuchtung unbcfruchteter eier. Med. Centralbl., 7: 403-4.
'75-6 Beobachtungen Uber die Befruchtung und die Entwickelung des Ka-
ninchen und Meerschweinchens. Ztsclir. f. Anat. u. Entwickl., i:

2I3-273-
Herlant, M.

'13 Etudie sur les bases cytologiques du mechanisme de la parthenogenese ex-

perimentale chez les Amphibicns. Arch. d. Biol., 28: 506-608.
Kindle, E.

'10 A Cytological Study of Artificial Parthenogenesis in Strongylocentrolus

piirpuraliis. Arch. f. Entwickl., 31: 145-163.
Huber, G. C.

'15 Development of the Albino Rat (Miis narvegicus) I. and II. Journ. Morph.,

26: 247-386.
Janosik, J.

'97 Die Atrophie der Follikel und ein seltsames Verhalten der Eizelle. Arch. f.

mikros. Anat., 48: 169—181.
Kingery, H. M.

'15 So-called parthenogenesis in the White Mouse. BIOL. BULL., 27: 240—258.
Kirkham, W. B.

'07 The Maturation of the Mouse Egg. BIOL. BULL., 12: 259-265.
Kirkham, W. B., and Burr, H.

'13-14 Breeding Habits of the Albino Rat. Amer. Journ. Anat., 15: 291-317.
Lams, H.

'13 Etudie de 1'oeuf de cobaye. Arch. d. Biol., 28: 229-323.
Lams, H., and Doorme, J.

'07-08 Maturation et fecondation de 1'oeuf des mammiferes. Arch. d. Biol.,

23= 259-343.
Long, J. A., and Mark, E. L.

'n The Maturation of the Egg of the Mouse. Publ. Carnegie Inst. Wash.,

142: 1-72.
Lau, H.

'94 Die parthenogenetische Furchung des Huhnereies. Inaug. Diss. Jurgen.

Dorpat.
Loeb, L.

'01 On Progressive Changes in the Ova in Mammalian Ovaries. Journ. Med.

Res., N.S., i: 30-46.
Oelacher, J.

'72 Die Veranderungen des unbefruchteten Keimes des Hiihnereies im Kilritrr

und bei Bebriitungsversuchen. Ztschr. f. wiss. Zoo!., 22: 181-234.
Newman, H. H.

"13 Parthenogenetic Cleavage of the Armadillo Ovum. BIOL. BVLL., 25:

52-64.
Rabl, H.

'97 Zur Kenntnis der Richtimgspinddn in degenerierenden
Sitzber. d. k. Akad. d. Wiss. Wien., 106: 95-!"'..

< VTOLOGICAL CHAN(.l-> IN KGGS OF RAT. 327

Rubaschin, W.

'05 Uber die Reifungs und Befruchtungs prozene des Meerschweinchenscies.

Anat. Hcfte 29: 507-553.
'06-07 ' I»T <li<- Vi-r.irnli-ningi-n der Eier in den Zugrunde gehenden Graaf'schnr

Follikeln. Anat. Hefte. 32: 257-276.
Schulin, K.

'81 Zur morphologie des Ovariums. Arch. f. mikr. . it., 19:442-51.'.
Sobotta, J.

'95 iJi.- Hi-iruchtunv: uml Futi hung des Eies der M.r:-. Arch. 1. Mrk;

Anat.. 4^: i.v -.
'96 Dii- I-nr«huMK -l<-~ \Virl.«-ltii-rcies. Ergehn. :\. Anat. u. Fntwickl.. 6:

493-500.
Sobotta, J., and Burckhard, G.

'10 Ki-ifuiiK urnl Hdrii. htiini; •]'- Eies der \V<-i--«-n Rattr. Anat. U.K.- 42:

. .-.

Spuler, A.
'oo 1'ln-i i|i«- reilungaeracheinungen der Eizellen in .l<xi •n.-ii«T<-n.|«'n l-'olli^.-ln

• I. iun AJI.T Ili-ite 16: K-- i i ^.

Tafani, A.
'89 La fa 1'ii'ia/ii'in- ' • DC I'tuiiii-- 'I. in- !• \n it.

ilalii-niir, .1. Iniil., i i : r i _• 117
Wilson, E. B.

"oi |-.\|ii-rii!n-ntal --l U'l . Aii li I. '-ntu i. U.. i .' : '-. ^>

BIOLOGICAL BULLETIN

OF THE

noannc Biological laboratory

VOL. XLYI JANUARY, 1924 . No. i

• « INTENTS

KOPEC, STEFAI i . '"/>-

at and ' i

1 I KM K, ( . I I . .1

; -

I1' i i aasmo M«>N UII.Y nv THE

MAklNl- i:i<>! OGICAL LABORATORY

PRINTED AND 1SSI El> BY

i INCASTER I-KI ss, i\

LANCASTER, r.\.
AGENT FOR GREAT BRITAIN

WHKI.DON & \\ y, I.IMIII i.

2, J anrf 4 Arthur . (Oxford Street, London, W. C. 2

Single Numbers, 75 Cents. Per Volume '6 numbers), S3.OO
Entered October 10. iooa, at Lancaster. I'a.. as second-class matter under Act of Congress of July 16, i8p4-

lEfcitortal Staff

E. G. COXKLIN — Princeton University.

GEORGE T. MOORE — The Missouri Botanic Garden.

T. H. MORGAX — Columbia University.

W. M. WHEELER —Harvard University.

E. B. WILSOX — Columbia University.

Efcitor

FRANK R. LILLIE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. I5th to June I5th, or
\Yoods Hole, Mass., June I5th to Sept. I5th. Subscriptions and other
matter should be addressed to the Biological Bulletin, Prince and
Lemon Streets, Lancaster, Pa.

BIOLOGICAL BULLETIN

THE

noarinc Biological laboratory

WOODS H< >!.h. .\\

VOL. XL\ I I EBRUARY, 1924 No. 2

CONTENTS

\l\-l, 5. ( ' . \M' .I/ '

I'l - II. I.. <

M \. \i.' i IIIK. I..M\ \V. .1 >:

//::

Entbr

\ll.i N, I i". \K.

HONDA, HIKOKUI

I1' i : Mos MII.Y II V THE

MAKINI-: i:l<>i OGICAL LABORATORY

PRINTED A liY

I \\( A-ri K F'KI SS, I
I VNCASTER, PA.

AGENT FOR GREAT BRITAIN

\VHELDON & \VI -I I V, F.IMIII n
2, j and 4 Arthur S!T,,!, New < '.\; • • /, London, If. C. 2

Single Numbers, 75 Cents. Per Volume 6 numbers*, S3.OO

••••^•••H-HH— ^^^^^^^••— •••••••i— _B_^^^«HB»«i^— ^— ^-^B— l^^MBI

Ent«rtd October 10, ipoa. at I-ancaster. Pa., as second-class matter under Act of Congreaa of July 16. 1894-

Editorial Staff

E. G. CoteKLIN-^-Princeton University.

GMORGK T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN — Columbia University.

\\ . M. \YIII-:J:LKR — Harvard University.

E. B. WILSON — Columbia University.

FRANK R. LILLII-; — The University of Chicago.

All communications and manuscripts should be sent to the Man
aging Editor, the University of Chicago, Sept. I5th to June I5th, or
Woods Hole, Mass., June I5th to Sept. i5lh. Subscriptions and other
matter should be addressed to the Biological Bulletin, Prince and
Lemon Streets, Lancaster, Pa.

BIOLOGICAL BULLETIN

OF THE

noarinc BtolOQical laboratory

H< M.L. M

VOL. XL\ I MARCH, 1924 No. 3

CONTENTS

YOCOM. HAM Y I'. Lutetil C< :d of th • KM

\\\ \ !'.i i • i I ). /;:.' ' ' /« /?<7(;-

//' : n-

• . Kid

w, THURLOW ( / >v .!/: •• r /.<mv 1. 1 ;

1'riu.isjiEii MONTHLY BY TH«
MAR INK lUoLOGICAI I AI'-< )RA TORY

PRINTED ASH ISSrED BY

I \\CAsTI R PR! SS, tMC.
! VNCASTER, PA

ACBNT FOR GREAT BRITAIN

WHKLDON & \YKSLEY, LIMITI-D
2,2 and 4 Arthur Street, .\ttt> Oxford Street, London, W. C. 2

Single Numbers, 75 Cents. Per Volume -6 numbers). S3.OO
Entered Octobw 10. tooa. at Lancaster. Pa., as second-class matter under Act of ConereM of July 16. 1894-

EDitorial Staff

E. G. CONKLIN — Princeton University.

GKORGK T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN— Columbia University.

W. M. WHI.KLKR— Harvard University.

E. B. WILSON — Columbia University.

FRANK -R. LILLIK — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. 151)1 to June I5th, or
\\'oods Hole, Mass., June 151)1 to Sept. 151)1. Subscriptions and other
matter should be addressed to the Biological Bulletin, Prince and
Lemon Streets, Lancaster, Pa.

BIOLOGICAL BULLETIN

fiDannc 3i3iolocueal laboratory

\V« >< >!>>> HI '! I -. \\ \NS

VOL. XI A I

APRIL. [924

\n. 4

Mi [/, CBAS, \\'..

-.Z, J»~K 1

PACK \i ( \KI.ES.

( i i.\ i i \M>, I.. K.

CONTENTS

Th, '. <uid Chr»>

jomo Dnrir in the Rob-

ber Fly, La5/. :<iritttit:<

S !>tibility of O//.> / - R-:.' :m RaJ

..

iologica . Rdat:

ships Bftuven l> -i of

md :h' :h Special Rcfcr-

Kollar ...... 178

M' >s mi > ii\ i UK

MAKIM- BIO1 « <HCA1 1 VBOR \ l« >RV

PRINTED A ED BY

I ANCASTER I'Ki 5S, INC.
I \NCASTER, I'.\.

AGENT KOR GREAT BRITAIN

WHI'I DON & \VI.SI I-.V, l.iMinu
2, .? and 4 Arthur Street, .Yr-ir Oxford Street, London, W. (.'. 2

Single Numbers, 75 Cents. Per Volume 16 numbers). S3.OO

Fnt«r^d October 10. IQOJ, at I anr utfr. Pa., as second-class matter under Act of Congress of July 16. i8o4-

Editorial Staff

E. G. COXKLIX— Princeton University.

GKORGK T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN — Columbia University.

\\ . M. WHKKLKR — Harvard University.

E. B. WILSON — Columbia University.

FRANK R. LILLIE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. I5th to June I5th, or
Woods Hole, Mass., June I5th to Sept. I5th. Subscriptions and other
matter should be addressed to the Biological Bulletin, Prince and
I.rmnn St reels, Lancaster, Pa.

BIOLOGICAL BULLETIN

OF THF

flDaunc Biological laboratory

\N< >< >! .^ H< 'I

\'OL. X1A I

M \v. 1924

\ i I,\MI, L. \\.

KlNDKKD, J AMI > I .

' kX)D&H ii, HUBERT 1

CONTENTS

The V> .nd Symbiotic Relation-

^3 bet- tinal Protozoa of

Termites and their ll<'-t, u-ith Special

R ft-r . /?./.<•;,.'• :lnr

The C Penvi^ccral

Fluid of Echinv I ^28

Cell Be . . . . . 252

\7,

I'l I M'lN mi -, MY IKK

MAR IN! i:H'| QGICAL I \n«l< \loRV

PRINTED A HD BY

I ANCASTER I'Ki SS, Ii

LANCASn-R, PA.

AGENT FOR GREAT BRITAIN

\VIIFI. PON >!v WI-.SI.KY, I.IMIII D
2,3 and 4 Arthur Street, \e-t- Oxford Street, London, W. C. 2

Single Numbers, 75 Cents. Per Volume '6 numbers). S3.OO

1 ncrrd October 10. iooa. at Lanc.uter. Pa., as second-class matter under Act of Congress of July 16, 1894-

Editorial Staff

E. G. Cox KLIN — Princeton University.

GEORGE T. MOORE — The Missouri Botanic Garden.

T. H. MORGAN — Columbia University.

\Y. M. WHEELER —Harvard University.

E. B. WILSON — Columbia University.

EDftor

FRANK R. LILLIE — The University of Chicago.

All communications and manuscripts should be sent to the Man-
aging Editor, the University of Chicago, Sept. I5th to June I5th, or
Woods Hole, Mass., June 15th to Sept. I5th. Subscriptions and other
matter should be addressed to the Biological Bulletin, Prince and
Lemon Streets, Lane-aster, Pa.

MBI. WH01 LIBRARY
III I I' "

17KJ A